€¦ · INTROD UCTIO N. The Congress Auxiliary of the World’s Colum bian Exposition was organized by authority and with the support of the Exposition Co r po r a tio n for the

U. S. DEPARTM ENT OFAGRICULTURE,

WEATH ER BUREAU .

BULLETIN NO . 11 .

R E P O R T

OF THE

NTERNA’I‘lONAM ORO0 A ONR

HELD AT

CHICAGO, ILL.,AUGUST 21-24 , 1893 ,

UNDER THE AUSPICES OF THE

‘

Congress Auxil iaryof theWorld’

s Columbian Exposition.

EDITED BYOLIVER L. FASSIG ,

SECRETARY.

Pu bl i s h e d by a u t h o r i t y o f t h e Se cr e t a r y o f Ag r icu l t u r e .

WASHINGTON, D . 0

WEATHER BUREAU .

1894 .

ART II is in pre ss . It wil l co n ta in th e pape rs Of

SECTION IV. H isto ry a n d Bibl io g ra phy.

SECTION V. Ag ricu l tu ra l M e te o ro l o gy.

SECTION VI. Atm o sph e ric El e ctr icity a n d Te rre stria l

M a g n e tism .

LETTER OF TRANSM ITTAL .

U . S . DEPARTM ENT OF AGRICULTURE,

WEATHER BUREAU,

Wa sh in g to n ,D . C .

,Sep tem be r 9 , 1893 .

SIR : I have the honor t o transmit herewith a document entitledReport Of the Inte rnational Mete orological Congress

,held at

Chicago,Ill .

,August 2 1-24

,and to recommend its publication

a s Weather Bureau Bulletin No . 11.

Very respectfully,

MARK W. HARRINGTON,Ch i ef of We a th e r Bu r e a u .

Ho n . J . STERLING MORTON,

Secr e ta ry of Agr icu l tu r e .

INTRODUCTION .

The Congress Auxiliary of the World ’ s Colum bian Exposition wasorganized by authority and with the support of the Exposition Co r

p o r a t io n for the purpose of bringing about a series o f conventions Ofleaders of the various departm ents of hum an thought .

The various congresses held their sessions in the Memorial ArtPalace in the city Of Chicago

,from May until October , 1893 ; those

in the Departm ent of Science and Philosophy were assigned to theweek comm encing August 2 1. In this department provision wasmade for a congress o n Meteorology

,C l im atology

,and Terrestrial

Magnetism . In Novem ber,1892

,the President o f the Congress Aux

il ia ry, Mr . C . C . Bonney,invited the Chief o f the Weather Bureau

to organize such a congress . In accordance with this request, I calleda conference o f gentlemen t o consult with m e in the arrangement o fa programme . The following persons responded to the call and metme at my office o n December 2 1 : Professors C leveland Abbe

,F . H .

Bigelow,Thomas Russell

,C . A . Schott

,Lieut . Commodore Richardson

C lover,and Mr . O . L . Fa s sig .

As a final result o f the conference the organization indicate d onpage iv was effected and the programme shown in the Table Of Co n

/

tents was arranged . The papers t o be submitted were to be of astrictly scientific character . Authors o f papers were to be requestedto present in the best manner the present state o f o u r knowledge o f

th e particular branch of the science under consideration .

It was the purpose of the officers o f the Congress Auxiliary to printin the Engli sh language all papers read at the va r io u s conferences

,

together with an account o f the daily proceedings . As this purposecould not be fulfi l led by the Auxiliary

,and as it was considered

desirable to publish the papers o f the meteorological congress as soonas practicable , other m eans of publication had to be sought . The

matter was presented to the Secretary of Agriculture,the H o n .

J . Sterling Morton,who approved the publication of the papers as a

bulletin of the U . S . Weather Bureau .

The failure o f the Auxiliary t o provide translators for the manypapers writte n in foreign languages caused the labor of translation todevolve upon the chairm en of the sections ; to these gentlemen , aswell as to Prof . Alexander Ziw e t , of the University of Michigan , andto Mr . Robert Seybo t h ,

of the Weather Bureau,I desire to express my

Obligation fo r their generous assistance .

MARK W . HARRINGTON,

Ch a irm a n .

ORGAN IZATION .

GENERAL COM M ITTEE.

CHAIRM AN.

M a r kW . H a r r i n g t o n , Ch ie f o f U. S. We a th e r Bu r e a u , Wa shin gt o n , D. C.

VICE-CHAIRM AN.

D r . H . C . Fr a n k e n fie l d , Lo ca l Fo r e ca st Officia l, Ch ica g o , Il l .

SECRETARY .

O l iv e r L. Ra s s i g , Libr a ria n , U. S. We a th e r Bu r e a u , Wa sh in gt o n D. C.

M EM BERS OF THE COM M ITTEE.

Pr o f . C l e v e l a n d Abbe , We a th e r Bu r e a u , Wa sh in g t o n , D. C., Ch a irm a n o f Se c

t io n o n Th e o r e tica l M e t e o r o lo gy.

Li e u t . W . H . Be e h l e r , U. S. Na vy , Hydr o g ra ph ic Office , Wa sh in g t o n , D. C

Ch a irm a n o f Se ct io n o n M a r in e M e t e o ro lo gy .

Pr o f . F. H . Bi g e l o w , We a th e r Bu r e a u , Wa sh in g t o n , D. C. , Ch a irm a n o f Se ctio n

o n Atm o sph e ricEle ct r icity a n d Te r r e st r ia l M a g n e t ism .

Pr o f . C h a r l e s C a r p m a e l , Dir e ct o r Ca n a dia n M e t e o ro lo g ica l Se rvice , To r o n t o ;M r . A. La w r e n ce Bo t ch , Dir e ct o r o f Blu e Hill Obs e rva to ry , Bo st o n , M a s s . ;

Ch a irm e n o f Se ct io n o n Na t io n a l We a th e r Se rvice s .

M a j . H . H . C . D u n vvo o d y, U. S. Arm y. ,We a th e r Bu r e a u , Wa sh ingt o n , D. C.

,

Ch a irm a n o f Se ctio n o n Ag ricu ltu ra l M e t e o r o lo gy.

M r . O l iv e r L. Fa s s i g , Wa sh in gt o n , D. C Ch a irm a n o f Se ct io n o n Hist o ry a ndBiblio g raph y.

Pr o f . F. E. Nip h e r , Wa sh in g t o n Un ive r sity , St . Lo u is , M o . , Ch airm a n o f Se c

tio n o n Clim a t o lo gy.Pr o f . Th o m a s R u s s e l l , Office o f U. S. En gin e e rs , Sa u lt Ste . M a rie , M ich ,

Ch a irm a n o f Sectio n o n Rive rs a n d Flo o ds.

Pr o f . C . A. Sch o t t , Co a st a n d Ge o de t icSu rve y , Wa sh ing t o n , D. C. ;

M r . H . H . C l a yt o n , Bo st o n , M ass ;

Ch a irm e n o f Se ct io n o n In st rum e n t s a n d M e th o ds .

LOCAL COM M ITTEE.

(CHICAGOJ

R . GRIGSBY CHANDLER. W . S. JAcxM AN.

ELIAS COLBERT. WILLIAM S . SEAVERNS.

Os srAN GUTHRIE. CHARLES B . THWING.

vi

M INUTES OF THE PROCEED INGS .

MEM ORIAL ART INSTITUTE,

Ch ica g o , EL, M o n da y, Au g u s t 2 1, 1893 .

Monday,August 2 1

,at 10 a . m .

,the congresses of the Departm ent

o f Science and Philosophy were form al ly opened at the Memorial

Art Institute with an address o f welcome by Mr . C . C . Bonney , Pres

ident o f the Congress Auxiliary of the Columbian Exposition . Atthe close o f this general session

,which lasted about one hour

,the

special congresses met in room s assigned to them for organization

and the reading and discussion of papers .

The Congress on Meteorology,C l imatology

,and Terrestrial Mag

n e t ism met in room No . 31,in which the regular sessions were held

daily,from August 2 1 t o August 24 .

At 11 a . m .,Prof . F . H . Bigelow , in the unavoidable absence of the

Chairman,Prof . MarkW. Harrington , Opened the Congress , welcom ing

the members and briefly stating its obj ects . The Con gress had n o

legi slative authority . The main purpose w a s to collect a series o f

mem oirs prepared by writers of recognized merit in their respectivefields o f labor , outlining the progress and summ arizing the present

state of knowledge of the subj ect treated . These reports are t o beprinted in full in the English language

,and will form a record o f

great and permanent value in the science o f meteorology .

At the conclusion of Prof. Bigelow ’ s remarks Capt . A . P . Pinheiro,

Director of the Brazil ian Meteorological Service,was called upon to

read his paper upon Storms in the South Atlantic .

”

Owing to the great number of papers and the absence o f authors,

the papers were largely read in abstract o r by title by the chairm enof the respective sections .Lieut . Beehler , chairman o f the section devoted to marine met eor

ology,read in abstract the following papers

“ The forecasting of ocean storms and the best method o f m aking

such forecasts available,

” by William Al lingham,London .

“ The secular change o f variation of the mariner ’ s compass,by

G . W . Lit t l e h a l e s , Washington , D . C .

“ Ocean te mperatures and ocean currents,

” by Lieut . A . Ha u t r e u x ,

Pari s .

1As a l l pa pe rs pr e se n t e d t o th e Co n g r e ss a r e prin t e d in fu ll in t h e fo llowin g pa g e s , n oa bst r act s a r e give n in th e acco u n t o f th e da ily pr o ce e din g s .

viii CH ICAGO M ETEOROLOGICAL CONGRESS .

The creation o f meteorological observat ories on islands scatteredover the ocean

,

” by the Prince Sovereign o f Monaco .

“ The barometer at s e a ,” by T . S . O

’Le a ry, Washington , D . C .

Mr. Fa s s ig , chairm an o f the section o n history and bibliography ,

presented for reading two papers o f his section“ The meteorological work o f the Smithsonian Institution ,

” by the

Secretary o f the Sm ithsonian Insti tution , read by Mr . H . H . C layton .

“ The meteorological work o f the office of the Surgeon General ,U . S . Arm y ,

” by Maj . Charles Smart , read by Mr . Fa s sig .

Prof . Charles Ca rpm e a l followed with the reading of abstra cts ofthe following papers o f the section devoted to national services andm ethods

,o f which he is one o f the chairm en

" The publication of daily weather m aps and bulletins,by Mr .

R . H . Scott,o f London .

Can we by automatic records at three selected station s determinethe energy o f a flash o f lightning by A . M cAd ie , of Washington ,D . C .

“ The utilization o f cloud observations in local and general weatherpredictions

,

” by A . M cAdie , o f Washington , D . C .

Adj ourned,at p . m .

,to meet Tuesday

,at 10 a . m .

TUESDAY,Au g u s t 2 2 , 1893 .

The m eeting was opened at 10 a . m . by the Chairm an,Prof . Mark

W . Harrington . The first paper of the day was by Lieut . Beehler

on “ The m eteorological work of the Hydrographic Office of the U . S .

Navy.

” During the reading o f this paper,which was devoted largely

to the work o f Com modore Maury,Lieut . Beehler had placed upon a

pedestal,f o r inspection

,a fine bust of the commodore by the sculptor

Valentine,of Richm ond

,Va .

Prof . Le m s t rOm ,of Hel singfors

,moved to hold a preliminary in

formal session a t 10 a . m .,Wednesday

,to decide upon a programme

for the day,the form al Session to begin at a . m . This was

agreed to .

Prof . M a sca r t , o f Paris , then gave a résumé o f his paper on Optical phenomena,

” referring particularly to the explanation o f thewhite rainbow . He also gave a résumé o f M . Chauveau ’ s paper onInstrum ents for the observation o f atmospheric electricity .

”

Capt . Pinheiro was called to the chair while Prof . Harrington readh is paper o n “ The history o f the daily weather map .

Tw o papers by Maj . Dunwoody,o f the U . S . Weather Bureau

,were

presented , Functions of state weather services and“ State weather

services of the United States .” Upon motion o f Mr . Fa s s ig , the dele g a t e s to the Convention o f Directors of State Weather Services

,who

were in session in an adj oining room,were invited to be present at

M INUTES OF THE PROCEEDINGS . ix

the reading o f these papers ; the invitation was accepted and the dele g a t e s attended in a body . At the close of the readings

,Dr . Duncan

,

of Ch icago,m ade some remarks upon the possibi lity of predicting

epidem ics as a result of the developm ent of State weather services .Adj ourned at 2 p . m .

WEDNESDAY,Au gu s t 23 , 1893 .

The informal conference agreed upon o n the preceding day washeld at a . no . It was decided t o read first the papers whoseauthors were present ; then the chairman o f sections were to presentthe papers o f their respective sections of which abstracts had beenpreviously prepared .

At the reading of papers was resumed . Prof . Harringtonrequested Lieut . Beehler t o take the chair .

Prof . Ca rpm a e l continued the reading o f the papers of hi s section,

as follows :

The prediction of droughts in India , by W . L . Dal las,o f Calcutta .

Plan fo r the prediction of fl oods ,” by M . Babin e t , o f Paris .

Dr . Veeder , o f Lyons , N . Y .,read a paper o n

“An internationalcipher code for correspondence relating to auroras and magneticdi sturbances .Prof . Bigelow ,

chairman of the section o n atmo spheric electricityand terrestrial magnetism

,presented the papers of his section

,read

ing some by title,some in abstract. He read at length his paper o n

“ The magnetic action of the sun upon the earth , and it was discussedby those present .

Father Faura,o f the Manila Observatory , presented a paper upon

Signs preceding typhoons in the Philippine Islands .” Father Faura

al so laid before the members an elaborate printed report upon terr e s t r ia l magnetism in the Philippine Islands , prepared by P . R .

C irera,S . J Di rector o f the magnetic section o f Manila Observatory .

Copies o f this report were distributed at the close o f the session .

Prof . Le m s t rOm ,of Helsingfors

,off ered a resolution proposing that

the Congress be divided into four sections , in which there Should be adiscussion as t o the most important questions pressing for solution ,

and that these sections place before th e General Congress a r e co m m e n

dation as to the method o f carrying on the necessary observations orinvestigations

,the General Congress to discuss such recom men dations

and take action thereon . The proposition was not agreed to , as such

action would be foreign to the purpose o f the Congress .Adj ourned a t ,

1.45 p . m .

X CH ICAGO M ETEOROLOGICAL CONGRESS .

THURSDAY,Au gu s t 24 , 1893 .

The m eeting was opened as usual in room No . 3 1, at a . In .

Lieutenant Beehler in the chair .

The first paper o f the day w a s by Father De riza o n“Alpine m eteo

r o l o gy, read in abstract by Father Al qu é .

Mr . Rotch , associate chairman o f the section devoted to national

services,read

,in abstract

,the following papers of h is section

Meteorological stations and the publ ication of results o f observa

tions,

” by Dr . J . Hann , o f Vienna .

‘ Present conditions o f t h e weather service— p ropositions for its

improvem ent,

” by Dr . W . J . van Bebber , o f Berlin .

“ The best method of testing weather predictions,

” by Dr . W .

Koppen,of Ham burg .

Prof . Bigelow then took the chair .

In connection with the reading of Dr . van Bebbe r ’ s paper,Prof .

Ca rpm a e l suggested that a statement describing the m ethod employedby the U . S . Weather Bureau in forecasting the weather be preparedand sent to Dr . van Bebber to be added t o his paper ; that he would

likewise prepare a statem ent describing the method employed by theCanadian Service . This would add greatly t o the interest and valueof Dr . van Bebbe r ’ s paper when published .

Prof . Le m s t rOm then read a paper by Prof . Lindelof , of Hel singf ors ,upon “ The influence o f the rotation of the earth o n movements at itssurface

,etc .

” This was followed by a paper o f his own o n“ The

cosmical relations manifested in the simultaneous disturbances of thesun

,the aurora

,and the terrestrial magnetic field .

The fol lowing resolution , offered by Lieut . Beehler,was then read

and agreed to

Recognizing that the members of this Congress do not possess legis l a t ive powers , be it resolved that the following statement be addedto the official report o f the proceedings : In view of the importance

o f a num ber o f the papers read before the Congress and impressedwith the desire o f international consideration o f certain questions

,we

request special att ention to the followin g points1. International co -operation in observations of auroras .2 . Simultaneous observations at the instant of Greenwich Noon , by

a l l observers o n land and at Se a independent o f , and in addition t o ,all other observations .

3 . Investigation o f the earth ’ s magnetic polar field,and exact de

termination o f the period o f solar rotation .

Mr . Fa s s ig , chairm an o f the section on history and bibliography,

then read abstracts of the following papers of his section“ Contribution to the bibliography o f m eteorology in the fifte enth

t o the seventeenth centuries ,” by Dr . Hellmann

,Berl in .

M INUTES OF THE PROCEEDINGS.

English mete orological literature o f the fifteenth t o the seven

te e n th centuries ,” by Mr . G . J . Sym ons , London .

Early individual observers o f the weather in the Uni ted States ,by Mr . A . J . Henry , Wash ington , D . C .

“ Contributions to theoretical meteorology in the United States

during the Espy-Re dfie l d period ( 1830 by Prof .Wm . M . Davis,

Cambridge,Mass .

“ Contribu tions to theoretical meteorology in the United Statesduring the Loomis-Ferrel period ( 1855 by Prof . Frank Waldo

,

Princeton , N . J .

A first att empt toward a bibliography o f Am erican contributionsto mete orology ,

” by Mr . Oliver L . Fa s sig , Washington , D . C .

The Congress wa s then declared adj ourned,sin e die

,by the pre

siding o ffice r , Prof . F . H . Bigelow .

Papers presented to the Congress and not especially referred to inthis report were read by title only .

OLIVER L . FASSIG,S ecr e ta ry.

TABLE OF CONTENTS .

P A R T I

Se ct io n I.— W e a t h e r s e rv ice s a n d m e t h o d s .

1

2

M e t e o r o lo gica l st a t io n s a n d th e pu blica t io n o f r e su lt s o f o bs e rva t io n .

Pr o f. Dr . J . Ha n n , Dir e ct o r Au st r ia n M e t e o ro lo gica l Se rvice , Vie n n aTh e publica t io n o f we a th e r m a ps a n d bu lle t in s . Ro be r t H. Sco t t , Se c

r e ta ryRo ya l M e t e o r o lo gica l Co u n cil, Lo n do nFu n ct io n s o f St a t e w e a th e r s e rvice s . M aj . H. H . C . Du nwo o dy , U. S.

Arm y , Assist a n t Ch ie f U. S. We a th e r Bu r e a u , Wa sh in gt o n , D. CTh e pr e dict io n s o f dr o u g h t s in In dia . W. L. Da lla s

,Assist a n t M e t e o r o

lo g ica l Re po rt e r t o th e Go ve r n m e n t o f In dia , Ca lcu t t aCa n w e , by a u t o m a t ic r e co rds , a t th r e e s e le ct e d sta t io n s de t e rm in e th e

e n e rgy o f a fla sh o f l igh t n in g ? Ale xa n de r M cAdie , M . A. , U. S.

We a th e r Bu r e a u , Wa sh in g t o n , D. C

Th e u t iliza t io n o f clo u d o bse r va t io n s in lo ca l a n d g e n e ra l we a th e r p r edict io n s . Ale xa n de r M cAdie , M . A . Pla t e I

An in t e r n a t io n a l ciph e r co de fo r co r r e spo n de n ce r e spe ct in g th e a u r o r a

a n d r e la t e d co n ditio n s . Dr . M . A . Ve e de r , Lyo n s , N. Y

Th e be st m e th o d o f t e st in g we a t h e r pr e dict io n s . Pro f. Dr . W. Ko ppe n ,M a rin e Obse rva t o ry , Ham bu rg

Th e pr e se n t co n dit io n o f th e we a t h e r se rvice — pr o po sit io n s fo r it s im pr o vem e n t . Pr o f. Dr . W. J . va n Be bbe r , M a r in e Obs e rva t o ry , Ha m bu rgAppe n dice s

I. Ca n a dia n se rvice .

II. Da n ish

IH. No rwe g ia n se rvice

IV. Ru ssia n se rvice

V. Au st ria n se rvice

VI. Hu n g a r ia n se rvice

VII. Ne th e rla n d se rvice

VIII. Br it ish se rvice

IX . Be r l in s e rvice

X . Swiss se rvice

XI. Unit e d St a t e s s e rvice

Se ct io n II.— Riv e r s a n d flo o d s .

1. Flo o ds o f th e M ississippi Rive r , w it h r e fe r e nce t o t h e in u n da t io n o f th e

2 .

3 .

4.

5.

a llu via l va lle y. Willia m St a rlin g , Ch ie f En g in e e r , M ississippi Le ve eCo m m is sio n , Gr e e n ville , M iss

Flo o d pla n e s o f th e M ississippi Rive r . J . A . Ocke rs o n , U. 8 . Eng in e e r ,M ississippi Rive r Co m m issio n , St . Lo u is , M o . Pla t e s Ir-Iv

Rive r -s t a g e pr e dict io n s in th e Un it e d Sta te s . Pr o f. Th o m a s Ru sse ll,Office o f U. S. En g in e e rs , Sa u lt St e . M a r ie ,

M e th o ds in u se in Fr a nce in fo r e ca s t ing flo o ds . M . Babin e t , Assista n tSe cr e t a ry o f th e Co m m issio n fo r Fo r e ca st in g Fl o o ds , Pa ris

Th e fo u r g r e a t rive rs o f Sibe ria . Dr . Fr a n z Ot t o Spe rk, Sm o le n sk ,Ru ssia

xiv TABLE OF CONTENTS .

Se ction II.— Riv e r s a n d no o d e — Co n t in u e d .

6.

7 .

8 .

Re gim e n o f th e Rh in e r e g io n : h igh wa t e r ph e n o m e n a a nd th e ir pr e dic

t io n . M . va n Te in , Ce n t r a l Bu r e a u fo r M e te o ro lo gy a n d Hydr o g ra phyo f Ba de n , Ka rlsr u h e

Th e Nile . M r . W. Willco cks , M . I. C. E. , Dir e ct o r Ge n e ra l o f th e Re s~e rvo irs o f Egypt , Ca ir o . Pla t e v

Th e be st m e a n s o f fin din g r u le s fo r pr e dictin g flo o ds in wa te r co u rs e s .

Se ctio n IH .— M a r in e m e te o r o l o g y .

1.

2 .

Pla t e

Th e fo r e ca st in g o f o ce a n s t o rm s a n d th e be st m e th o d of m a king su ch

fo r e ca st s a va ila ble t o co m m e rce . Willia m Allingh am , Lo n do n

Th e cr e a t io n o f m e t e o r o lo gica l obs e rva t o rie s o n isla n ds co n n e ct e d byca ble with a co n t in e n t . Albe r t , Prince o f M o n aco

Th e m a r in e n e ph o sco pe a n d it s u s e fu ln e ss t o th e n a viga t o r . Pr o f. Cle vela n d Abbe

,U. 8 . We a th e r Bu r e a u , Wa sh in gt o n , D. C. Pla t e vr

Th e ba r o m e t e r a t s e a . T. S. O’Le a ry, U. S. Hydr o g r a ph icOffice , Wa sh

in g t o n , D. C

Th e s e cu la r ch a n g e in th e dir e ct io n o f th e m a g n e t ic n e e dle ; it s ca u se a n dpe r io d. G. W. Lit t l e h a l e s

,U. S. Hydr o gra ph ic Office , Wa sh in gt o n ,

Re la t io n s be twe e n th e ba r o m e t r icpr e ssu r e a n d th e dir e ctio n a n d st r e n g th o fo ce a n cu r r e n t s . Lie u t . W. H. Be e h le r , U. 8 . Na vy , Ch ie f o f Divisio no f M e t e o r o lo gy , U. S. Hydr o g ra ph icOffice , Wa sh in g t o n ,

D. C . Pla t e

Th e pe r io dic a n d n o n -pe rio dicflu ct u a t io n s in th e la t it u de o f st o rm t racks .

Dr . M . A . Ve e de r , Lyo n s , N. Y

No r th Atla n t ic cu r r e n t s a n d su r fa ce t e m pe r a tu r e s . Lie u t . A . Ha u t r e u x ,

Fr e n ch Na vy . Pla t e s vmSt o rm s in th e So u th Atla n t ic. Ca pt . A . P. Pin h e ir o , Ch ie f o f t h e M e t e o r

o l o gica l Se rvice o f th e Br a zilia n Na vy,Rio de

LIST OF PLATES .

P A R T I .

Re la t io n be twe e n t e m pe r a t u r e a n d clo u din e ss . M cAdie .

Lo ca t io n o f g a u g e s o n th e Lo we r M ississipppi Rive r . Ocke rs o n .

Hig h e s t a n n u a l s t a g e s o f th e M ississippiRive r a n d da t e s o f th e ir o ccu rr e nce ,1872 Ocke rso n .

IV. Hydr o gr a phs o f th e M issis sippi, t h e M isso u ri, a n d th e Oh io rive rs , 1872Ocke r so n .

Th e Nile . Willco cks .

Th e m a r in e n e ph o sco pe . Abbe .

Ba r o m e t r ic pr e ssu r e a t se a a n d oce a n cu rr e n ts . Be e hle r .

Cu r r e n t s o f th e Atla n t ic in 1892 . Te m pe ra t u r e s o f th e su rface o f th e se a

in t h e Bay o f Bisca y . Ha u t r e u x .

Drift ing bo t tle s , Ju n e , 1893 . Ha u tr e u x.

Te m pe ra t u r e s o f th e s e a fr o m th e Gir o nde to th e La Pla ta Rive r . Ha u t r e u x.

https://www.forgottenbooks.com/join

S E C T I O N I .

WEATH ER SERVICES AND M ETHOD S .

1.— M ETEOROLOGICAL STATIONS AND THE PUBLICATION OF

RESULTS OF OBSERVATIONS .

Pr o f. Dr . J. HANN.

I.

— WHAT ADD ITIONAL STATIONS ARE DESIRED FOR M ETEOROLOGICALAND FOR CLIM ATOLOGICAL PURPO SES

In many im portant fields o f meteorology the progress o f our

knowledge depends upon the uniform distribution o f the meteoro

logical stations over the earth ’ s surface,so that large districts shal l

not remain without observing stations .I wil l only point o u t that the important question , whether the

mean tem perature of the entire earth ’ s surface a s well a s the quantity o f precipitation

,etc .

,undergoes periodic or continual changes

,

can only be settled when no great part o f it remains without stations .Only then shal l we be certain that changes in the mean condition o f

the atmosphere observed at certain stations are not compensat ed forin a contrary sense on those parts o f the earth which lack stations .

The vast extent of the oceanWil l always be a great obstacle to theinvestigation of the mean condition and variation o f the atmosphereover the whole earth . It is the m ore important that all oceanicislands should

,if possible

,have m eteorological stations

,and this i s

especially true o f the islands o f the Pacific . There should be atleast uninterrupted records o f temperature and rainfall . Much progress latterly has been m ade in this respect but much remains to bedone . The southern oceans

,unfortunately

,rem ain alm ost without

stations . Still,by buried thermometers on the islands in the South

Pacific,Atlantic

,and Indian oceans som e tem perature determinations

may be m ade,since the determination of the constant earth t e m

p e r a t u r e a t su it abl e points may be em ployed as a substitute for theestimation o f the mean air temperature when there i s no prospect ofestablishing permanent stations .In the first place

,I would insist o n the occupation of the oceanic

islands by m eteorolog ical sta tions , since here appear the great gaps

2 CH ICAGO M ETEOROLOGICAL CONGRESS.

in o u r knowledge of the meteorological conditions o f the whole earth ’ s

surface .

Referring to certain portions of the globe,it i s very important that

a ring o f meteorological stations surrounding the North Pole Shouldbe in constant operation . The Polar region north of Europe andAsia i s tolerably well surrounded by the meteorological stations ofRussia

,Norway

,and Denm ark

,but a perm anent station on Nova

Z e m bl a i s perhaps attainable ; a similar station in Spitzbergen r e

mains perhaps only a hope,but it would be of great im portance f o r

the determ ination o f the climatological variation o f the European

frozen ocean .

We have to thank Denmark f o r the instal lation o f m eteorologicalstations on the coast o f west Greenland up to very high latitudes .Thence

,further west

,there exists a deplorable gap for which

,h o w

ever,the explanation and excuse are not far to seek . Nevertheless ,

when possible,eff orts should be m ade to establ ish a perm anent me

t e o r o l o g ica l station in Arctic North Am erica between 60° and 165°

west from Greenwich,near the seventieth parallel

,the further west

the better . Point Barrow would be a suitable point f o r such a permanent station . Perhaps this desideratum f o r science i s already a

reality . On e o r tw o o f the stations in northern A laska Should be inconstant operation . In the Antarctic latitudes there can be noquestion o f permanent stations . The m ost southerly stations inSouth America and in New Zealand are

,therefore

,very important a s

being those in the highest latitudes which i t is possible to reach inthe southern hemisphere . Much value consequently attaches to thepermanency of these stations and to the regular publ ication of theresults Of the observations .In the temperate latitudes of both hem ispheres

,so far a s they are .

not occupied by the oceans,a sufficient num ber o f stations has gen

e r a l l y been provided , and the existing gaps will no doubt Shortly befil led . Matters are not so favorable as regards the occupation o f the

tropical zone by meteorological sta tions .

The greatest gaps we find in South Am erica . In tropical SouthAmerica meteorological stations are almost completely lacking

,at

least in the interior . Som e stations in the great Amazon Valleywould be of .m uch importance . In Para

,Manaos

,Teffe (Ega ) , Taba

tinga , and Iquitos , the establishment o f meteorological stations wouldpresent no impossibility . In the same way s tations could be establ ish e d at some of the capitals o f the inte rior Brazil ian states . It isto be regretted that neither in Quito

,which possesses an astronomical

observatory , nor in Bogota , n o r in Lima,are there m eteorological

stations which publish their observations . ‘ In short,tropical South

1Th e n e w m e te o r o lo g ica l o bse r va t o ry “Un a n u e , in Lim a , wil l pr obably fil l t h is,wa n t . —ED1TOR.

ADDITIONAL STATIONS DESIRED . 3

Am erica remains the t e rm in cogn i ta as regards clim atological a n d

m eteorological data .

Even in tropical Africa,there is im provem ent in this respect,

notwithstanding the fact that tropical South America i s occupied bycivilized states

,which is only true t o a limited extent o f the interior

o f Africa . If the Egyptian equatorial province was n o t , by theshortsighted and foreign pol icy o f a European power , given up to

the Mahdists,we should now have continuous meteorological data

from Lado and the countries o n the banks o f Lake Victoria . Agood beginning had already been made when barbarity interfered .

Stations in the interior o f the Congo States would be very desirable ,but we m ust stil l wait for them

,a s well a s f o r Stations in the British

and Germ an claim s in equatorial East Africa . There is , however ,every prospect that in Germ an East Africa meteorological stationswill be established .

Australia is already partially provided with stations , and , to allappearances

,the num ber will be increased .

‘

Fo r the study o f certain interesting questions as to the dailyperiod Of wind direction and the daily period o f the barom eter

,

stations would be valuable if situated in the m idst o f a large,even

plain . They should be provided with self-recording barom eters and

anem om eters,and the observations should be published in exte n s o ;

a series o f five -year observations would only suffice to answer theproposed questions

, viz . ,daily perio d o f wind direction and am pli

tude of o n e diurnal oscillation of the barom eter . Those mete oro

logical s e rvice sgp o s s e s s in g such stations are requested t o m ake the

fact known .

Stations o n tropica l plateaux , or better , o n high mountains in thetropics

,could contribute with advantage to the question o f the ex

is te n ce of long periods in the m ean air temperature .

Long ago,in the Zeitschrift f iir Meteorologie

,

” I stated my opinion

that a lofty barometric station in the equatorial regions would givethe best explanation o f the tem perature variation of the stratum ofair lying between the station and s e a level . The observations of

pressure would give a m uch better indication o f this than the therm o m e t e r itself

,which gives only the local tem perature and is subj ect

to m any disturbing influences .The barom eter on a mountain is

,therefore

, t o be regarded as a good

air therm ometer,o r at least a kind o f diff erenti al thermometer when

the true height o f the barometer is not known . It indicates the t e m

p e r a t u r e o f the whole underlying a ir stratum,or at least the vari a

tions . There m ust also be a base station whose horizontal distancefrom the high station is so sm all that no considerable pressure gradient ( in a horizontal direction ) can be suspected between the two stations during a period o f som e length , such as a year

’ s mean .


If we designate by B the height of the barometer at the base station ,by bthat at the high station , by h the difference o f height betweenthem

,by Tthe air tem perature in absolute measures ( that is , t

by R the known constant ( for dry air the equation

b bh bd b_ d B

B +R T

2d b— d B -

E

i s the pressure change at the h igh station,with the pressure variation

at the corresponding place o n the earth eliminated , which is only dependent o n the temperature and vapor capacity o f the air . The true

thermic pressure variation at the height h i s accordingly, for a station like Quito (b 548 millim eters , h about meters , t z é

(2 7° T being therefore R

d b' = 0 .62 d t o r d t = 1.61 d b

’

If the mean air temperature of the stratum between s e a level and

Quito changes the barom etric level in Quito alters mil limeters .

Changes of two-tenths o f a degree centigrade correspond,therefore

,to

a pressure change of something more than m il lim eter,which

al lows o f accurate determ ination in the means of the year .

If,f o r exam ple

,a period corresponding to the s u n -spot period ex

i ste in the mean air temperature , it must al so be equally wel l shownin the pressure variations of the high station to allow its magnitudeto be calculated . While the therm ometer only gives the local airtemperature o f Guayaquil and Quito

,for exam ple

,Which is much

influenced by chance circum stances,clouds

,precipitation

,etc .

,the

barometer furnishes the true air temperature o f the whole meterso f air , as wel l a s the effect produced in the sam e way by changes inthe amount o f vapor

,that i s to say

,in a certain degree the “ poten

tial tem perature . The mean barom etric pressures,therefore

, o f the

tropical high stations give much more precise indications o f the variation o f the a ir temperature

,and thereby o f the solar radiation

,than

the therm om eter itself . In order to derive the ful l advantages of thismethod o f m easuring the air temperature by the barometer

,the lower

station,where the higher pressure i s observed

,m ust n o t be s o far re

moved from the upper stratum that the relation d b dB (b B) holds .Quito

,therefore

,would not be a good station f o r this purpose . A per

manent station o n the Do d abe t t a Peak,in South India

,o n the con

t r a rywould be very suitable . If the Indian Government would erecta first class observatory o n the Do d abe t t a Peak , in the Nilgiri Hil l s ,science would be much benefited . Such a station would aid m eteor

ology greatly in other directions . Still better would be a permanentstation o n the Kamerun Peak

,in West Africa

,but the erection would

present much greater d ifli cu l t ie s than that on the Do d abe t t a Peak ,which could be easily carried out . (The high station o f Nuwara

Eliya , meters , in Ceylon , if only the barometer correction is

ADD ITIONAL STATIONS DESIRED . 5

sufficiently constant,I would already place in this category . ) Even

if the exact altitude above sea level o f such a tropical station i s n uknown

,sti ll

,by the introduction o f an approximate value for h in

the above form ula,the variation of the lower air stratum can be cal

cu l a t e d even if the m ean tem perature o f the whole air stratum cannot .The greatest im portance i s to be attributed to the constancy o f thebarom etric correction o r to the accurate determ ination of any changetherein . Short , but entirely hom ogeneous , series o f pressure meanscan be used to determine the variations of the mean air temperature .

II.

-SHOULD THE PUBLICATION OF CLIM ATIC DATA BE FOR PLACESOR DISTRICTS AS REPRESENTED B Y PLACES ?

Each observing system should publi sh,for a certain number o f

chosen stations , whose number corresponds to the size o f the country ,thrice-daily observations in exte n so

,and besides these

,at certain prin

cip a l stations , hourly observations , as i s in fact done by m ost of thegreat European system s .Besides these

,for as m any stations as possible

,the monthly and

annual means should be published according to the internationalscheme

,as h a s been done in the last reports of the Signal Service .

Only in this way can the records of the m eteorological stations bemade useful general ly

,and the progress of the science toward e ffi

cie n cybe promoted .

It is to be very m uch regretted that th e observations are n o t publ ish e d f o r m any stations

,which

,from their positions

,fil l important

gaps in o u r climatological and meteorological knowledge , whereby allthe labor which has been given to m aking the observations i s rendered useless .

In other cases the‘ publication is in an entirely unsuitable form,so

that the results cannot be used scientifically,o r they appear only in

local papers which do n o t r each the specialists .The installation o f stations and the best equipm ent of them with

instruments,their care and reading

,are useless

,if the results of the

observations are n ot sufficiently made publ ic . Economy in money

in the publication o f Observations must be characterized as thegreatest prodigality

,since all the outlay expended o n the Station and

the care given to reading the instrum ents are thus rendered useless .It is to be remem bered that the worth o f the m eteorological datam ay be increased in a notable way , since for data which only go intothe archives

,the zeal and care diminish . The observer who sees his

observation s,or important extracts thereof

,printed and distributed

,

will always try to make them correctly . Criticism of the observations

,and i t s beneficial influence on their value , wil l be greatly in

creased by their publication .

In the most l iberal form o f publication of observations in any

CH ICAGO M ETEOROLOGICAL CONGRESS .

m eteorological system,as for example

,that o f the Central Physical

Observatory at St . Petersburg , the cost o f printing forms only a smallpercentage o f the cost of the whole observing system ,

even when the

labor o f the observer i s not considered . The perm anent value of the

activity o f a meteorological system lies in it s annual reports,and o n

these the greatest efforts should be concentrated . The annual reportso f the various observing systems o f the world form the evidences whichseem destined to be laid before future generations as proofs of the present condition o f the atmosphere over the earth ’ s surface and for thestudy and progress of science . Therefore , w e owe it to o u r successors tohand over to them yearly as detailed reports as possible o f the

.

m eteorological occurrences over the entire globe

,in order that with the lapse

of tim e they may be able to answer the question as to ' the secularvariation o f the m eteorological elem ents . It i s thus always better topubl ish too much than too little

,since what i s missed cannot be

recovered,and avenges itself by retarding the progress o f the science .

The publication o f meteorological means f o r whole di stricts has novalue scientifically . It can

,perhaps

,fo r purely practical purposes

be used to advantage,but fo r all scientific work such combined

m eans are wholly unserviceable . It i s unnecessary to insist on this,for anyone wh o has em ployed m eteorological means and data ingeneral f o r scientific work will agree with me . Neither the stationso f a country nor of a district

,the instruments and their exposure

,

nor the local influences at the various stations remain constant longenough to make the m eans for whole di stricts appear even tolerablycomparable .

The means f o r districts from different series of years are not co mparable with one another and can not be employed to Show t h e

changes o f the meteorological elemen ts with time . In general such

meteorological means and data f o r whole districts should be confined

strictly within the limits to which they be long . They are only to be

employed a s rough approxim ations,which

,occasionally

,may be very

useful practically,but are unserviceable from a scientific standpoint .

2 .—THE PUBLICATION OF DAILY WEATHER M APS AND

BULLETINS .

ROBERT H . SCOTT .

The subj ect which h a s been placed before m e is o n e which hardly

adm its of any very decided treatm ent,inasmuch as the

‘

Scale andcharacter o f the maps to be published in each country must dependfirstly o n the amount of m oney which can be appropriated to the

service of preparation and issue of these maps , and secondly o n theextent o f area which the m aps are intended to cover .



From these observations there are two charts prepared for 8 a . m .,

o n e showing the barom eter , wind , and sea disturbance , ( the wind byarrows and the s e a disturbance by hatching ) , the other Showing thetem perature by i sotherms at 10 ° apart , and the rain in figures

,where

it exceeds inch . Changes in pressure o r tem perature are printedin words across the face of the respective maps .

The chart for the previous evening i s not published by the office

except in it s weekly weather report , which will be described presently ,but a copy is Supplied to “ The Times newspaper

,and appears in it s

morning i ssue o f the fol lowing day , and so secures a very extensive

circulation .

A copy o f the 8 a . m . chart i s also forwarded t o The Times , and

incorporated in the second edition,but the circulation o f that edition

is n o t very extensive .

Both o f these copies are prepared expressly for “ The Tim es,and

at the sole cost o f that j ournal,which for more than thirty years ,

ever since m eteorological telegraphy w a s organized by Adm iral FitzRoy in 1860 , has been conspicuous by the prominence it has given init s colum ns to m eteorological inform ation . In fact , f o r som e years ,the entire service for the preparation o f these 6 p . m . charts was car

ried on at the sole cost of The Tim es,a fact which affords strong

evidence o f the public interest in weather intel ligence evinced in thi scountry .

Th e we ekly we a th e r r ep o r t— This was commenced in 1878 at the

suggestion o f em inent agricultural authorities,in order to supply for

the diff erent agricultural dist ricts statem ents of the temperature andam ount o f rain for the week

,and o f their differences from their

respective averages . In 1884 this report w a s m aterially im proved bythe insertion of figures illustrating the weekly m arch of cumulativetemperature , that i s , o f the num ber o f day degrees ” o f tem peratureabove o r below 42 ° F . (approxim ately 6

° which,according to the

late Alphonse de Candol le,is the degree of tem perature at which

active vegetable growth may be assum ed to comm ence . A popularexplanation o f this cum ulative tem perature will be found in a paper

read by me before the International Heal th Comm i ssion in 1874 .

An explanation o f the scientific principl es on which the calculationof the values published weekly i s based will be found in a paper by

Lieut . Gen . R . Strachey,which appeared in the “ Quarterly Weather

Report ” for 1878 .

At the present date,1893

,this report contains o n the first page

,f o r

each o f twelve districts :Fo r tem p e r a tu r e .

— The average and the absolute maxim um and

minim um . The m ean for the week,and it s diff erence from the aver

age for the week .

Fo r a ccu m u l a te d h e a t — The num ber o f day degrees above and be

STATE WEATHER SERVICES. 9

low 42 ° F . f o r the week,and their respective differences from the

mean,with similar information for the interval elapsed from the

beginning of the current year to the last day in the report .

Fo r r a in — The num ber o f wet days . The total fal l f o r the week,

and its difference from the average,and sim ilar inform ation

,as be

fore,for the interval from the beginning o f the year .

Fo r su n sh in e — The number o f hours recorded during the week , its

percentage o f the possible duration , and i t s difference from the average

,with similar data for the in terval since the commencement of

the year,and general remarks o n the weather for the week .

Page 2 gives information f o r each of the stations,as regards tem

p e r a t u r e , rain , and sunshine , with d ifie r e n ce s from averages fo r the

week .

Then follow weather m aps for the whole o f Eu r 0 p e as far eastwardas Odessa

,Moscow

,and Archangel

,giving

,respectively

,pressure and

wind f o r 8 a . m . and f o r 6 p . m . ,and temperature and weather f o r 8

a . m . only .

Remarks are given f o r each day,and the report concludes with a

table o f sunshine values for additional stations in the . United

Kingdom .

Appendices have,in successive years

,appeared in connection with

the “Weekly Weather Report,

” and in te r a l ia these have containedfigures giving

,f o r each o f the districts into

-

which these i slands have

been divided,the weekly and progressive values of the different ele

m ents for each year a s far back as 1879 .

The daily and weekly weather reports are accompanied by monthlysumm aries

,giving

,for calendar months

,a brief summary o f the

weather over the United Kingdom .

This is a brief account o f the am ount and character o f the inform ation which the experience o f this office has led it to issue daily

,

weekly,and m onthly

,for the use o f the public .

3 .— FU

'

NCTIONS OF STATE WEATHER SERVICES .

M aj o r H. H . C. DUNWOODY,U. S. A .

State Weather Services are organizations f o r the collection and

dissem ination of climatological and other information . They depend

almost whol ly upon the voluntary co -Operation o f intel ligent and

public-spirited citizens,whose individual reports collected at the s e v

eral central stations form the basis o f their publications . These publ ica t io n s are reviews o f the prevail ing weather conditions published

monthly,and bulletins issued weekly during the season of planting,

cultivating,and harvesting o f crops

,giving the m ore important

weather features and their effect upon growing crops from week to


week . Through S tate weather service organizations,the daily weather

forecasts and special warnings o f the National Bureau are distributedto large num bers of stations throughout the country .

There are three independent l ines of work,each dependent

”

upon

its special class of contributors who serve in the capacity of (1) met e o r o l o g ica l observers , taking observations o f temperature , rainfal land m iscel laneous data ; (2 ) crop correspondents , wh o , during thecrop season

,render weekly reports o f farming operations , the growth ,

m aturing,and harvesting of crops

,and the efie ct s of the prevailing

weather conditions thereon ; (3 ) the forecast disp l aym e n,who displ ay

flags o r sound whistle signal s repres enting the weather forecasts ofthe National Weather Service . It not infrequently happens that o n eperson serves in more than one capacity and sometimes co -operatesin al l the three distinct l ines of work .

In the United S tates there are less than 175 m eteorological stationsconducted by the regular paid observers o f the Weather Bureau , orabout o n e station for each square miles o f territory . The

utter inadequacy o f the data supplied by these stations fo r pur

poses o f detailed investigation o f special localities is therefore plainly

apparent,m aking the State weather service an absolute necessity

for the prosecution o f such work .

Although the work of collecting volu ntary m eteorological observations and publishing the results was begun in Iowa as early as 1875 ,and in Missouri in 1878, the organization of State weather services

for the active prosecution o f work o n the lines previously referred to

may be said to have begun in 1881 and 1882,Since which time the

number o f meteorological stations has steadily increased,there being

now about stations taking and recording meteorological obs e r

va t io n s daily . With this exte nsive system it i s possible to determinethe special cl im atic features of every section o f the country to anextent that would be entirely im possible were it not for the exi stenceof local weather services .

Al l State weather services issue monthly reviews of the prevail ingweather conditions

,and m any o f these publications are i ssued in

elaborate and attractive form,rendering them valuable and interest

ing . In m any o f these m onthly reviews,besides giving a general dis

cu s sio n o f the daily temperature and precipitation,observations are

published in detail . While it would be difficult t o correctly estimatethe great value o f this particular l ine o f State weather service work

,

a more popular feature i s the weather-crop service . From the begin

ning o f the crop season until it s close,weekly reports o f the weather

conditions a n d the effects o f the sam e upon farming Operati ons , the

growth o f crops , etc . ,are co l lected at the several State weather service

centers . These weather-crop reports are m ailed by the correspondentss o as to reach the central station o n Tuesday morning

,and

,as far as

STATE WEATHER SERVICES. 11

possible,cover the week ending with Monday . Upon receipt they

are carefully summarized and a brief discussion o f the general con

dit io n s prepared,which

,with the detailed reports from the several

correspondents,forms the State crop bul letin . The ofli cia l in charge

of each State service o n Tuesday morning sends a telegraphic sum

mary o f the more im portant features o f the week t o the National

Weather Bureau in Washington .

The entire territory o f the country being covered by local Services ,com plete information a s to weather and crop conditions i s had from

every section o f the United States . These telegraph ic reports are

publ ished in full in the National Weather-Crop Bulletin,and

,with

the charts o f temperature and precipitation departures , form thebasis of a general discussion of the weather and crop condition s forthe whole country .

The charts o f temperature and precipitation departures are prepared from the data collected principally from U . S . lVe a th e r Bureau

stations and serve in a general way to Show how the temperature andrainfall o f each week com pares with the normal o f the corresponding

period .

This weather-crop service is,with the exception o f the general

weather forecasts,the most valuable work being done by the National

Bureau,and is the most popular feature o f State weather service work ,

being'

o f greatest interest to agriculturists , although the bulletinsare eagerly looked for by those interested in other pursuits . To the

intelligent farm er it affords a m eans o f supplying accurate and impor

tant information as t o the condition o f crops,enabling h im t o form

reliable estim ates as to supply and demand . In some States the edi

tions o f the local weather-crop bul letin have already grown to very

large proportions,and the dem an d f o r the bulletin is constantly

increasing . More than copies o f the Ohio weather-cr0 p bul

l e t in are printed and distributed weekly . As an illustration of theimportance o f this work

,i t m ay be stated that a material change in

the condition o f the cotton crop in the State o f Texas influences the

cotton m arkets of the world ; and it i s the work o f the State weatherService that presents weekly im partial and reliable information as tothe actual weather and crop conditions prevailing throughout eachseason .

The publ icity given the State and National weather-crop bulletinsthrough the press o f the country is so extensive that an accurate estimate o f the combined bulletin and newspaper circulation would bedifficult of com putation : The full text o f the National bul letin ,including the special telegraphic reports from the various States , is

telegraphed each week by the press associations and printed o n

Wednesday in the large dailies . The agricultural press make a

specialty of the bulletin,and some reproduce in their columns the


charts o f precipitation and temperature . The patent-sheet papersalso find the bulletin an attractive item , and they extensively printthe bulletins of the States covered by their circulation . The Missouri

bulletin i s printed in nearly o n e hundred patent-sheet papers issuedby the Kellogg and Western newspaper companies .

The late Prof . George H . Cook , for several years Director o f theNew Jersey Weather Service

,in the work o f organizing the New Je r

sey service,summarized the importance o f the State service as fol

lows

It willbe th e m e a n s o f so o n s e cu r in g be t t e r p r e dict io n s o f w e a th e r ch a n g e s a n d s t o rm s .

It w ill br ing th e be n efit s o f t h e Na t io n a l We a th e r Bu r e a u o f th e Un it e d St a t e s in t o

e ve ry co u n ty pa r t icipa t in g in th e St a t e lo ca l o rg a n iz a t io n .

It w ill so o n pr e pa r e th e Sta t e fo r a sys t e m o f s t o rm Sig n a ls displa ye d fr o m r a il f'o a d

t r a in s t h a t will be w ide ly be n eficia l t o a g r icu ltu r a l in t e r e s t s .

It will g ive t o e ve ry co u n ty th e Go ve r n m e n t s ta n da rds fo r t e m pe r a t u r e , r a in fa ll, win dve lo city , h u m idity , e tc. ,

wh ich a r e s o u rce s o f u se fu l pu blic in fo rm a t io n .

It w ill p u t with in r e ach o f lo ca l a g r icu lt u ra l s o cie t ie s m e a n s o f a ccu r a t e o bse rva t io n swh ich

,in t h e co u r se o f ye a r s , m u s t be va lu a ble t o a nylo ca lity in th e st u dy a n d a da pt a t io n

o f ce r e a ls .

It w ill br in g t h e scie n ce a n d m e th o ds o f th e Na t io n a l We a th e r Bu r e a u with in th e

r e a ch o f th e h igh sch o o ls o f th e St a t e,o ff e r in g t e a ch e r s a n d pu pils alike e xce lle n t

Oppo r t u n ity t o st u dy a wide r a n g e o f th e a pplica t io n o f scie n ce t o fo st e r a n d pr o t e cta g r icu lt u ra l in du st ry .

It will le a d t o t h e co ll e ctio n o f r a in fa ll s t a t ist ics t o e n a ble e n g in e e rs t o be t t e r e s tim a t e

t h e su pply o f ca n a ls , a ls o th e su dde n do wn po u r s t o g u a rd a g a in s t in la yin g o u t s e we r s

in cit ie s . It will le a d t o a co r r e ct kn o wle dg e o f r a in fa ll o ve r t h e diffe r e n t w a t e r sh e ds o ft h e St a t e , fo r th e pu rpo se o f g iVin g da t a fo r su pplyin g th e w a t e r w o rks o f cit ie s , t o wn s ,a n d villa g e s .

It w ill le a d t o th e fo rm in g o f r e lia ble m e t e o r o lo g ica l r e co rds fo r u se in l e g a l ca s e s .

It will le a d t o publish in g th e t e m p e r a t u r e o f su m m e r r e s o r t s , dra win g a t t e n t io n o f o u t

side pa r t ie s t o th e ir de sir abil ity a s su m m e r r e side n ce .

It wil l le a d t o a be t t e r pr a ct ice o f m e dicin e , wh e n ph ysicia n s th r o u gh o u t th e Sta t eca n s t u dy dise a se with r e liable an d a ccu r a t e m e t e o r o lo g ica l fa ct s by th e ir side — a n d fo r

sa n it a ry pu rpo s e s co r r e ct m e t e o r o lo g ica l s ta t is t ics a r e inva lu a ble t o th e pr a ct it io n e r ina pplyin g pr e ve n t ive r e m e dieSfo r t h e pu blic g o o d .

The g rowth and popularity of these services were such that inNovember

,1885

,Gen . W . B . Hazen

,Chief Signal Oflice r

,invited the

directors o f al l State weather services to assemble in Washington f o rthe purpose of mutual conference and discussion . Arrangementswere accordingly m ade f o r a convention o f the directors

,which met

February 24 and 2 5,1886 . At this conference m any im portant sub

j e ct s bearing upon State services were discussed looking to improvedmethods o f taking and recording observations

,and a general inter

change o f Views regarding State service work was had . Much g ood

resulted from th is conference,and a report of its proceedings was

publ ished wi th the Annual Report o f the Chief Signal‘

Ofii ce r .

A second and more largely attended convention met in Rochesterin Augus t , 1892 , at which the

“ American Association o f StateWeather Services wa s formed

,the constitution o f which provides

DROUGHTS IN INDIA . 13

for annual m eetings,and the convention of 1893 will be held in

Chicago in August during the time of the meeting of the Meteorological Congress

,August 2 1—24 .

During the past year there have been prepared by m any Stateweather services

,f o r exhibit at the World ’ s Fair

,valuable and inter

esting charts il lustrating graphically the special clim atic features ofthe several States . Some o f these exhibits have been prepared atmuch expense of labor and considerable pecuniary cost

,and have

been very favorably commented upon .

4 .— THE PRED ICTION OF DROUGHTS IN IND IA .

W. L. DALLAS .

The following gives an account o f the m ethod employed in Indiafor the preparation o f the seasonal forecasts issued by the IndiaMeteorological Department

,the ch ief obj ect of which is to give

warning o f the probable occurrence o f severe drought in a ny large

area in India .

In northern India there are two distinct periods of rainfal l o f

importance for agricultural operations . The first is the period of

the sou thwest monsoon rains from June to October . They are

heaviest in the coast districts and at the foot o f the Himalayas,and

are most intermittent and irregular in the more interior districts ofnorthern India . The second period is that o f the cold weather rainsfrom December to March

,when light to moderate Showers are

received during the passage o f feeble cyclonic storms across northernIndia .

The chief causes o f fa ilure o f cr0 ps in northern India are

1s t . Deficiency o f rainfall,m ore especially in the southwest mon

soon period .

2d . Early termination o f the southwest m on soon rains .Under these circumstances the great rice crop in the parts o f north

eastern India aff ected withers away and is a m ore or less com pletefailure . In northwestern India it prevents the cold weather cr0 psbeing sown

,except in l o w-lying o r irrigated districts .

In southern India,the Deccan

,and Burmah

,the only period o f

regular rains o f Value f o r the crOp s i s that o f the southwest m o n

soon from May t o November o r December . In the Deccan and

southern India it i s moderate in May and June,l ight from July to

September,and moderate to very heavy in October , Novem ber , and

December .

In‘

t h e s e districts the rains m ay fai l more o r less com pletely duringa part o r whole o f the period . The most serious partial failure i swhen the rainfall of the second maximum (October to Decem ber ) i slight and irregular .

14 CH ICAGO M ETEOROLOGICAL CONGRESS .

Hence,in northern India the most serious droughts are due to

the combination of a more o r less complete failure o f the southwestmonsoon rain s followed by a failure of the cold weather rains . Inthis case both crops

,the kh a r if and r a bi

, fail .

In southern India failure Of the crops and consequent famine i sdue to a m ore or less serious and large failure o f the rains o f a com

pl e t e southwest m onsoon period . The intensity of the scarcity orfam ine consequent on the failure o f the crops under either of theseconditions depends largely upon the character of the previousseasons . If the preceding two o r three seasons have been u n s a t is

factory,s o that the accum ulated food stocks have been depleted , the

famine may be of the m ost intense character .

The preceding remarks have shown that the most important factorin determining the character of the crops is the rainfall o f the southwest m onsoon

,and hence long period forecasts in India have been

chiefly confined to the prevision o f the southwest m onsoon dis tr ibution of rainfall .These forecasts are usually issued in the first week o f June

,and

attem pt to give a rough estimate of the general character o f the rainfall o f the next four m onths in t h e larger provinces o f India

,and

more especially to indicate any area in which there is a strong probabil ity th e rainfal l will be seriously below the normal , o r to point o u twhen there Is a probability of unusual delay In the comm encement ofthe rains o r of their abnorm ally early term ination in northern o r

central India .

Rainfall in Europe occurs chiefly during the passage o f cyclonic

storms,and hence i s apparently fortuitous in its occurrence .

In India at least four-fif th s o f the rainfal l occurs as a normal

feature o f the southwest monsoon circulation . The lower air currentso f that circulation advance into India from the adj acent sea areas

,

determined by the regular periodic pressure and tem perature changesin India and central Asia . The circulation is m ainly m aintainedand continued by its internal energy , o r rather by that of the energyset free o n the condensation o f the aqueous vapor brought up in itover India . It varies to some slight degree in intensity from year toyear , and its extension also varies in different years , dependent uponthe antecedent m eteorological conditions .

It i s this fact,that the rainfall o f this period IS due to the p r e va

lence o f a massive and steady current,and n o t to local cyclonic dis

t u rba n ce s in a region o f irregular winds,that makes it probable long

prevision can be successfully attempted and carried o u t in India .

In order that the attem pt to forecast the character and distributiono f the monsoon rainfall from the meteorological conditions prevailing anterior to the advance o f the rain giving southwest monsoon



fall are related to persistent variations in the strength and extensionand other characteristics o f the great currents o f the period . It wil lsuffice to give o n e case . The monsoon rainfall was very largely in

excess in Burmah in 1891. The following table gives the deflection o fthe mean winds at three representative stations in that area duringeach month o f the season

We s t e r ly d eflectio n , 1891.

St a t io n .

O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

The winds at these stations during the southwest m onsoon are

from directions between south and west,and increased westing

ardently indicates a greater determination o f the Bay monsoon cur

rent to Burm ah and Tenasserim than usual .

Again,the m onsoon currents were both stronger than usual in 1892

during the period July to September . The fol lowing data will Showthat the increased strength w a s marked throughout the whole o f theperiod

,m ore especial ly in the case o f the strongest current in that

year,viz .

,the Bengal current

Pe r cen ta g e va r i a tio n of s tr e n g th .

Nam e o f cu r r e n t .

The relations of the variations of the strength and direction of thelower air currents during the southwest m onsoon to the rainfall varia t io n s require further investigation

,but s u flicie n t data have already

been accum ulated to establish that there are marked differences inthe strength and extension of the monsoon currents and in the dist r ibu t io n o f the rainfall from year to year

,and that these are directly

related to each other .

These relations m ight have been inferred from the fact that themonsoon rainfall is n o t due to the passage of cyclonic storm s

,but to

the continued prevalence of a steady,strong current charged with vast

supplies o f aqueous vapor . Assum ing that the Character of the d ist r ibu t io n o f the rainfal l i s fairly persistent through out each season

,

and that the rainfall is due to the advance and prevalence o f a

strong s e a current into the Indian land area,it i s evident that the

extension o f this current will be to som e degree determined by any

abnormal meteorological conditions present before o r during its

advance . The following gives a brief Statement o f some o f thesedetermining conditions :

DROUGHTS IN INDIA . 17

1) Unusually heavy and prolonged snowfall in the HimalayanMoun tain area has been shown by Mr . Blanford to exercise a verypowerful influence . It modifies the pressure and tem perature co n

d it io n s in n orthern India,and usually n o t only retards the com

m e n ce m e n t o f th e m onsoon but modifies it s intensity . The manner

in which snowfall modifies the hot weather conditions and the subsequent rains has been investigated and is fairly wel l known . Ab

normally deficient snowfall and its usual correlative,m ore intense

h o t weather conditions than usual,o n the other hand are found t o

precede almost invariably stronger and ste adier monsoon than usual .

(2 ) The abnormal pressure conditions establ ished during the h o tweather

,more especial ly if they are m arked

,exercise a large in flu

ence in m odifying the set o f the monsoon currents . The general

rule in In dia lis that the h o t we a t h e r'

t e n d s to exaggerate and developlocal p e cu l a r it ie s o f pressure

,and the rains to smooth them away .

Thus,if the hot-weather develop a local deficiency o f pressure in any

area it tends to becom e a sink to which the m onsoon current i s morelargely directed than usual

,and hence al so affects the rainfall in

neighboring districts . If,o n the other hand

,a local excess o f pres

sure i s formed,as occasionally happens in Guzerat

,northwest Ra jpu

tana,etc .

,it usually accom panies a considerable o r large dim inution

o f the rainfall in Rajputana o r northwestern India . Much rem ains

to be done to work out ful ly the influence exerted by high and l o wabnormal pressure areas in modifying the distribution o f the monsoon rainfal l

,but several useful relations have been established and

are used in drawing up these long-period forecasts .Sim ilarly

,the consideration o f the temperature conditions o f India

during the h o t weather throws light o n the causes of the general andlocal pressure conditions obtaining before the setting in o f the m o n

soon,and hence enables their probable importance to be estim ated .

An im portant point to be taken into consideration is the relativestrength o f the two currents

,a s upon this depends largely the posi

tion of the monsoon trough of low pressure,and hence also the mean

tracks o f the cyclonic storm s o f the rains an d o f the heavy rainfal lthat accom panies these storms . A strong Bom bay m onsoon tends todisplace it northward and a strong Bengal monsoon southward .

Another important point is based o n the results o f Mr . Blanford ’ s

investigations (given in the “ Rainfal l o f India ”) o f the relationsbetween the rainfall variations in different parts of India . He has

worked o u t very fully the areas in which the rainfal l variations areusually similar o r opposite in character

,and the measure o f the prob

ability o f sim ilar or opposite variations occurring for any given year .

The previous gives a few o f the m ore important principles andfacts upon which the fore ca sts of the distribution o f the m onsoonrainfall are based ,

2


A consideration of the s nowfall data o f the cold weather , o f them eteorological conditions prevailing during the hot weather , andmore especial ly the character and persistency o f the pressure variations

,usually enables a rough estim ate o f the general Strength o f the

monsoon currents a nd the distribution o f the rainfal l to be made .

This is first done and afterwards a com parison is made with previou syears in which similar conditions are known to have obtained . By

taking into consideration the actual conditions,the relations estab

l ish e d by Mr . Blanford between the rainfall variations in d ifferentareas

,and the rainfall distribution o f previous years o f similar me

t e o r o l o g ica l conditions , not only the probable character o f the rain

fal l can be estimated,but also the probability of the occurrence of

deficiency or excess o f rainfall in any area as dependent upon orresulting from these conditions . This is what i s now attempted tobe done in the forecasts i ssued annually in June by the department

,

and which have had a fair measure o f success . For example,a full

warning wa s given in June,1891

,o f the drought in Raj putana dur

ing the monsoon rains o f that year .

It is hardly necessary t o point o u t that the m ethods employed andsketche d above are practically identical with those em ployed in givin g warning o f the approach o f storms

,and I m ay again point out

that these long-period forecasts in India are rendered possible by thepeculiar features o f the southwest monsoon air m otion over India

,

and by the rem arkable persisten cy o f many o f the abnormal conditions o f the meteorology o f that current .

5 .— CAN WE BY AUTOM ATIC RECORD S AT THREE SELECTED

STATIONS DETERM INE THE ENERGY OF A FLASH OF LIGHTNING ?

ALEXANDER M CADIE,M . A .

I may begin this paper with an answer in the afli rm a t ive . It would

be a good plan to have these observations m ade . The lightning flash

has been regarded up t o the present time as a thing accom p l ished,a

discharge betwe en the electrified cloud and the earth,over about as

soon as seen . It must be a discharge of very high potential becauseo f the length of the spark ; and the potential being great the capacitymay be small , i f, as we have som e reason to suppose , the quantity ofelectricity in a flash is not great . CV: Q .

To-day we are beginning to look at a l ightning flash from a different point o f View . We study the strain in the dielectric

,where pre

vio u s l yw e thought only o f the surface electrification ; and the ch a ra cte r o f the discharge is n ow o f great importance

,and where before

we talked o f forked,zigzag

,and Sheet lightning

,a classification som e

what l ike Luke Howard ’ s cloud classification , we talk now of “ in k

ENERGY OF A LIGHTNING FLASH . 19

pulsive rush” discharges,m eandering flashes

,etc . What we need

,

and we have in part,i s a systematic classification o f the e l e ct r i

cal discharges in the atmosphere . At one end of the list we mightplace the impulsive rush discharge

,a most intense flash

,and at

the other the gentle glow discharge which we find so frequently o n

Pikes Peak and Ben Nevis,and now definitely connect with certain

meteorological conditions . Observe,too

,that the conditions for the

protection of life and property are very diff erent for these differenttypes of discharge . Points fail to be effective under the im pulsiverush

,wh ile most eff ective with the glow .

We want,then

,to classify o u r flashes ; and to get m ore accurately

at the ch a r a ct e r Of the flash,perhaps we should attem pt to get at the

energy of each particular flash . Dr . Lodge,in his book o n

“ Lightning Rods

,

” in Chapter xv,gives the suggestion o f the editor of the

“ Electrician ” that,where thunderstorm s are frequen t and violent

,

it might be possible to set up lightning conductors for experim entalpurposes

,and thus accumulate experience concerning their behavior

m ore rapidly than at present . On a preceding page it i s al so noted

how much work could be done at meteorological stations and o bs e r

va t o r ie s“ in the matter o f accuratel y observing and recording light

ning,photographic records

,obtained by proper appliances f o r distin

g u ish in g multiple from successive flashes , being, o f course , superior

to all others . An experimental lightnin g conductor o n a flagstaffnear every m eteorological Observatory would al so be a m ost desirableaddition . It need not be associated with danger . A system o f fuses

o r cut-outs , o r an east or west steel bar , might be used to record thepassage of a flash

,and the rod need not be examined until after the

cessation o f violent disturbances . By having the conductor of diff e r e n t thickness at diff erent parts o n e could learn what size is reallylikely to be m elted . On e could al so arrange so as to gain inform a

tion about side flashes .” In the “ Philosophical Magazine,

” August,

1888,Dr . Lodge applies the mathem atical expressions fOr the real

resistance and inductance of a conductor under an alternating current t o the case of a lightning flash .

“An air-condenser with platesof any size separated by a distance h (height of cloud ) . and chargedup to bursting strain gramme weight per square centimeter ; theless strength of rare air i s hardly worth bothering about ) . Let asm al l portion o f this condenser

,o f area nb

’

,n o w discharge itself

,

being separated from the rest after the trap-door and guard-ringm anner . A volum e of dielectric rrb

zh is relieved o f strain

,and the

9

21nb

zh ergs .

K b2

4 h

ca n be pu t a t 110 h e l ectro static units .” He then calculates the

energy o f the spark is E

The capacity discharged is S and the m aximum potential

2 0 CH ICAGO M ETEOROLOGICAL CONGRESS.

inductance o f the circuit L: h (a n t i where u may be a numbernot very different from 4 o r 5, and now knowing S and L proceeds

to find the crite rion f o r the discharge to be oscillatory and to determine the rate o f alternation . The discharge will be oscillatory unlessthe resistance it meets with exceeds a certain critical value , viz

4 h u u 4 h u /u v

S

4 h

1where — ~

K )the ve l o cIty of l ight

so the critical resistance is

R02 12 0 -2 ohm s .

Suppose h t o be a m ile meters ) , b50 m eters,and a a milli

meter ; the critical resistance comes out about ohms . When

the resistance then fall s below this the discharge wil l be Oscillatory .

The impedance to a condenser discharge comes o u t

impedance 60 ohms .

Or , it i s half the critical resistance ; it depends almost en tirely uponthe amount o f space magneti zed round it ; and upon the capacity o f

the discharging condenser . Magnetic permeability,specific resist

ance,or even the thickness of the conductor

,hardly matter . The

length of the conductor does figure .

Now,while we may n o t erect a conductor a mile high

,it i s feasible

by kites,balloons

,or a er o p l a n e s to carry up a wire a millimeter

thick some 2 0 0 meters . The critical resistance would come o u t

som ething like Ohms and the impedance one-half Of this , andthe frequency constant

,n L: impedance

,something l ike

Now,the total maxim um energy “ of a given area o f cloud is easily

estimated,

” says Lodge,

“ by remembering that as soon as the electrictension of

.

the air reaches the limit o f about o n e -half gram m e weightper square centimeter disruption occurs ; and the energy o f the dielec

1tric per CIl bl C centimeter be i ng

9

; ergs , per cubi c m i le It would be

X 1012

any ordinary flash can be accounted f o r by the discharge of a verysmall portion o f charged cloud

,for an area o f 10 yards square at a

height o f a mile would give a di scharge o f over foot tons Of

energy .

” And f o r the case we have taken,some 2 0 0 m eters

,we should

have from 2 0 0 to 30 0 foot tons,or very roughly in the neighborhood

of horse power .

With three stations grouped then around a common center,pro

vide d with cameras with some type of electrometer and with meteoro~

CLOUDS AND WEATHER PREDICTION. 21

logical apparatus,we might get first the exact times of occurrence o f

all visible discharges ; and the exact a pp e a r a n ce o f the flashes,i .

not as referred to o n e plane which a Single camera would give,but the

character and dir ecti o n o f the flash in space . Many flashes startingfrom a given point undoubtedly meander

,turn and twist upon them

selves,and som e Of the seeming thickenings in single photographs

are doubtless points simply of change o f direction o f flash . Next we

would get from the potential fluctuations,as shown on the electrometer

records,the exact tim es and som ething of the individual strains

,and

,as

I have elsewhere shown,evidences o f discharges n o t visible

,and in this

way could,from a com posite o f the records o f our three stations

,get

at a very good approxim ation of the strains to which our dielectric,

the a ir between the thundercloud and ground,had been subj ected ;

and like a piece of plate armor,when the firing i s over

,w e could

exam ine and locate the places and times o f rupture .

6 .— THE UTILIZATION OF CLOUD OBSERVATIONS IN LOCAL

AND GENERAL WEATHER PRED ICTIONS .

ALEXANDER M CADIE, M . A .

In o u r daily work of forecasting weather changes,we have reached

the point where we feel the necessity o f some knowledge o f the con

d it io n s o f the upper air strata . We m a p with great success the conditions of the bottom o f the aerial ocean in which we live . By the a id

o f the telegraph we m ake invaluable synoptic charts . We have exce l l e n t ground plans o r horizontal sections , but we attempt nothingin the way Of vertical sections Of the atmosphere . The telegraph isn o t available ; som e other agency must be sought for . We

,in part

,

attem pt the exploration o f the free air by balloons and by mountainObservatories

,and when aerial navigation i s an accom plished fact we

shall doubtless have system atic and extensive surveys of the atmos

ph e r e . But until that happy time arrives,clouds must remain the

best exponents of conditions prevail ing at d ifle r e n t level s in theatmosphere . They can be made to give us even now

,with most crude

methods,information concerning the currents at d ifle r e n t heights ,

and indirectly,temperature and moisture conditions . Studied closely

and in connection with the surface i sobars,i sotherms

,and winds , the

forecaster will find in cloud m otions and formations portions of thestorm mechanism otherwise hidden from him . Fo r special as wel l

as general forecasting cloud study is important,and I desire to em

ph a s ize the need o f cloud study at places along the coast . I thinkthat if we had well equipped stations at Capes Fear

,Lookout, and

Hatteras,with cloud conditions a subj ect of special attention , we

would receive timely warning o f the occasional storm s that slip inupon us from the seaboard .

2 2 CH ICAGO M ETEOROLOGICAL CONGRESS .

C loud nomenclature and the various m ethods o f cloud m easure

ment do n o t fall strictly within the limits o f this paper . Both topicsrequire special papers . But f o r the purposes o f general forecastingwe need

,first

,a codification Of what f o r want of a nam e I shall call

“ cloud laws ”,i . e .

,the results of studies Of cloud form ation and

m ovem ent ; and secondly , some cipher scheme at once flexible and

defin ite that wil l convey to a distance the a ctu a l aspect o f the sky .

H il d ebr a n d s s o n in hi s paper (read before the Royal MeteorologicalSociety

,London

,February divides the problem into two

sections— how t o best study the relation o f formation to the physicalprocesses at work ; and then the determination o f the bearing o f

these on weather changes . In a footnote he instances the great valueand interest a series o f cloud observations would have i f m ade by asociety o f persons spe cially interested in cloud studies and Observingsystematically over a large area o f country

,

“ keeping strictly t o thesame detailed nomenclature

,e . g .

,that o f C lem ent Ley .

”

There can be n o question that forecasting would be more certainif we could connect certain types Of cloud formation with certainconditions of atmospheric circulation . It being impossible to getthe series o f Observations o f the character referred t o , I thought thata rough approxim ation m ight be made by carefully charting cloudObservations made sim ultaneously by the observers o f the WeatherBureau . The classification is that o f Luke Howard

,and I can only

repeat here the remark m ade in the discussion o f Captain Toynbee ’ spaper on cloud nam es

,by C lement Ley

,viz

“ Before the dawn o f

synoptic meteorology,Luke Howard ’ s system fi l led a need

,though it

did little to prom ote inquiry . Since that era it m ay safely be m ade

the basis o f a carefully discriminating and eclectic system o f termin o l o gy. But any endeavor to restrict ourselves to its use cuts o ff

the possibil ity o f obtaining what becomes more and more necessary ,viz .

,the power o f either comm unicating from distant localities the

actual aspect o f the sky so that this may be represented graphicallyor of recording such an aspect

,so as t o call up in the mind a vivid

idea o f the Observed phenom ena ; I believe that ten thousand yearsof Observations conducted o n Luke Howard ’ s system would give us

an absolutely futile record .

” The language i s a little strong,but

there is som e j ustification f o r i t . However,it is n o t altogether an

easy task to devise a classification so detailed as to definitely picture

up any o n e o f the numerous and Often n o t easily definable sky a spects . ‘

Taking,then

,the observations o f the Weather Bureau Observers

,

charts were m ade in the Forecast Room each morning and night,

and prove first that it is entirely practicable to construct such cloudmaps within the time allotted

,and

,second

,that we can make use

o f the same in forecasting . The particular point which these bringout is that it is possible to fix with considerable accuracy the storm



Temp e r a tu r e a m p l i tu de s a n d m e a n cl o u din ess .

S ta t io n .

'I-9

C l o u dsPh il a de lph ia , Pa

Clo u dsSa l t La ke Ci ty,Ut a h

0

Cl o u d s

C l o u d sDe n ve r

, Co l o

C l o u d sEl Pa s o , Te x

C l o u dsM i lwa u ke e ,Wis

C l o u dsCh e ye n n e , Wy

OB‘ O O O O 0 0 0 0 0

C l o u dsSa va n n a h , (wa

a c e

0 0 0 0 0 0 0 0 0 0 0 0 0 0

CLOUDS AND WEATHER PREDICTION. 25

No paper upon cloud work would be complete without reference tothe work o f Ekholm

,Ha g s t rOm ,

and H il d ebr a n d s so n in putting beforeus the question of cloud measurements . And I must n o t om it reference to the law of relative directions of lower and upper currentsas announced by Ley— high currents coming from a direction to theright o f the lower cu rrents

,and the higher up the more marked this

twist . Or in Ferrel’ s words The higher currents of the atm osphere

while moving commonly with the highest pressures,in a general way

on the right of their course,yet manifest a distinct centrifugal ten

d e n cy over the areas of l o w pressure and a centripetal over those ofhigh .

” It i s also not o u t o f place to m ention briefly the more promin e n t of the proposed cloud systems , viz .

,Luke Howard ’ s

,the essay

read before the Aske s ia n Society , 180 2—180 3 ; C lem ent Ley’ s observa

tions ; Abercrom by and H il d ebr a n d s s o n’s ; Wilson-Barker

’ s ; Abercr om by

’s ; H il d ebr a n ds s o n and Neumayer

’ s ; Dr . Carl Singer ’ s ; KOppen and Neumayer ’ s ; Dr . Ve t t in ’

s table Of average altitudes,and the

work at Blue Hil l in this country . As a matter o f perhaps more

literary than scientific value,I append a list of English cloud d e s ig

nations .Lis t of En g l ish cl o u d de sig n a tio n s

Alt o -cir r u s , a lt o -cu m u lu s .

Arko’th e cl u ds .

Au r o r a l clo u ds .

Ba l e fra e — ba le fir e .

Ba ll clo u ds,ba lly .

Ba n n e r , ba n n e r e t .

Bise .

Bu ddh a ’ s r a ys .

Ca tst a il s .

Cir r u s , cir r o -cu m u lu s,cir r o

fil u m, cir r o -n e bu la , cir r o

h a z e , cir r o -s t ra tu s , cir r o

st r ipe s , cir r o -ve lu m .

Clo u d ,clo u d a r e a , clo u d ba n k ,clo u d rin g , clo u d Sh ip , clo u dwr a ck , clo u d wr e a th , clo u dw r a ith .

Co rm iza n t .

Co rpo sa n t .Co r o n as .

Cu m u lu s, cu m u lo -cir r u s , cu

m u lo -n i m b u s , cu m u l ost ra t u s .

Da pple sky.

Da rk s e gm e n t (a u r o r a l ) .Diabl a t o n .

Fa lse cir r u s .

Fe st o o n e d , fe sto o n e d cu m u locirr u s , fe s t o o n e d cu m u lu s

,

fe st o o n e d st r a t u s .

Filly t a ils .

Fir e ba lls .

Fo g bo w .

Fract o cu m u l u s,fr a ct o

n im bu s .

Fu n n e l -sh a p e .

Glo bo -cu m u lu s .

Go a t s h a ir .

He lm -ba r .

He n sca t .

Iride sce n t .Le e side .

Lu m in o u s.

M acke r e l Sky, m a ck e r e lsca le s , m a cke r e l ba ck .

M a m m a tO-cu m u lu s .

M a ry ’ s sh ip .

M a r e’s t a ils .

M e r ry da n ce rs (a u r o ra l) .Na cr e o u s .

Nigh tca p .

Nim bu s .

No a h ’ s a rk.

Nim bo -pa lliu m ,n im bo

st r a t u s .

Pa ck e t bo ys .

Pa llio -st r a t u s .

Pa lliu m .

Po cky clo u d .

Pr o ph e t clo u d .

Pla g u e clo u d .

Po la r ba n ds .

Ra in ba lls , r a in clo u dsr a in bow .

Ra dia t io n fo g .

Rim e clo u d .

Ro cky .

Ro ll cu m u lu s .

Sa lm o n .

Scu d .

Sco tch m ist .

Sa in t Cl a r a ’ s fir e .

Sa in t Elm o ’ s fir e .

Sn ow ba n n e r s .

Spin dr ift .St o rm clo u d .

Sn o w clo u d.

Spe ct r e o f Br o cke n .

St r a t u s,st r a t o ci r r u s ,

st r a t o -cu m u lu s .

Table clo th .

Th u n de r h e a ds , th u n de r

squ a ll clo u ds .

Tu r r e t e d cu m u lu s .

To r n a do .

Un r a ve le d .

Wa t e ry sky.

We a th e r ligh t s .

Wo o lly h e a ds .

Wu lst cu m u lu s .

Wo o l ba g s .

Wrack .

Wr a ith .


7 .-AN INTERNATIONAL C IPHER CODE FOR CORRESPONDENCE RESPECTING THE AURORA AND RELATED CONDITIONS .

Dr . M . A(VEEDER.

It is presum ed that the assignment o f this subj ect to the writer i sintended to cal l f o r the results of the experience which he has hadin attempting to secure concerted observations o f the aurora . The

adoption of a special plan o f Observation or code correspondencerespecting such a phenomenon as the aurora presupposes the selectiono f the points thought to be most important in order that they maybe made the subj ect o f special observation and record for purposes o finterchange and com parison . The purpose in view in any such case

determines the character o f the record to be made . If the plan involvesnothing more than the preservation o f mem oranda

,such as may

happen to be secured incidentally in the ordinary course Of meteorological observation , and without reference to the requirements ofserious s t u dy,

‘

it i s scarcely worth while to di scuss the subj ect at anygreat length o r Offer many suggestions .Al l that can be expected

,i f nothing more than this is to be attempted

,

i s the recording o f dates and localities with perhaps some items ofdescription m ore o r less condensed

,it may be

,by the aid of the sys

tem s of classification already in ordinary use which have reference tothe presence o r absence o f arches

,streamers

,auroral waves

,the corona

,

and the like . Sti l l,the gathering into suitable records and m aking

accessible inform ation n o t more complete than thi s,has been the

means Of affording a knowledge of certain broad features . The relative prevalence of the aurora in diff erent years and its conformity tothe records Of

l

su n spots and magnetic storms has thus been shown,as

has al so the predominance o f auroras near the equinoxes,and at

interval s o f about twenty-seven days,corresponding to the tim e o f a

synodic ro t ation of the su n . By such means also the distribution o f

the aurora in belts surrounding the magnetic poles has become known .

If s o m uch i s to be learned by the aid o f Observations that have beenf o r the most part l ittle better than merely desultory, what might notbe expected from the elevation of the subj ect into a specia l department o f research to be undertaken formally and o f set purpose?In the paper on the “ Periodic and Non-periodic Fluctuations in

Lati tude o f Storm Tracks,

” presented by the writer in the Sectiono f Marine Meteorology o f this Congress , it is shown that im portantrelations to meteorology may be involved in the operation o f theforces concerned in the production of the aurora . This being thecase

,i t s behavior i s as worthy of careful record as is temperature ,

pressure,o r any other meteorological elem ent .

The experience which the writer has had in this regard has had

INTERNATIONAL CIPHER AND AURORAS. 2 7

reference m ore particularly to m ethods o f recording Observations,

and not to any system o f code correspondence based thereon . Aspecimen o f the form s which he h a s em ployed for securing such

recor ds is appended to this pa per . The points upon which thegreatest possible stress is laid are the giving o f the tim es o f observations and o f all prom inent features

,and the noting specifically o f

verifications o f the absence o f the aurora as well a s it s presence,and

the recording o f frequent estimates of the extent o f sky covered . By

the aid o f such data it becom es possible to attack the questions as togeographical distribution

,altitude

,coincidence with magnetic per

t u rba t io n s , and the like , positively and directly , and not rem otely andinferentially .

Some o f the results o f this system of Observation are indicated in thepaper on Storm Tracks

,to which reference h a s been made

,and in other

notes and articles of similar tenor,and do n o t need to be rehearsed in

the present connection . Suffice it to say that the observations recordedin this precise way are proving to be extrem ely valuable .

As regards the general description s to be given in connection withthese Observations

,it is found that great freedom and fullness in giv

ing detail s are very desirable . No t unfrequently item s o f descriptionthat would be om itted in a code system o f abbreviating and sum marizing prove t o be o f the very h ighest interest . Further experience i srequired before it Ca n be fully known what points are Of such immediate and practical interest as to j ustify or require the adoption o f SO

elaborate an arrangement as an international cipher code for theircom m unication . Stil l

,there are indications that something in this

line is worth attem pting, and that the tim e is surely com ing when asystem o f correspondence having reference to the wh o l e

x

r a n g e o f phe

h om ena Of which the aurora is the visible expression will be wellnigh indispensable for purposes o f weather predictions as wel l as theadvancem ent o f general scientific research . It wil l n o t be advisable ,perhaps

,to be t o o urgent in attempting to bring about such an ar

rangem ent prematurely . No t until the facts and principles involvedare fully appreciated and recognized by the scientific world generallywill the dem and f o r the adoption o f an international code systembecome SO em phatic that it can n o t be disregarded .

The question n ow i s a s to the best m eans o f arousing such livelyinterest m ost rapidly and eff ectually . It wil l contribute som ewhatto this end

,perhaps

,and compel attention to the m erits o f the case ,

to describe briefly what would be an ideal system Of inter-com m uni

cation a t the present stage o f progress Of the research respecting theaurora and related conditions .

If there could be brought together up on daily synoptic charts , alongwith other meteorological data which it is now customary t o presentin this way

,information also in respect to all auroras seen over as

2 8 CH ICAGO M ETEOROLOGICAL CONGRESS .

wide an area as possible,and likewise som e indication Of the extent

of prevalence of thunderstorm s,together with notes from the mag

netic observatories a s to the times and extent o f any perturbationsrecorded

,and information al so as to the geographical distribution o f

any earth-currents that may have been felt o n the telegraph lines,

and in addition,some description o f the coincident solar condition s

o n which th is class Of phenom ena evidently depends , any ch a r a ct e r istic relations to intensification o f storm s o r changes in the d is t r ibution o f atmospheric pressure would soon become apparent. Such an

arrangement would require simply an extension o f the telegraphic codesystem n o w in ordinary use for the com munication of meteorologicaldata s o a s to comprise features n o t heretofore taken into the account .It i s evident that even tentative eff orts in this direction would

arouse a lively interest , and would certainly stimulate critici sm which ,whether adverse o r favorable , would tend to increase o f knowledge , suchas could never result from the utter stagnation and neglect t o whichthis class o f research has been subj ected for extended periods . Such

a plan would inevitably bring to a practical test the suggestions thathave been m ade in various quarters recently as to the part whichelectro-magnetic forces of solar origin play in atmospheric control

,

and would tend to eliminate errors and crudities which are,to a cer

tain extent , unavoidable in the prosecution o f a new line Of research ,and if there be a residuum o f truth it would be shown beyond a per

adventure,and its practical value dem onstrated .

These suggestions are the outgrowth o f the practice which the writerhas m aintained f o r m any years o f j ournaliz ing phenom ena o f thisclass o n a daily record . As the result

,the conviction has grown that

the principle o f electro-magnetic induction o f dynam ic origin plays

a far-reachin g part in the economy o f the solar system,and that it i s

concerned in atm ospheric control in ways that are only j ust beginning

to be understood . From h is point o f view,therefore

,the scheme o f

com municating and recording Observations above described is wel lworth trying . How it will impress other m inds remains to be seen

.

!Co py o f fo rm u s e d by t h e Pe a ryArct icExp e di t io n in r e co rd in g a u r o r a l phe n o m e n a ]

Na m e a n d a ddr e ss o f o bs e rve r Da t e 189 .

La t it u de a n d lo n g it u de o f Kin d o f t im e u s e dOBSERVATIONS OF THE AURORA in co -o pe r a tio n with Civil En g in e e r Pe a ry , U. S. N.

,

in No r t h e rn Gr e e nla n d , a r e t o be e n t e r e d a s fo llo ws : Th e a bs e n ce of th e a u r o r a ist o be in dica t e d by e n t e r in g in th e p r o p e r spa ce th e fig u r e s sh ow in g th e m in u t e s o f th eh o u r du r in g wh ich s uch abse n ce wa s ve r ifie d by Obse rva t io n . Th u s th e e n t ry “ 0 —10 ”

in th e co lu m n h e a de d 7 t o 8 p . m . w o u ld be u n de r s t o o d a s sh o win g t h a t o bse rva t io n sw e r e m a de fr o m t o p . m . a n d th a t t h e r e wa s n o a u r o r a a t t h a t in st a n t . IfObse rva t io n is im po s s ible fr o m clo u din e ss o r a ny ca u se it wil l be s u fficie n t t o le a vet h e spa ce s e n t ir e ly bla n k . Th e p r e s en ce of th e a u r o r a is t o be in dica t e d by w r it in gAURORA in th e p r o pe r spa ce a n d g ivin g t h e e xa ct t im e s a n d o th e r it e m s u n de r th e h e a do f De scr ipt io n s . NO m a t t e r wh a t e ls e m aybe r e co rde d

,it is o f th e u tm o s t im po r t a nce

t o g ive a s a ccu ra t e ly a s po ss ible th e t im e s o f a ny su dde n in cr e a s e o r dim in u t io n in



chance or to the knowledge (even if i t be occult ) Of certain laws , sin ce ,in the latter case

,the weather

,according to opposite predictions , would

have been sensibly different , while in the first case it wo u ld remainthe same

.The supposition is always that a sufficiently large number

o f predictions has been included in the verification , s ince s o long

as the law o f large numbers does not enter , the separation betweenchance and law is destroyed and no argum ent i s possible , unless it

be that th e prognosticator i s n o t infall ible .

For a better understanding o f the above let u s take an example

from the summ er o f 1883 . There were two kinds o f predictions in

ve s t ig a t e d ,which we will designate by classes A and B . The second

columns give the contents o f the predictions :

Th e s u cce e d in g w e a t h e r wa s

W a rm . No rm a l . Co l d .

Per cen t .

It i s seen that,according t o the prediction “A ,

” the real weatherremained the same whether the prediction read either warm o r cold

but showed a m arked contrast with respect to the prediction “ B .

Now the predictions “A ”

(made for a month in advance ) are truechance predictions

,although they excited great surprise and by m any

are regarded as sati sfactory .

“ B ” are synchronous daily predictions

o f a meteorological institute . In cases o f thi s sort this m ethod isentirely conclusive and sufficien t .

Concerning the “ B ” predictions , this table Shows that between themand the following weather there exists an evident relation . In general

there can be no question as to the predictions f o r only twenty-fourhours ahead . Even with the m ost unskilful predictor there would bea general agreement for s o short a time . It is to be asked

,therefore

,

h o w close this connection i s , and , indeed , i f this relation can beexpressed by a single numerical value . Detailed researches o n thesubj ect show that

,unfortunately

,a series o f unknown values enters

,

and that an irreproachable derivation o f such a simple num ber for

the expression o f the worth Of a prediction i s im possible— almost as

im possible as to estimate the value o f a person o r a nation by a

numerical expression . Such expressions have been proposed,for

exam ple,in America in 1884 , by Messrs . G . K . Gilbert and C . S . Peirce ,

but the formulae given,although in genious

,are o n e sided

,and have

not a universal application .

Stil l less , as a measure o f the value o f a prediction , can the usual

M ETHOD OF TESTING PREDICTIONS. 31

percentage o f success be used in which the influence o f chance is

neglected and the phenom ena are treated without regard to theirfrequency . If o n e wishes to have an answer to the question

,how

Often a certain prediction is followed by such and such weather,

without regard to the reason,naturally such a calculation o f per

centage is quite satisfactory,only in the first place we d o not know

what this can teach us,and secondly

,as we shall see directly

,from

the method of calculation it i s largely influenced by the personal int e rp r e t a t io n of the verifier . It is

,f o r exam ple

,very instructive t o

know that the cases o f tornado predictions are s o divided that

in cases,when n o tornado w a s predicted

,in only 2 3 wa s one

observed,in cases not

,a n d am ong the 10 0 cases where tornadoes

were foretold they happened in 28 cases and did not occur in 72 .

To r n a do o ccu r r e d . No t o r n a do .

Per cen t . Pe r ce n t .

No t o r n a do pr e dict e dTo r n a do pr e dict e d

From this statement it i s seen that the predictor has in the lastcases found the tendency to tornado formation . But what good

2 70 8

2 80 3

success of this prediction,a seemingly high num ber

,but which i s

,

nevertheless , inferior to that which would be had if, without trouble ,a daily prediction o f

“ no tornado ” had been made,for then thi s

number would be 3323—2 per cent .

The second cause which makes the value o f the percentage of success

,as usually calculated

,seem very sm all , is the indefinite nature

of the fundamental material,and therefore , the consequent im p o s s i

bil ity o f making the verification com parable .

The determination o f what follows a prediction has been generally

sought o n the basis o f the verification of the general character o fthe day over a large territory

,which is fixed by estimation . As to

the sincere general wish to get at the truth by this estimation therecan be no doubt . Only according to the interpretation o f the words

and the pessim i stic or optimistic tendency of the verifier such estimates must differ greatly ; and even when in an institute , by preciseinstructions they are rendered independent o f the person

,it will

never be possible to introduce such instructions and their in t e rp r et a t io n in other institutes , o r even to render them so invariable at thesame institute that slight changes in their application wil l n o t affectthe results in an uncontrollable m a nner . For

,in order to determine

the influence of each o f these rules o r usages o n the result,com pre

h e n sive investigations would be needed , which have not yet been madeand for which time would be required that could better be used in the

i s it when the.

fraction i s given as the percentage o f


extension o f the basis of weather predictions . Besides , these percen tages o f success , although as a rule they are u n d e r s t o o d

'

t o be therelation which the successful predictions bear to the whole number

,

yet they are generally to be cons idered a s the means o f the testswhich the prediction would give if arranged in various gradations .

Thus,most of the German institutes have arranged the predictions

according to their results in three grades— the Bavarian Institute,

since January,1882 , h a s used five grades— whose values are 10 0

,50 ,

and O,or 10 0

, 75 , 50 , 2 5 , and 0 , a s percentages of their entire accuracy

,but the arithmetical mean o f these num bers h a s been taken as

the percentage Of success of the prediction .

Fo r the above reasons , at the S e ewa r te since 1886 the calculationOf percentages of success has been abandoned , and in its place statistical summ aries in the form already indicated

,with as accurate a

basis as possible,have been introduced . Al so

,in the MonthlyWeather

Review o f the Washington Weather Bureau,since January

,1892 , the

heading “ Ve r ifica t io n s ” has disappeared— a sign that at this greatinstitute the value o f these numbers i s less considered than formerly .

1

The method followed by the S e ewa r te carries out these principles1. In place of an approximate representation o f the whole space

and time covered,a precise determination is given f o r 8 a . m . and 2

p . m . at only three places in the German Empire .

2 . For this verification the conditions , as shown by the meteorological observations

,are laid down in the following scheme

8 a . m . 2 p . m .

No ve m be r 4 k u l e d n u m s h

No ve m be r 5 n z m n r

In which place 1 gives the deviation o f the temperature from thenorm al k cold (negative deviation greater than n = h o rm a l

(deviation 0 to w warm (positive deviation greater thanPlace 2 gives tem perature change in twenty-four hours : a de

crease ; u stationary (change less than 2 increase . Place3 gives the wind force : l : l ight (0 to m m oderate (3 toetc . Place 4 gives the wind direction . Place 5 gives the p r e cipit a t io n .

3 . As far as the det ermination o f the expression permits,the pre

dictions are also arranged in the same way in classes , and then theagreement o f p r e d l ct io n s and weather are worked up in comprehensivecalculations and put in a tabular form . Further detail s will be foundin the m onthly reports o f the S e ewa r te from 1886—1891

,and especially

in a supplem ent for 1886 .

Starting from similar ideas,the storm warnings o f the S e ewa r te

1Th e ca lcu la t io n o f th e p e rce n t a g e s o f ve rifica t io n s h a s n o t be e n a ba n do n e d,bu t th e

r e su lt s in s t e a d o f be in g pu blish e d m o n thly in th e We a th e r Re vie w a r e publish e d a n n u

a lly ia th e An n u a l Re po r t o f th e Ch ie f o f th e We a th e r Bu r e a u — EDITOR.

M ETHODS OF TESTING PREDICTIONS. 33

since 1886 differ from those previously verified,and in this way— by

a representation of the changes in the wind velocity according to theanemometer indications at the normal Observing stations of theS e ewa r t e in the six hours preceding and in the thirty-six hours fo llowing each storm warning .

The comprehensive tables which have been compiled and published

by the Se ewa‘

r te present much material for study which i s well consolidated and is thoroughly controllable and comparable . Herein is

a,precise answer to many questions whose clear com prehension andsolution are o f great im portance for the u s e o f the prediction service .

For making up reports to the public,o n the contrary

,they are too

complicated,and demand more knowledge o f the difficult questions

concerning calculation of probabilities than is often found . The

ultimate question,which for practical purposes is preeminent

,is

,

what i s the value o f weather predictions Still,this question must

remain under all circumstances unsettled,no matter h ow many o r

how well founded are the figures cited concerning the measure o f thesuccess Of the predictions . When

,for exam ple

,an accuracy o f 80 per

cent i s found f o r the prediction,one person m ay consider this per

centage as very good,another that the 2 0 per cent o f failure is suffi

cient to invalidate it,a third may say that he does n o t find just what

he needs in the prediction,t h e fourth may state that the prediction

is worthless if it does not com e t o h im sooner,the fifth may declare

the most successful warnings to be useless because he might havepredicted j ust as wel l the bad weather

,and s o o n . A strict proof of

the practical worth Of weather predictions and storm warnings is impossible

,but the general impression o f competent persons must be

considered conclusive,and in this combined impression the fulfil lment

Of the prophecy counts for only o n e,though perhaps the most im

portant,o f many circumstances .

If we a sk ourselves whether the great working experience whichhas been gained in weather predictions during twenty years be ars aproportionate relation to the results Obtained

,we must decidedly say

no . Among the people there i s a widespread belief that the predictor,

by a strict and thorough verification of the prediction made by him,

m ust make great progress in h is j udgment and experience . Were thisso

,then the verification s would be a duty and worthy of the attention

o f the meteorologists to whom the task o f issuing predictions fal ls .Really

,this task is

,as anyone wh o has prosecuted it long will adm it ,

o n the whole a very thankless one,and the time which is expended

o n it can be much better em ployed upon meteorological investigations in stati stical meteorology upon which the predictions rest . Ifo n e will study critically the prediction service , the investigationsshould at least be as thorough and carried out according to as strictmethods as at th e Se ewa r te . A report to the public should , in thi

3


case,be as free as possible from numerical data , because figu res on

this subj ect are generally misunderstood and real ly carry only prao

tical convictions in certain directions and only to special ists . For

such a report,a statement

,as direct and as free from excuses as pos

sible,with concrete examples and the opinion Of special ists , i s the

most convincing and suitable .

9 .— PRESENT COND ITION OF THE WEATHER SERVICE

PROPOS ITIONS FOR ITS IM PROVEM ENT.

Pr o f. Dr . W . J . VANBEBBER .

In the states in which weather telegraphy has been used in connection with weather forecasts

,either for agriculture o r navigation , the

experience has been generally that the views on the value of forecastshave been partly overstated and partly understated , and that thehopes which the weather service first inspired have n o t been fulfi l led .

The reason f o r thi s l ies in the fact that o u r knowledge o f the causesof meteorological phenomena is so meager that the practical weatherprediction i s always accompanied by frequent

.

and great failure,

which j eopardize m ore or less its usefulness,that the progress o f p r a c

tical meteorology is uncommonly slow,and scarcely noticeable

,and ,

finally,that the knowledge o f meteorology generally , and especially

of the principles underlying prediction,i s so slight that such influ

e n ce s as the moon , have , even with learned people , equal o r greatervalue than the prediction issued by scientific institutions .It i s incontestably true

,that the predictions o f the institutes really

have a basis which i s capable o f further development,and that

already in the present state o f our knowledge of atmospheric phen o m e n a and their changes , they may be useful for practical vocations ,provided that all available means shall be used in their i ssue anddistribution

,and that the public understand how to estimate their

worth .

On account of the great practical m eaning o f the forecasts whichincreases with their accuracy as well as with the extension of theinterval of time covered , it appears our duty t o strengthen and tospread the fundamental principles in order to accomplish what ispossible and to m eet the wants o f the public in every way.

It seems difficult to put together the success which has beenattained in the several states in order to have an adequate survey o fthe util ity of the weather service in these states . True

,figures show

ing the percentage o f success have been published by the variousInstitutes , but these have for the estim ation o f the real worth andutil ity of weather prediction but very little value

,since in the veri

fica t io n a series o f circumstances must be considered which influencethe final result greatly . Chief among these are the proba bi l ity o f

THE WEATHER SERVICE. 35

the occurrence of a weather phenomenon,o r chance

,also the persist

ent tendency o f the weather,and finally

,the greater o r less use which

can be derived from the forecast . So long as all these points do notfind recognition in the verification

,the percentages o f success

,no

matter h o w scientific and rigorous the verifications may be,are only

of secondary im portance,and can n o t , as may easily be shown , be

regarded as proper criteria for the success o r failure Of the prediction . But

,in the efforts to take account o f al l these circumstances

we come to the difficulty , which appears the more insurm ountableand the greater

,that much must always be left to arbitrary decision .

Upon these points a detailed report of Prof . Dr . KOpp e n i s pre

sented to the Congress,so that further discussion is unnecessary .

‘

The best,and up to n o w

,the only correct and definite scale for

the utility of prediction i s the j udgment of the public and of that portion of the public which is most interested in the predictions , that i st o say , th e coast inhabitants an d the agriculturists , who , from theirvocations

,are most dependent upon wind and weather . As regards

the first class,opinions have been given in the United States

,by

the London Meteorological Office,and by the De u tsch e Se ewa r te .

According t o these the coast dwellers regard the storm warnings,in

spite o f the frequent failures,as a desirable arrangement . As a fur

ther confirmation of the util ity o f the storm warnings I would addthe fact that o n the German coast the Provincial authorities andsome private persons have erected and maintained signal stations attheir o wn cost

,where those of the Governm ent were n o t su flicie n t .

The j udgment of the public on weather predictions for agriculturaland industrial purposes differs generally so much that it appears im p o ssible at present to draw valid conclusions as to the value o f the same ;however

,I have always had the experience , as i s the case elsewhere ,

that those persons who have a direct interest in the predictions placea greater value upon them than those who regard the thing as m orein difle r e n t to their professions .On the whole , it can be a fli rm e d that even in the present state o f

weather prediction great use can be derived for practical life,so that

the practice of it by Institutions ought not t o be given up or o u r

tailed . Therefore it appears a pressing necessity t o employ al l suitable means to advance weather prophecy

,such as (relating to weather

telegraphy ) the speedy col lection and distribution of the reportsand ( regarding prediction ) the application o f accum ulated experiencewhich can add t o the progress o f the service , that is to say, the extension o f the predictions .

1 Se e Se ctio n I, p . 30 . Se e , a lso , Va n Be bbe r : “ Die Erg e bn iss e de r We t t e rpr o g

n o se n ,” M o n a tsber ich t d e r D eu tsch en Se ewa r t e , 1886

“Erg e bn isse de r St u rmwa rn u n g e n ,

”ib. , 1889 a n d Da s We t ter , 1889 , p . 2 68. M o n a t l ich e Uebe r sich t dcr

Witt eru n g , Hambu rg , 1882. M on a tsber ich t de r Se ewa r te , Ham bu rg , 1889 .


The greatest,and also the best organized , system o f weather teleg

r a phyi s in the United States

,s o that we may take this as a m odel

f o r the other countries . In Eu rope there must be radical reforms ,the most pressing o f which we will briefly , but emphatically, mention

Of the greatest importance i s the acceleration o f the telegraphicservice

,both f o r the incoming and outgoing telegrams , and upon an

international basis. Fo r this purpose the introduction o f a circuit

system (as is used in the United States and latterly for other purposesin Germany ) appears necessary . It would suffice , if, immediatelyafter the observation

,the lines should be placed at the disposal of

the Central Institutes for o n e -half to o n e hour , during which timethey could be placed in possession Of the united telegraphic material .

This telegraphic m atter,including the num ber Of stations and the

scope o f the data , must Of course be reduced a s much as possibl e .

With this the introduction o f simultaneous (world ) time is necessary .

The advantages and disadvantages as regards local time would seem

to counterbalance o n e a nother .It i s greatly to be deplored that in E u rope , although it i s recog

nized as a necessity,uniform hours of Observation can not be agreed

upon . On this point the Institutes Should com e to a decision .

As soon a s the data are collated they should be sent to a numbero f secondary stations so that within a few hours after the working up

o f the Observations the deduced synopsis and predictions can begiven t o the public . A suitable remuneration fo r the trouble of theobservers is desirable .

More frequent information,at least three times a day

,as is now

done at many Institutes,increases greatly the efficiency o f the weather

service,a s well as more frequent i ssue o f predictions when it suits the

requirements o f the public,for which a principal hour can be easily

agreed upon by the Institutions . The idea first suggested by BuysBallot to connect the registering apparatus of the chief station cont in u a l l y o r occasionally

,SO that at any time the course Of the distant

weather elements may be known tele-m eteorology ) was nothing morethan a dream of the meteorologist which should Ope rate for a shorttime over a small region . An d yet this idea seem s very useful forthe purposes o f weather forecasts

,and especially for storm warnings

,

so that from its m oderate cost it i s t o be recom mended f o r in t r o du ction over l im ited areas

,a s I have already shown . If the chief tele

graphic lines , which at certain hours , especially during the night , aren o t used , could be arranged f o r tele-m eteorology , the changes of theweather elem ents at a series o f distant stations might be registeredcontinually o r intermittently at the principal stations .The principles wh ich in o u r region are used f o r the m akin g o f fore

ca s t s , ,I have given in m y book o n

“Weather Predictions,

” and the'

Opinion expressed is that the fundamental principles. a re th e same



wind ; next, those which depend chiefly o n the wind and air trans

po r t a t io n ,that is to s ay, the temperature phenomena and to some

degree the hygrometric conditions o f the air . According to this , theprediction o f clouds , fog , and precipitation is always more difficultthan those of wind and temperature , because here , besides the airmovement

,other factors occur

,such as vertical air currents , topo

graphical conditions,etc . From what has been said it follows that

the results o f the storm warnings are more favorable than those ofthe agricultural predictions

,although the percentage o f success seems

to prove the contrary . The reason is that with cloud , fog , precipitation and in less degree with tem perature , n o t only chance but alsothe persistent tendency play an important role , while the probabil ityo f the occurrence o f a storm and o f it s continuance is relatively ex

t r e m e l y small , so that even percentages o f success which hardly reach

50 ought not t o be regarded as unfavorable , though predictions o f theseverity Of the storm and the timeliness o f the warnings are of muchim portance . Evidently predictions which rest only o n chance o r o n

persistence are absurd ; for it is the changes of weather mainly whichmust be predicted .

Snowdrifts occur m ost frequently in certain types of weather , andtherefore

,the communication of such warnings to the railroads ought

to be valuable . Such an arrangement exists in Russia since 1891,and the results have n o t been unfavorable , seeing that most railroadsregard the warnings as useful .That predictions o f the changes Of the height o f rivers (floods )

can well be made has been sufficiently proved by the experience inthe Unite d States , in Bohemia , and in other countries .The value o f the predictions

,aside from their trustworthiness

,i s

also dependent upon the length o f time which they cover . Up tonow m ost of t h e Institutes issue forecasts only for the followingtwenty-four hours o r for the ensuing civil day . The shortness of

this time does not correspond to practical needs ; often the prediction only com es to the public o n the day f o r which it applies . Aprediction f o r two

,three

,or more days in advance

,if the accuracy

w a s n o t very much less than that f o r o n e i ssued o n e day ahead,

would be o f the greatest value . It i s only a question whether in thepresent state o f practical weather knowledge such a step is desirable .

Everywhere experience h a s shown that in general over large regionsthe sam e weather condition lasts a considerable tim e and then

,

final ly , either slowly or suddenly , i s transformed into another moreor less Opposed ; s o that

,for exam ple

,periods o f dull

,rainy

,and

storm y weather are followed by clear,dry

,and calm weather with

wh ich the te m perature phenom ena,chiefly dependent upo n the air

m ovem ent, o n the season , and on the cloudiness , have to do . If a

typical weather condition is form ed,it appears that a long-range


prediction can safely be made,and it only remains to announce an

alteration in the character o f the weather o r a change o f weather .

In long-range predictions two cases occur, viz (1) To determine

the degree o f probabil ity that the weather character will last a longeror shorter tim e , and 2 ) to predict this change . The last is by farth e most difficult, and we ought not to forget that a critical point o fweather prophecy l ies here— a n uncertainty which it must be the aimof practical meteorology to remove

,but whose full accompl ishm ent

ca n not be expected at present .

In weather forecasting , the position o f the country relative to itssystem of stations and to the track o f low p

’

res'

sure areas is im portant .

The European countries lying to the westward,such as the Iberian

Peninsula , France , Great Britain , and Norway, are hardly in aposition to issue long-range predictions whose success would becomparable with those issued by the countries lying to the eastward .

Towards the south o f the globe the changes in general weathercharacter become less

,but

,o n the other hand ; local phenomena are

more marked . In northern Europe typical weather phenomena areth e rule , but the propagation and the changes o f the depressions

( secondaries ) Show many changes in which the disturbances comingfrom northwest and west present the greatest anomalies .It would be of the greatest advantage fo r long-range forecasts if

the weather service stretched westward into the ocean,both by weather

telegrams from the Faroe Islands,Greenland

,and the Azores and by

telegraphic reports in the eastern regions of the North Atlantic Ocean,

from the ports of the trans-Atlantic steamers,which Often outstrip

in speed the depressions . Thereby would the costly trans-Atlantictelegrams received from Washington

,concerning the weather in the

west o f the North Atlantic,have more value .

The situation of the eastern United States i s favorable for weatherpredictions f o r some time ahead

,although the movemen t o f the

maxima and minima is much quicker than in Europe,and co n s e

quently the changeableness o f the weather i s much greater than inEurope

,and the more s o

,since the northern and southern air cur

rents present such extreme contrasts as occur nowhere else over soextensive a region .

In spite of al l these ideas,long-range weather predictions can only

,

according to m ymind , be recommended when they are issued withthe necessary prudence and when they are made with a strong probabil ity o f success . In conj unction with these the customary forecastfor the following day should be m aintained

,a s i s d o n e .a t the Weather

Bureau at Washington . Recently,the S e

‘

ewa r t e has given in its

weather sum m ary opinions o f the probable course o f the weather foran indeterminate time

,as soon as the weather situation warranted it ,

and wi th good results . Latterly,also

,in Switzerland long-range


predictions have been made (but n o t published ) , and not without

success .The efficiency o f the weather prediction can be increased in a high

degree by teaching the public the ruling principles and to connectthe local Observations

,m ade with or without instruments , with the

general atmospheric conditions,so that in certain cases it can j udge

why the actual course o f the weather agrees o r does not agree withthe predicted

,and

,under certain conditions

,h o w far the forecasts

must be modified . Local Observations in com bination with the general weather situation give results which are n o t to be undervalued ,sinde in most cases they take into account the changes which theweather conditions undergo in a certain place . In order to enablethe public to follow the weather conditions day by day , the circulation o f newspaper weather charts appears very desirable , and here Imay cite the efforts o f the Berlin Weather Bureau , which , increasingyear by year

,at present furnishes weather maps to seven of the great

daily papers,besides those posted o n the numero

'

u s Urania columns .

On the other‘

hand,it i s to be desired that the chart i ssued by the

Institute should be sold at a moderate price,or

,i f possible , dis t r ib

u t e d grati s . The free distribution of the forecast telegram is also tobe recom mended .

Unfortunately,we have to admit that such a desirable understand

ing o f practical weather lore i s unknown to the public,as well as to

learned persons,and we must further adm it that the blame rests partly

upon the fact that most meteorologists consider it sufficient to present to the public the relatively few principal doctrines o f practicalmeteorology

,and d o n o t expose superstitious views o r l im it their

belief . I have already stated the fact that the weather bulletins andcharts i ssued by the m eteorological institutes

,and partly al so the

predictions,have only a smal l practical value if their comprehension

be wanting . It is the duty o f every meteorologist,s o far as he can

to strive in this direction,not only o n the ground o f util ity but al so

for the advancement o f science .

Finally , the accuracy of weather forecasts can be greatly increased ifthe specific cases be com pared with similar previous cases . Therefore

it i s very desirable to arrange the weather charts o f the previous yearsaccording to general m ethods

,such as storm tracks

,in order that

these comparisons m aybe at once instituted . Thereby our experience

and al so o u r skill in making predictions wil l be greatly advanced ;and we are soon able

,at the sight of any weather chart

,to form a j udg

m ent a s to t h e .p r obabl e sequence o f the weather conditions . Such a

procedure must be attended with good results,as m y o wn experience

has shown . It appears desirable that f o r vast regions,such as North

Am erica and Europe , with the neighboring oceans , a num erous collection Of system atically arranged charts should be published which


would permit the course o f the elements to be traced from the preceding to the following day . Such an atlas would be o f great value

,

n o t alone for the Institute but also for the public,which Is now Often

able to form an idea Of the prevailing weather conditions from thenewspaper weather charts . A special atlas for the agricultural forecasts and o n e for the storm warnings appears in any case necessary

.

Since the Vienna Congress and the Utrecht Conference there hasbeen little accomplished by the meteorological congresses and confe r e n ce s ; therefore it is to be hoped that at the present Congress them ost pressing needs wil l be sati sfied .

APPENDICES .

!Ext r a ct s fr o m le t t e r s r e ce ive d by Dr . Va n Be bbe r fr o m r e pr e s e n t a t ive s o f we a th e r

s e rvice s in r e ply t o qu e st io n s r e la t in g t o t h e pr e s e n t co n dit io n o f th e w e a th e r.

se rvice

in th e ir r e spe ct ive co u n t r ie s ]

I.

— METEOROLOGICAL SERVICE,CANADA .

1. Th e deg r e e of a ccu r a cy of o u r f o r eca s ts of t emp e r a tu r e , r a in,a n d

win d fo r o n e,two

,o r th r e e da ys in a dva n ce .

2 . As t o th e p r in cip l e s o n wh ich th e f o r eca s ts dep e n d , a n d th e ch a r a cte r

of th e we a th e r we a r e a bl e t o p r e dict .

3 . As to th e a dvisa bil i ty of p r edictin g r a in,a n d th e ext e n t to wh ich we

sh o u ld p r e dict t emp e r a tu r e o r o th e r ch a n g e s vi ewe d f r om th e s ta n d

p o in t of wh a t i t is p o ssibl e t o p r e dict wi th a f a ir d eg r e e of su c

ce ss,a n d wh a t i t i s th e p u blic ca r e s t o kn ow

1. The ordinary forecasts o f t h e Canadian Service are issued fromthe Central Office at Toronto at 11 p . m . daily

,and are distributed

by the various telegraph com panies to nearly every telegraph oflice

in the Older provinces , and in Manitoba . The forecast i s for thetwenty-four hours from 8 a . m . Of the m orning after i ssue to 8 a . m .

Of the following day,i . e .

,practically a thirty-Six-hour prediction ; a

supplementary forecast is made at 10 a . m . each day , m odifying , if

necessary,that o f the previous night

,but is n o t very generally util

ize d,a s the Toronto and Montreal evening papers and the Toronto

Board o f Trade are the only means by wh ich the publ ic can Obtainthem

,unless by direct inquiry from Toronto Observatory o r the tele

graph Oflice s at Toronto .

In making the ordinary forecasts the predicting officer at Toronto

endeavors to give the p u bl ic a s accurate an outline a s possible o f theweather during the prescribed period ; he i s not bound by any hardand fast rule to predict for every subj ect that m ay be included in the

general word “ weather,such a s wind velocity and direction , tem per

ature,rain

,and weather in the more restricted m eteorological sense

,

although in every instance he feels bound,even when grave difii

‘

cu l t ie s are to be contended against,to forecast as to the probability


o f rain for at least that portion o f the prescribed period which liesbetween 8 a . In . and 11 p . m . of the next day .

A s a rule,a prediction is made fo r wind velocity and direction ,

weather,tem perature

,and rainfall

,both as regards time and am ount

f o r each of the various districts into which the Dom inion has been

divided . Each morning the forecasts o f the previous day are compared with the actual weather

,and each item entered in a table under

the headings “ number o f predictions ,” number fully verified ,

”

“ number partly verified,

” number not verified .

” At the end o f themonth

,when more num erous reports have been received , these figu res

are checked,and if necessary changes made

,and a percentage struck

by taking half those partly verified ” as verified,and t h e other half

as not verified . In every instance rain i s counted as a prediction ,absence of rain when predicted being coun ted a failure , as is al sorain when no prediction o f it has been m ade . The tables on pp . 43

and 44 show the degree of success with which we meet in forecastingrain

,wind

,and temperature

,and that

,roughly speaking , we are about

as likely to fall into error by predicting it too Often , as by not predicting it often enough .

It sometimes happens,in the more settled weather

,that the pre

dicting Officer feels very sure o f his ground and makes a two o r threeday prediction

,but no separate percentage o f verification o f such

predictions h a s been kept . Telegraphic and telephonic inquiries forexte nded predictions are being continually received at the Observatory

,a n d the fact that certain firm s

,of various descriptions

,have f o r

years made a practice o f asking for extended forecasts,and by word

o f m outh and by letter , acknowledged their usefulness , proves thatsuch forecasts are

,to say the least

,fairly successful .

2 . The forecasts a s issued from Toronto depend altogether o n aknowledge Obtained from practice and study of the movements ofareas o f high and l o w pressure o n this continent

,and the weather

which,under various circum stances

,accompanies these areas

,taking

fully into account in predicting f o r certain districts the influencethat winds Of d ifie r e n t directions and increased or diminished mois

ture wil l have in the various cases . The predicting Officer has byexperience learned m uch a s to the influence that areas of high pressure and Of low pressure will probably have o n each other as regardsdevelopm ent o r dispersion and rate o f movement , and makes his predictions o n the ba sis that the pressure changes wil l be as he anticipates . It is n o t an uncom m on thing for h im

,when an abnorm al

m ovem ent Of a cyclone has occurred,to base a prediction o n the

assumption that such movem ent w a s either directly o r indirectly

caused by another cyclone beyond the region of observation,for

instance at s e a, and in th is m anner m any storms on o u r seaboard

have been foreseen , the existence o f wh ich would otherwise not havebeen suspecte d .


Although,up t o the present time

,very little use has been made o f

the kind and direction o f upper clouds,it i s full well recognized that

a greater knowledge o f the relationship o f the upper currents tocyclones and anti-cyclones may ultimately lead to more exact andextended forecasts . Even at present weather prognostics dependenton clouds are n o t by a n ymeans ignored , as in cases o f doubt as to thehovering o r change in direction of movement o f a cyclone

,such indi

cations are at times o f decided value .

3 . What i s it the public cares t o know? The mariner wants toknow

,j ust prior to sailing, the force and direction o f wind he may

expect during periods varying from a few hours to several days ; fishermen are ordinarily satisfied with twenty-four-hour forecasts ; theagriculturist wants to know general ly as t o the likelihood of rainduring harvest time

,and

,in the case o f the Northwest farmer

,the

likelihood o f early frosts ; the shipper of perishable goods wants toknow the best time to Ship in order to escape severe frost ; the general public wants t o know the general character Of the weather to beexpected o n any given day and the day fol lowing .

Ou r ordinary forecast percentages Show that we are able to predictdirection and velocity Of wind with tolerable accuracy f o r a period ofthirty-six hours

,and

,at tim es

,for a longer period

,and we have a per

centage o f verification o f storm warnings of 84 per cent ; therefore ,the meteorological service i s o f benefit to the mariner .

Ou r percentages Show a verification o f 74 per cent f o r thirty-sixhour forecasts o f rain . We know that the public generally

,and the

agriculturist in particular,consult the probabilities and believe in

them ; therefore , it is obviously advisable to predict rain .

Ou r percentages show a verification o f 84 per cent for temperaturefor thirty-six-h our forecasts

,and during winter the service i s con

t in u a l ly consulted as to cold wa ves , etc . This shows that temperaturepredictions are possible

,and are appreciated by that portion o f the

public more directly interested — R. F. Stup a r t , Pr e dictin g Ofiice r .

Tabl e sh owin g p er cen ta g e of ve rifica t io n of p r edictio n s of r a in , t emp e r a tu r e , a n d

win d ve l o ci ty.

Pr e d ict io n s .

PRECIPITATION.

153 131 137 136 12 2 142

10 8 94 10 3 82 69 10 42 0 16 15 2 5 2 7 17 2 44

159 156 160 151 152 159113 93 10 4 10 5 112 10 2

2 4 3 2 14 2 0 17 2 8 2 83

181 155 155 155 155 166

12 8 94 10 0 98 9 2 110

3 0 2 4 2 3 2 8 32 32 360


Ta bl e s h owin g p e rcen ta g e of ver ifica t io n , e ta— Co n t in u e d.

Pr e d ict io n s .

PRECIPITATION— Co n tin u ed.To t a l fOI t hr e e ye a r sNu m be r 438

Fu l lyve r ifi e d 2 7 2

Pa r t y ve r ifi e d 10 5

To ta l pe rce n t a g e fo r thr e e ye a r sTEM PERATURE.

189 0Nu m be r 156

Fu l lyve r ifi e d 12 3

Pa r t 18

141

Fu l l ve r ifie d . O O O O O O O 9 0

Pa r t y ve r ifi e d . 2 6

Nu m be r . 148

Fu l ly ve r ifie d 10 4Pa r t ly ve r ifi e d . 2 3

To t a l fo r t hr e e ye a r sNu m be r 445Fu l l y ve r ifi e d 3 17Pa r t ly ve r ifi e d 67

To t a l p e rce n t ag e fo r t hr e e ye a r sW IND VELOCITY.

Nu m be r 58

Fu l ly ve r ifi e d 35Pa r t ly 15

Nu m be r 39Fu l ly ve r ifi e d 2 6

Pa r t ly ve r ifi e d 41892Nu m be r 81

Fu l ly ve r ifi e d 50

Pa r t ly ve r ifi e d 18

To t a l fo rbt hr e e ye a r su m be r 178

111

Pa It y ve r ifi e d 37

To t a l pe rce n t a g e fo r t hr e e ye a r s

3832 63

59

11389

94

11

142

1 0 0

3 1

3492 62

46

50

33

7 0

598

46

166

12 0

18

9 0 .

42 6

2 6776

12 1

9 2

15

14512 4I4

151

417

98

7 0

18

84

12

60

3 1

12

2 42

158

4 2

4352 88

7 1

12 2

10 1

11

I4710 8

2 1

142

12 0

I4

41I

32 946

430

2 74

78

119

18

141

10 1

32

12 8

10 2

I9

3882 88

44 1

2 84

74

10 8

7 1

12

146

12 3I8

138

10 1

2 7

392

2 95

57

816710

11488

16

57

40

12

2 52

195

38

Ra in p r ed ictio n s f o r Ju n e .

Ye a r .

Ifie d . Ifie d . Ifie d .

D ays of n o r a in p r edict io n s .

Ye a r .

Nu m be r o f Fu l ly ve I Pa r t ly ve i No t ve Idays . ifie d

O O O O O O O O O O O O O O O O O O O O O O O O

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ifie d .

Fu l ly ve r Pa r t ly ve r No t ve r

ifie

44 2 452 442 42 9 4672 81 3 0 7 2 85 2 73 316 3 .459

7 2 52 73 76 77

0 0 0 0 0 0 0 0 . 0 0 ' 0 0

P e rce n t a g ef u l

l

lfiyd

ve r

e d

887

73 9

Pe rce n t a e

fu l ly a npa ri

t lydve r



weather signs in Scotland at 8 a . m . have begun to give warning,

S torm warnings are n o t sent farther north than Bo dO. The shore

comm unity is satisfied . We have received latterly pecuniary aidfrom the S torth ing in order to undertake special weather studies withthe intention o f establishing local systems of weather warnings withlocal centers in Bergen

,Th r o n d h e im ,

and at the great fi sheries

(Lofoten , fo r example ) under the direction o f local meteorologi sts asdirectors

.The principles according t o which we issue warnings are

hard to analyze— scie n t ific instruction , local experience , im agination .

It i s quite a s much an a r t as a science which one practices — Dr . H .

M o h n,Dir ecto r .

IV .

— RUs s1A : CENTRAL PH Y SICAL OB SERVATORY .

SAINT PETERSBURG .

I send yo u reports Of the observatory f o r the years 1886—1891

,as

wel l as the report concerning the service o f the forecasts of snowdrifts on the railroads . With the permission of the director , Dr . H .

Wild,I add an extract from the report for 1892 . The rules which

we follow for weather predictions are about the same as stated inyour (Dr . Van Bebbe r ’ s ) excellent book , which is used as a text book

in the forecast departm ent .

Since July,1892

,after previous trial s

,the general weather predic

tions have been subj ected to regular tests,as published in the bul

l e t in . The predictions for the districts Of European Russia (comparemap in supplement to Daily Bulletin ) have been separated and veri

fie d f o r the four elem ents Of precipitation , cloudiness , temperature ,and wind . For each o f the elements three grades were distinguished

,

a high,a mean

,and a fair grade . The predictions were considered as

successful when,at the m aj ority o f stations in the given district

,the

predicted phenomena were Observed in the predicted degree . Aspartly successful were taken those predictions for which the phemomena did not show the indicated grade but the next o n e to it . Asunsuccessful predictions were regarded those f o r which the oppositeOf the predictions w a s Observed .

Pr e dict io n . Obse rva t io n . Ve r ifica t io n .

Th ick clo u ds

In the final verification o f the degree o f success o f the predictionthe num ber o f partly verified prediction s i s distributed

,half to the

successful and half to the unsuccessful prediction .


The exact determination of the notation,the sub-division o f the

is t r icts , and the choice o f the grade f o r the characteri stic weathere m a n d

,however , special meteorological researches ; consequently ,

for 1892 is t o be regarded as a first attemptproper criteria .

the verification are given in percentages in the fol

Ver ifica tio n of th e g en er a l we a th e r p r edictio n s f o r s ix m o n th s (Ju ly—D ecem be r ) of 1892 .

Pr e d ict io n s .

Su cce s s fu l . Un s u cce s s fu l .

DISTRICTS Per cen tNo r t hwe s t e r n Ru s s iaWe s t e r n Ru s s iaCe n t r a l Ru s s iaNo r t he a s t e r n Ru s s iaEa s t e r n Ru s s iaSo u t he a s t e r n Ru s s iaSo u thwe s t e r n Ru s s ia

ELEM ENTS

Besides the general weather predictions which are publ ished dailyin the bulletin

,more than one hundred and fifty replies are given

by the department to private inquiries (m ostly from Perm and Nizhnee-Novgorod ) concerning the expected weather . Besides this

,dur

ing three months ( in summer and autumn ) daily weather predictionsare made for Pawlowsk .

From the preceding table it i s evident that the percentages o f thesuccessful predictions f o r the diff erent districts Of European Russiaare dissimilar

,being larger in the east and smaller in the west . This

difference is more marked in the comparison Of the results of predictions fo r single places . The number o f successful predictions forthose stations lying in the east (Perm ,

Nizh n e e -Novgorod,Saratov

,

e tc. ) i s 80 to 81 per cent, while in Pawlowsk it only reaches 63 to 64per cent .In the supplement we reproduce three letters 1 from Mr . Pa n a ew and

On e from Mr . Ba t ju schko w ,in which the writers express their thanks

for the predictions sent them . From these letters it appears that o u rpredictions are practically useful this resul t i s for us the more sati sfactory-since it i s not based on single chance forecasts

,but upon a

whole systematic series of predictions,o f which naturally a certain

percentage m ay be bad . It should here be remarked that both theabove-mentioned gentlemen were subscribers to the telegraphic

1Th e s e le t t e r s we r e n o t r e ce ive d by m e .—EDITOR.


weather predictions and in consequence paid f o r each o n e a tax of 50kopecks besides the usual tax . In both letters the wish i s expressedthat further progress may be m ade .

Sto rm wa r n in g s f o r th e Ba l tic a n d f o r th e Bl a cka n d Azof Se a s in 1892 .

!Th e g e n e ra l r e s u l t s , e xp r e s s e d in p e rce n t a g e s , a r e fo r a l l d is t r ict s .]

Wa r n in g s . Ba l t ic. Black.

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O

The percentage o f storms which were not predicted,whose force

exceeded that indicated by o n e ball,was

,fo r the Baltic Sea

,10 per

cent ( 1891, 13 per cent ) for the Black Sea , 19 per cent (1891, 2 3 percent ) . If we com bine the verified and partly verified warnings theresult o f the successful warnings f o r 1892 i s as fol lows : Baltic Sea

,

85 per cent (1891, 76 per cent ) ; Black Sea, 76 per cent ( 1891, 69 percent ) .— Oh . Ryka tch ew .

V.

— THE WEATHER SERVICE IN AUSTRIA .

VIENNA .

In Austria the issue of a telegraphic weather report wa s begunJanuary 1

,1877 . This service is carried o n by th ree officials of the

Section for Weather Telegraphy,

” whose office since the year 1879i s in the center o f the city in the building of the Imperial Academyo f Science . The telegraphic material has increased notably Since1877 . At present there are received daily weather reports fromthirty-six Austro-Hungarian and from Sixty-six foreign stations .These are dispatched in twelve com bination telegrams to 158 domesticand foreign station s . The monthly subscription price for the weatherreport i s 1 fl.

, 50 kr . The weather report goes t o the printer not laterthan 3 O ’clock , and is sent out regularly before 5 O ’clock . At thepresent tim e there are ninety-nine subscribers to the weather report ;the number of free , exchange , and Obligatory Oopies amounts to eighty .

Besides the printed weather report,there have been sent since 1877

daily prediction telegrams which go chiefly to the landed proprietors,

health resorts , and the larger provincial papers . Since the year 1884the prediction telegrams have been in cipher

,whereby the su bscr ip

tion price h a s been reduced 50 per cent to it s present price of 5 fl.per

month . The telegram s are issued general ly between 1 and p.m .

The success o f the predictions is,o n th e whole

,satisfactory

,aver

aging 85 per cent . The subscribers to the prediction telegramshave varied since 1878 from fifty to eighty three . AS regards therelation o f the public and their interest In the weather to the prediction service , the j ournalists hold an important place . In Vienna all


the great newspapers are forced by their readers and correspondents,

both in the morning and evening editions,to print the telegraphic

weather reports with th e predictions . When , from lack of space orother reasons , the weather report does not appear , complaints comeat once to the publishers . The confidence in the weather prediction smanifests itself most strongly in the fact that n o t only the Viennese

,

but a l so the great proprietors in the provinces take pains to procurethe predictions — Dr . J. H a n n

,Dir e cto r Ce n tr a l Bu r e a u fo r M e te o r o l ogy

a n d Te r r e s tr i a l M a gn e tism .

VI.— HUNGARIAN METEOROLOGICAL AND MAGNETIC BUREAU

,

BUDAPEST.

Weather predictions were issued in 1881 by direction Of the Minister of Agriculture , at that time Baron K e m sny, by Dr . Sze n tgyOrgyi

Weisz , wh o , however , wa s independent o f the M eteorological Institute . The dissemination o f the forecasts was accomplished throughthe newspapers ; the telegraphic dispatches were insufficient andafter eight years ’ continuance this arrangement was given up .

In the year 1888 the task o f m aking weather predictions was e utrusted to the Meteorological Institute

,but stil l

,this service

,in the

to o narrow limits to which it was confined by the previo us organization

,could get no firm hold .

A noteworthy advance in this field can be recorded when I tookthe direction of the Central Bureau . On June 15

,189 1

,our first syn

optic charts appeared,whose issue was facilitated by a subvention

from the Minister of Agriculture . The forecasts were publiclyposted in Budapest and their circulation was confined to the dailynewspapers . On account o f the great importance of the predictionsin an agricultural state like Hungary

,a quicker and more general

dissemination o f the forecasts was desirable . The present Ministero f Agriculture

,Count Andreas Be th l e n

,manifests a lively interest

in the m atter and,acting o n my suggestion

,has succeeded in enl ist

ing the Ministry o f Commerce . The creation o f such a method Of

forecast dissem ination as exists in Hungary,according to my knowl

edge,has n o t come about in any other European country .

There was inaugurated,primarily

,a general official method o f dis

tributing the predictions in the interest o f the farming community .

Accordingly,o n August 1

,1892

,the first forecast dispatch ( in cipher )

was i ssued as an appendix to the official “ circular dispatch,

” which

contains the quotations of the Budapest merchandise and grain exchanges

,and by way o f experiment this forecast dispatch wa s sent to

one hundred and thirty telegraph Oflice s for official distribution tothe public . Each o f these telegraph operators received

,also

,a suit

able bulletin board o u which the data and the announcements formin g the forecasts were to be hung . There belong with each board

4


twelve small tablets for the d ifie r e n t months , thirty-o n e tablets

(bearing the numbers 1—31) f o r the days , and thirty-one f o r the pre

dictions. On the last are , together with the cipher , al so the a n

n o u n ce m e n t s belonging to each . The telegraph Operator is thus r e

l ie ve d o f all trouble o f translation ; hi s task consists only in takingfrom a box that tablet which bears the cipher o f the forecast dis

patch just received and in hanging it up .

This method o f spreading the information deserves to be mentionedbecause it i s a free o n e and is solely for the benefit o f the agriculturalcommunity . Much public interest is everywhere manifested , and theagricultural societies in many counties have applied to the Government f o r an increase in the num ber o f telegraph Operators wh o havebeen entrusted with the receipt and the publication o f the forecast

dispatches,s o that at present one hundred and thirty o flficia l s have

been appointed,and after May 1

,1893

,two hundred and sixty tele

graph o ffice s will serve to dissem inate the predictions . Further , f o rrem ote places and for Pu szt e n ( inns in o u t-o f-the-way localities ) anoptical method of signaling i s proj ected

,so that by displaying dif

f e r e n t colored flags,baskets

,and cones (m ade Of Osier ) the forecasts

wil l be m ade known . On my estate in O’Gya l l a and on that of theMinister in Be t h l e n this scheme is already in operation .

The forecasts relate,exclusively

,to the c onditions wh ich are o f

interest t o farmers , that i s to say , especially to precipitation andtemperature . and

,incidentally

,to cloudiness . Storm warnings are

not issued . The forecasts apply for the following twenty-four hours,

but in consequence o f the importance which a forecast o f the weatherf o r two days ahead possesses f o r agriculture

,in certain cases o f

weather stability o n e i s undertaken f o r two days with the announcement

,

“ weather conditions persistent (w ) , as well as,when the

weather condition adm its,a change for the second day is indicated

by “ later precipitation o r“ later clearing The varia

bil ity of the clim atic conditions requires that from the extent o f

Hungary,in special cases

,the rain probabil ity should only relate to

certain regions,which i s indicated by the announcements “ p r e cip i

t a t io n in the west “ precipitation in the south e tc. The

forecast dispatches are sent until 3 O ’clock each afternoon from theInstitute to the telegraphic centers .At present the Meteorological Institute i s hoping f o r a r e o r g a n

iza t io n Which wil l create a Special section f o r Weather Telegraphyand Forecasts .” It will only then be possible to study the weatherconditions in detai l from the point o f view o f forecasts

,and to pre

pare the needed statistical data .

Regarding the success o f the forecast dispatches,I can only now

cite the computed resul ts which relate to the quarter August-October ,1892 . The figures are obtained by a comparison o f the forecasts


with the records,a t the hours 7 , 2 , and 9 , o f twelve uniformly-dis

tributed meteorolog ical observing stations . These forecasts wereverified in the above-mentioned quarter

,f o r precipitation

, te m

p e r a tu r e , and cloudiness,

per cent . With the establ ishm ent o f a special section it is the intention t o develop furtherthe whole forecasting service

,o f which I may speak more at length

later .— Dr . N. vo n K o nko ly, Dir e ct o r .

VII.

— THE WEATHER SERVICE IN THE NETHERLAND S .

UTRECHT.

In s t a l l a tio n — At Utrecht , after the arrival of the dispatches fromthe Netherlands

,about 10 to 11 a . m .

,the report i s placarded and

publish ed in the m orning edition of the local newspapers whichappear about 1 p . m . Beneath the report is entered the greatestbarometric deviation which is (graphically ) represented on thea ér okl in o sko p (cf .

,In vo e r in g e n Ve rkl a r in g , translated by Dr . Je l in ek

with the title Th e In the afternoon,about 3 p . m .

,

the weather chart,manifolded by the hektograph process

,i s i ssued ,

beneath which i s a summ ary of the greatest barom etric deviation at8 a . m . and p . m .

,and a weather prediction for the twenty-four

ho urs following the observation . The weather charts published atUtrecht are sent elsewhere o n payment o f postage .

Re su l ts — The results o f the predictions are s e l d o m e xp r e s s e d by

figures . Results o f the storm warnings,which are based on the dif

f e r e n ce s o f the barom etric deviations,are published annuallyin the

Niede r lan disch e n M e te o r o l ogisch e n An n a l e n . The results of the two lastyears are that 60 per cent of the storm warnings were not verified .

Pr e se n ta ti o n to th e p u bl ic.

— The organization of the weather service ,which is essential ly represented by the Te l eg r ap h isch We e rbe r ich t t e ndi e n s te va n de n La n dbo u w and which dates from the close of 1882 , wasfor several years unfavorably regarded by the public

,j ust as the

storm warnings were for the fifteen o r twenty years preceding . There

followed a period,which still continues

,when the publ ic

,took little

or no notice of the weather reports . Only about 1889 w a s thereShown a greater interest by the public which manifested itself in t h eappearance o f a daily weather chart in tw o newspapers , the first andup t o now the only ones in the Netherlands to publish these charts .Th e deg r e e of a ccu r a cy in p r e dictio n s of t emp e r a tu r e , sn ow ,

r a in,o r

win d,two o r th r e e d a ys in a dva n ce . Wh a t p r in cip l e s a r e a dop te d in su ch

p r edict io n s? Utrecht issues a definite wind prediction , which i sfounded o n the difference of the barom etric deviation , and a more orless detailed weather prediction for the ensuing twenty-four hours .A te mperature and wind prediction i s seldom made f o r the second

twenty-four hours,never for the third

,and never

,as regards rain , for

the second period . These predictions , like those for the first twenty


four hours,are based on the determination o f the current gradient

(not barometric gradient ) o f the air in a vertical and horizontal d ir e ction

,o n the modifications which they may undergo during the day ,

and o n the physical changes which may result.It is desired chiefly to predict those weather elements which are

most relate d t o the duration o f the prediction , such as temperatureand rain

,f o r example .

Rain prediction i s the most important f o r the public , and is al sothe most uncertain ; even the thunderstorm prediction , which haseven more value and also is the surest, has often only a local ve r ification — Abr idg ed tr a n s l a ti o n by M r . En g e l e n bn rg .

AM STERDAM .

The weather service in Amsterdam furnishes information about theweather not only to the public in Amste rdam , but al so to the S ta a ts

a n ze ig e r and the Ha r l em m e r Ze i tu n g . These papers receive as full a

summary a s those in Amsterdam . The report i s telegraphed to the

S t a a tsa n ze ig e r each noon , while the H a r l em m e r Ze i tu n g sends for it .

The latter paper wishes no tabular data , but o n ly‘

a summary . Be

sides,there i s sent to other places in the Netherlands a weather sum

mary,with a wind prediction added to the Te l eg r ap h isch We e rbe r ich t te n

die n s te va n de n La n dbo u w. These reports are publi shed by harbormasters

,provincial newspapers

, e tc. Finally , in many places in Am

s t e rd am ,and also in Harlem and at the office of the H a r l em m e r Z e i

tu n g , h ekt o g r a ph ic weather reports are posted .

I believe the value o f.the Te l eg r a p h isch We e rbe r ich t te n di e n s te va n

d e n La n dbo u w to be very small . A mere statement that a d e pr e s

sion l ies northwest and far o ff ” o r“ a hig h pressure area northeast

and in the neighborhood ” i s incomprehensible and of n o value tothe public .

The abstracted tables containing data for some domestic and foreigh stations may

,however

,be useful

,but

,in general

,I do not think

that the ordinary newspaper readers care much for the report . It i sotherwise with seaports

,where great differences in the barometric devi

a t io n s may serve as warnings to seamen . This branch o f the serviceshould

,according to my ideas

,be better supported . The a e r okl i

n o sko p ,which i s set up in some places

,does not fulfi l l the want .

To the ports (especially Delfzyl , Nieuwe D iep , Ymuiden , Zandvoort ,Scheveningen , Maassluis , Vlaardingen , Hellevoetsluis , Brouwershaven ,Vlissingen ) storm signals , analogous to those o f the De u tsch e S e ewa r t eo r those long employed o n the English coast

,should be supplied .

The a er okl in o sko p can only be read at a little distance , and demandsto o much im agination on the part o f the sim ple fishermen and seamen .

The reports which the Am ste rda m e r Ze i tu n g , the H a r l em m e r Ze i tu n g ,


K


a be n den Wi t t e r n n g sku n de Of Dr . W . J . Van Bebber , and more recently

the We t te rvo rh e r sa g e o f the same author , are used as guides .If one considers that in the Netherlands only once in twenty-four

hours a review o f the situation over Europe can be had , and in Ger

many,for exam ple

,this i s made three times a day, it appears t o

me that the result o f the predictions in Amsterdam i s not wholly

unfavorable . It is very m uch to be desired , that ( for example at 1p

.m

. ) som e dispatches should be received directly from England ,France

,and Germany ( for exam ple six o r eight ) and immediately

charted . If that could be done , the value of the prediction wouldthereby be g r e a t ly increased . Nevertheless , I repeat“ n o dispatchshould be more than o n e hour late .

In brief,m y Opinion is , that apart from som e progress which may

stil l be made in m eteorology as a science , it is very desirable , as may

n o w be done , by an acceleration of the dispatches and a proper means

o f distributing and publishing the reports , that th e public shouldderive greater advantage from the reports than is at present th e

case — Abr idg e d tr a n s l a ti o n byL. Ro o se n bn rg .

ROTTERDAM .

The weather service in Rotterdam is confined to the com munication

o f weather reports to the papers published in Rotterdam and . the d is

t r ibu t io n o f this report within the parish . This is accom plished by

m ere mechanical working up of the dispatches received ; it seem s tom e

,therefore

,that a central bureau can hardly be spoken o f . Predic

tions are n o t m ade here,unless t h e paragraph which follows the

barom etric deviations from the normal be regarded as such , f o r exam ple

,the paragraph “ indicate s a wind

,

”o r in the h ekt o g r a ph ic

weather reports “ according to,the Buys-Ballot l aw

,there should be

a wind .

”

I have left these expressions because it is difficul t to give a ful lexplanation each day

,but I do not consider them as forecasts . It i s

only the statement o f two phenom ena between which there exists acertain relation without our being able to s ay in general that onei s the cause and the other the consequence .

The value o f weather predictions published for the public appears

to me doubtful,especially because to the forecasts wh ich are u n su c

ce s sfu l much m ore attention is paid than to the others .Fr o m

m

p r iva t e telegrams , especially in com m ercial circles,it has

often been shown to m e that much interest attaches to weatherreports which m ake known the true situation

,and that it i s to be

regretted that o u r reports do not cover a larger portion o f Europe .

The method of procedure here is as follows : At noon,the Sch e ep

vo o r t newspaper sends for a l ist o f the reports already received . This

li st is posted outside a window o f the newspaper office in the neighbor


hood o f the Exchange . At certain seaso ns,dependent o n the sugar

crop,a copy of this is posted in the Exchange . Between 2 and 3

o ’clock a com plete weather chart is h ekto g r a ph e d , and the Sch e epvo o r t distributes sixteen copies throughout the city , which are accessible to the public at large

,and are al so pos ted in localities where

many interested persons assemble,as for example

,exchanges

,com

m e rcia l clubs ; societies , etc .,fo r the benefit o f navigation

,navigation

schools,trades unions

,etc . Further

,about 6 p . m .

,there appears in

the Sch e ep vo o r t a sma l l m a p ,printed by the Rung system

,and about

5 o ’clock a full list with a general review is published in the Ne u eRo t te r da m e r Ze i tu n g . This paper

,also

,has received f o r some time

past a graphic representation,in a form devi sed by the editors

,o f the

highest and lowest barom eter and the temperature during the fivepreceding days .

The great expense wh ich the Sch e epvo o r t incurs , as wel l as the largespace which the Ne n e Ro t t e r da m e r Ze i tu n g , from interested motives ,devotes to this m atter

,and the questions which occasional ly reach

m e through the editors o f these papers,im press me with the fact that

they count am ong their readers m any who are interested in this subj ech — Abr idg e d t r a n s l a tio n byArke n bo u t Sch okke r .

VIII .

— LONDON METEOROLOGICAL OFFICE.

Dr . Van Bebbe r ’ s letter asks various questions : 1) As to thedegree of accuracy in our forecasts of tem perature

,rainfall

,wind, etc . ,

f o r o n e,two

,o r three days in advance . 2 ) As to the principles on

which the forecasts depend,and the character of the weather we are

able to predict . (3 ) As to the advisabil ity of predicting rain , andthe extent to which we should predict tem perature o r other changesviewed from the standpoint o f “ what it i s possible to predict with afair degree o f success and what it i s that the public cares t o know .

”

The following replies are drawn up,not exactly in the order in

which the queries are put,but in such an order as enables me to r e

ply more clearly and briefly than I could otherwise do .

Forecasts ” or “ predictions ” are issued by this office , as followsThe first are prepared at a . m .

,and issued at 11 a . m . (Sun

days,Good Fridays

,and Christmas days excepted ) , and are mainly

dependent on the observations taken at 8 a . m . daily ( see copy of theDaily Weather Report ) and relate to the weather to be expected durin g the twenty-four hours ending at noon o n the following d ay. They

a r e intended chiefly f o r publication in the Daily Weather Report ,in the afternoon newspapers

,and f o r exhibition at certain positions

in the city and west end o f London,including most o f the clubs .

The second are prepared at p . m . (Sundays , Good Fridays ,and Christmas days excepted ) from 2 p . m . observations , and madeat a lim ited num ber of stations

,as supplementary to the 8 a . m . ob


s e rva t io n s . They relate to the weather o f the ensuing civil day .

They are always posted at the door of the office for inspection by the‘ public and during the hay and wheat harvests are telegraphed g r at u it o u s l y to a selected number (about twenty-eight ) o f agriculturistswh o make their contents known a s widely as possible and keep a

careful check o n their accuracy .

The third are prepared at p . m .,daily , and are issued at

p . m . These al so relate to the weather o f the ensuing day,and are

dependent o n observations made at 6 p . m .,a s supplementary to those

m ade at 8 a . m . and 2 p . m . They are intended m ainly for publica

tion in the morning newspapers o f the following day .

They are,therefore

,all o f them

,f o r a period o f rather more than

twenty-four hours in advance of the tim e of issue,and are util ized in

answering inquiries by the public as t o comin g weather .Special “warnings ” a s to the advance o f storm s are sent by telegraph

to the coasts threatened,whenever the indications are believed t o be

o f a stormy character . These may be sent at any hour betweena . m . and 8 p . m . , and are made known by the hoisting o f a cone

(point up for northerly, point down for southerly gales ) at the portsto which they are sent .

In the forecasts the wind (direction and force ) and the weatherare predicted separately , in a som ewhat general m anner, as the dis

t rict s for which they are prepared cover a considerable area . In

the weather portion any kind o f weather is included,if it i s likely t o

be a prominent feature , but at present hardly any attem pt has beenmade to estim ate the intensity of coming rain— the local variationsin the character o f the country and the variations in intensity o f

thundershowers being t o o abrupt to m ake m inute detail desirable .

Such expressions , however , as rain at times— heavy locally ” areemployed when deemed necessary .

With regard to changes o f tem perature,two distinct classes are

kept in view ,1) those o f a general and (relatively ) of a perm anent

character affecting the mean temperature o f the approaching period,

and referred to in such expressions as “ colder,

” “ much colder,

”

“ warm er ,” “

m uch warm er ,” etc .

, and (2 ) those of a diurnal character which , in such periods as that recently experienced over o u r

i slands , are very large , and are referred to in sen tences such as “ cold

at night , warm during day .

” No attem pt has been m ade hitherto t ocheck the accuracy of such forecasts

,except as forming part of the

weather portion of the predictions,but it i s believed that they are as

good as those f o r any o f the other features included in the forecasts .With regard to the success which has attended the issue o f the

forecasts , reference may be made t o many distinct sources : (1) tothe Official checking of the p . m . issue

,carrie d on in thi s office

from the information received daily by wire . The results of this


checking wil l be found o n pp . 11 and 63 o f the Report of the Meteorological Council for year ending March

,1892

,and are very fairly

satisfactory . 2 ) To a similar checking o f the p . m . forecasts,

based o n inform ation suppl ied by the recipients o f the forecasts ( seesame report pp . 12—13 ) and the favorable opinion expressed by themin their letters to the council ; also , to the fact that the samegentlemen are glad

,year after year

,to receive the forecasts

,to

make them known,and to keep the record necessary to check them .

(3 ) That among those who m ake inquiry privately,the same

nam es appear regularly in successive years,whenever the informa

tion i s required,although a fee i s levied for it

,and the costs o f trans

mission by wire (when necessary ) are paid by the applicant . (4 ) Tothe facts that ( a ) the National Lifeboat Institution applied recentlyto have the forecasts telegraphed daily to the officers in charge attheir various stations

,a s a guide to them in their duties (a request

which wa s reluctantly declined only because the cost of nearlyper annum wa s more than the Meteorological Council were

able to meet ) , -and (b) that“ the Agricultural Department i s even

now endeavoring to make arrangements f o r telegraphing the p .

m . forecasts daily to all agricultural districts during harvest time .

(5 ) That the authorities at Her Maj esty’ s Dockyard

,Devonport

,

now have the 11 a . m . forecast telegraphed to them every day for

guidance in sending the sm aller vessel s to sea ; and that Her Maj estynever puts to s e a without having the latest forecast transm itted toher by wire . (6 ) That the newspapers in all parts o f the kingdom

have not only published the forecasts regularly for many years , but inmost cases pay a considerable su m for cost o f telegraphy

,when the

offices are too far distant for them to be delivered by hand,and that

Th e Tim es paid £50 0 per annum f o r the exclusive use of the 6 p . m .

forecasts,and subsequently the three leading London papers paid

£90 0 per annum between them f o r the u s e of the same forecast untilthe Government made a grant to defray the cost and make the in f o rmation free for all papers .With regard to what the public wish to know— they would u n

doubtedly like (a ) to have the forecasts i ssued f o r a longer period inadvance

,probably fo r seasons

,and (b) that m ore m inute detail

should be Observed in localizing the regions likely t o be affected by

rain,and the intensity o f the com ing fall . At present

,however

,it

has not been found possible to gratify these Wishes .This brings us to the consideration o f the principles adopted in pre

paring the forecast,and the line o f work o r study which promises to

increase their accuracy .

Wi th r e g a r d t o th e p r in cip l e s a d op t e d— They depend mainly upon a

recognition o f the well-known characteristics o f cyclonic and anti

cyclonic system s,prim ary

,secondary

,or V-shaped

,and upon the


ind ications afforded by the three daily Observations a s t o their m ove

m ents and the questions o f their tendency to increase o r decrease in

area o r intensity . The general distribution of pressure (al so whetherfa vorable f o r a continuous prevalence o f cold o r warm ,

o f dry o r wet,

currents o f air ) , the effects of such currents when coming from o ff

the s e a,o r vice ve r sa

,the relatively rap id motion o f a ir o n coasts

when compared W i th that over land , and the variations produced bythe sea sons o n such phenomena a r e ‘ a l l careful ly thought o u t before

preparing the forecasts or issuing warnings , besides the questionwhether the disturbances are o f a “ thunderstorm

”o r other characte r .

Wi th r e g a rd to th e l in e of wo rko r s tu dy m o s t l ike ly t o in cr e a se th e

a ccu r a cy of th e p r e dicti o n s — It appears probable that som e rearrange

ment of the districts f o r wh ich they are prepared,the separation o f

coast from inland parts of the countries and of the west from theeast portions of the Irish districts

,is desirable . It is probable

,also

,

that a better knowledge o f the upper currents of the air,as shown

by h igh clouds,is necessary

,and that a more careful study of the

distribution o f rainfall under varying types of pressure distributionand at diff erent seasons o f the year (distinguish ing between thevarious classes o f rains ) m ay im prove the forecasts mate rial ly . Thatevery effort to bring about such an improvem ent is desirable mus tbe patent to all— sailors

,agriculturalists

,and dwel lers in towns being

all interested in the results .

At present i t h a s be en found impossible to institute seasonal forecasts with any reasonable hope o f success — Fr e de r ic Ga s t e r

,Ch i ef of

Fo r eca s t Divisi o n .

IX .—BERLIN WEATHER BUREAU.

l

The telegraphic reports are received from the De u tsch e S e ewa r te in

Ham burg, and during the past year another telegram -h a s been

received from the Royal Bavarian Central Bureau,containing the

reports from the four stations,Zu rich

,Genoa

,Lugano

,and Bozen

,

wh ich have proved very useful .We u se the telegraphic material f o r the purpose o f weather fore

ca sting in the construction o f isobars and isotherm s,and for about a

year I have worked with an assistant ( formerly Dr . Siih r in g , now Mr .

Basilius ) in plotting lines of equal pressure and temperature variation in twenty-four hours . The m ethod o f verification o f '

o u r fore

ca s ts has undergone a great change since it s com mencement in thespring o f 1884 . Together with the system atic verification o f the fore

cast , resolved into the elem ents,I have verified also

,a s a whole

,each

forecast a ccordin g to the S e ewa r te method ( I , entirely successful , up t o

1Th is is a p r iva t e bu sin e s s e n t e rpr ise w ith it s h e a dqu a r t e r s a t t h e Ag r icu lt u r a l HighSch o o l.— ED ITOR.


V,entirely wrong ) , in which I have endeavored t o take account o f the

views of the local public as m uch as possible,and

,therefore

,to give

proportionately more weight to the predictions concerning rain andtemperature than to those relating to cloudiness and wind conditions .From these rules the fol lowing percentages o f success were obtaine dfor the years 1885

II. III. IV. V. Su cce ss .

Win t e r (De c. 18. 2 48. 1 2 9 . 6 4 . 1 0 . O 81. 1

Spr ing 2 1. 6 47. 6 2 6. 4 4 . 4 O. O 82 . 4

Su m m e r 15. 3 49 . 1 2 9 . 7 5. 7 0 . 2 79 . 2

Au t u m n 18. 2 44 . 9 32 . 6 4 . 3 0 . O 79. 4

Ye a r 18. 3 47. 4 2 9 . 6 4 . 7 0 . O 80 . 5

The most favorable month was April,with per cent ; the most

unfavorable, October , with a n d next , July , with I would

here remark that the forecasts are i ssued between and p . m . ,

and apply for the whole of the following calendar day . In single

years an increase in the figures denoting success is not evident,and

indeed the first years Show the greatest success,viz .

,1885

,per

cent,and 1886

,per cent . This arises

,however

,from the fact

that we now attempt to give to the forecasts a more definite meaning,

especially t o emphasize the expected weather changes,and perhaps

,

also , o u r o wn j udgment about the m istakes has becom e harsher . The

continually growing interest of the public i s perhaps best shown bythe results outside o f the Bureau

,and I quote below the number o f

newspapers t o which,at the com mencem ent of each year

,we were

furnishing weather charts and forecasts :

Ja n u a ry 1. 1885 . 1886 . 1887 . 1888. 1889 . 1890 . 1891. 1892 . 1893 .

Fo r e ca st s 4 6 6 7 7 16 17 16 17

We a th e r ch a r ts 2 4 4 4 6 6 7 7

The great increase from 1889 to 1890 i s partial ly explained by the

fact that Mr . O. Jesse , wh o up to that tim e had made forecasts forfour local papers

,gave th is up in 1889 . For the representation o f

weather charts an attempt was m ade by us in 1885 to stam p them o n

type metal,and in the following year the schem e had been so far

perfected that it was applicable to stereotyped papers,am ong which

the Be r l in e r Ta g ebl a t t , which had already for m any years printed

weather charts by an etching process adopted ours ; this underwenta further improvement in 1889 by the substitution o f black fo r whitefigures and sym bols . Bes ides this newspaper block

,which i s ready

at p . m . ,we have delivered since May

,1892

,dai ly

,except Sun

day,at o ’clock

,h ekt o g r a p h ic weather charts f o r the local

Urania colum ns .”

I will now,after this long and detailed account of the working o f o u r

weather bureau,state briefly the experience we have obtained in o u r

efforts to improve the forecast methods . It appears to m e that the

eff orts toward longer forecasts,even if they are more general , prom ise


better success than the more exact ones f o r the next day . In additionto the thorough researches o f Va n Bebber , upon the typical tracks

o f minima , it appears to me that a m ore detailed investigation o f

Te is s e r e n c de Bort ’ s weather types would be very profitable , and I

believe that,to the knowledge o f their frequency , duration , change ,

sequence,etc .

,my weather charts might contribute som ewhat .

As to the wish o f the publ ic respecting the forecasts f o r the ensuingday

,it has been my experience that no little importance i s attached

to the explicitness o f the forecast . So,f o r exam ple , it wil l not suf

fice to give the announcement “ changeable weather .

” Here,as every

where,the greatest attention m ust be given to the precipitation fore

cast,and it should be the aim . gradually t o separate more and more

the local from the general rains,and in summer the simple “ tendency

to thunderstorms ” might be replaced by a greater o r less “ probabilityo f a thunderstorm .

” Perhaps,a systematic investigation of the vari

ation of the absolute humidity,f o r which I have for several years

been collecting data , might contribute som ething to this question ,although , naturally there are other more important researches . Ibelieve that a more exact knowledge o f the recent precipitation dist r ibu t io n is very necessary , f o r wh ich reason I would urge that ino u r telegraphic dispatches the tenths o f degrees of temperature bereplaced in summer by the amount o f precipitation

,and in winter

by the depth o f snow . But the subj ect i s too large a one to beexhausted in o n e letter

,so that I fear I have already entered too

much in to details — Dr . E. Le ss , Dir ecto r .

X .-THE FORECAST SERVICE IN SW ITZERLAND .

The creation of a daily weather forecast,based upon synoptic

meteorological data, was first undertaken in Switzerland by theundersigned , in the summer o f 1878, in a private way. In the yearfollowing it wa s done o flicia l l y by direction of the Swiss Government and the i ssue of a daily weather bulletin was made o n e of theduties o f the Swiss Central Meteorological Institute

,established in

May,1881.

As Switzerland is divided into several (3—4 ) d i stricts whose climatic conditions are essentially different from o n e another

,it w a s

originally planned that the forecasts Should be m ade at severalpoints , based upon a synoptic summ ary telegraphed t o each from thecentral office . But the observatory at Berne has been the only o n e

to undertake the issue o f special forecasts for the surrounding territory ; so that, as a m atter o f fact , the forecasts o f t h e central instituteat Zurich have continued to be the only generally di stributed progn o s t ica t io n s . But in view of the peculiar geographical position o f

southwestern Switzerland (o n account o f the influence of the lower



much to know whether there will be a slight rise or fall in te mperature ( leaving o u t entirely , as above indicated , wind direction andl ight wind ) , as to be informed o f the general character of theweather

.The absolute state o f the tem perature is only in a few cases

( in spring ) o f special in terest, and the same may be said as regardsthe am ount of rainfall (when there i s danger o f floods ) .In most cases only the general character of the weather i s under

consideration,and con sequently the greatest importance i s attached

to its correct determination . Sudden changes , especially , demandthe most attentive consideration . The s o -called local influences p e cu

l iar to o u r mountain country are im portant factors in this connection

. Of very great im portance is the distribution o f the atmos

ph e r ic pressure o n both sides o f the Alps . It depends upon this dis

t r ibu t io n whether o r not the influence o f a depression moving from

north to south wil l extend to the foot o f the mountains . There needbe n o well-d efin e d wind movement (FOh n ) in this case , at least n o tin the lower regions . The s o -called FOh n -eff ects (otherwise dammin g up at the m ou ntains and favoring precipitation ) make themselves felt even with a comparatively small barometric gradient , andmay delay f o r two o r three days , o r even entirely prevent, the formation of clouds and precipitation

,while at no great distance from our

frontier the weather undergoes a radical change .

The study of the influence exerci sed by the Alpine mountain chainas a climatic factor (which may, perhaps , be al so felt in dynamicmeteorology ) is o f the utmost im portance f o r the improvem ent of theforecasts in o u r country

,and this study requires the m ost careful

fostering o f the meteorology of the upper regions through the establ ish m e n t o f good Observations at mountain station s — R. Bi l lwi l l e r ,Dir ecto r .

XI.

— THE UNITED STATES WEATHER BUREAU .

The forecasts and warnings o f the United S tates Weather Bureauare based upon a study o f reports o f observations taken daily at 8a . m . and 8 p . m .

, seventy-fif t h meridian tim e , at o n e hundred and

twenty-four regular reporting stations in the United States and nineteen points in Canada . These reports are promptly telegraphed in

cipher to the Central Office at Washington and t o the m ore importantWeather Bureau stations , and al so transmitted by telegraph to theCanad ian Central Office at Toronto . During the West India cycloneseason provision is made for timely reports by telegraph o f disturba n ce s noted in that region .

Instructions covering the m aking o f observations and fi l ing o f r eports a t the telegraph offices allow o f no deviation from prescribed

methods n o r departure from fixed rules . A reference to these m ethodsand rules , and a brief statement of the processes involved from the


making o f the observations at the several stations to the issue Of theforecasts and warnings at the Central Office

,will probably best i llus

trate the system o f the Bureau .

Prom ptly to the hour and m inute the work o f observation is begunat each station

,and simultaneously the sm all army o f observers per

forms the several operations pertaining thereto . Within a specified

period the enciphered reports are filed at the telegraph offices andplaced upon circuits devoted exclusively to their transm ission . At

a . m . and p . m .

,daily

,the work of deciphering the reports

and charting the data is begun by a force of trained experts atWash ington . The average tim e required f o r this work is about o n ehour . Upon the com pletion o f the charts the Forecast Official dictatesa statement of the general and special m eteorological features presented by the reports

,prepares the forecasts f o r the various districts

,

and issues such signal orders as the conditions may require . The

dictation covering the synopses,forecasts

,and warnings is s e t in type

and also telegraphed as it progresses,and at the expiration of the

thirty to forty-five minutes required f o r the perform ance of this workthe utterances o f the Forecast Official have been filed for transmis

sion to all points in the United States reached by electric telegraph .

The press associations furnish the daily press with the regularforecasts and warnings

,and also transmit special statements o r bul

l e t in s i ssued in anticipation of unusual o r alarming m eteorologicalconditions . In addition t o dispatches transmitted by the news associa t io n s , weather and temperature , cold wave , and frost messages aretelegraphed at Government expense t o specially appointed displaymen and selected points

,exclusive o f regular observers and stations

of the Weather Bureau , as follows :

Displ aym e n o f w e a th e r a n d t e m pe r a tu r e sig n a lsDispl aym e n o f co ld-w ave sig n a lsDispl aym e n o f fr o s t s ig n a ls

To t a l pa id m e ssa g e s .

In addition t o the above , messages f o r public display are telegraphed to railroad stations ; messages are telegraphed or tele

phoned to 62 0 places ; forecasts are sent by mai l to points’

; and

are delivered by cooperating railroad train services to stations .The total number of places to which the forecasts o r warnings aresent i s As before stated

,this num ber represents only regularly

authorized d i splay stations , and does not include thousands of persons and places furnished by the various local Weather Bureau offices

throughout the country . In addition to the above , and exclusive o fregular stations o f the Weather Bureau , Signals giving warning o f

dangerous gales are displayed at 12 1 points o n the sea coasts and theGreat Lakes .


While the comparative degree o f accuracy o f the forecasts for defined periods is shown by the percentage Of verification , the bestproof o f their value to the public i s the increasing demand for thepredictions . Their distribution i s n o w l im ited by , and coextensivewith

,the scope of the electric telegraph and telephone .

The regular forecasts o f the Weather Bureau are i ssued from theCentral Office at Washington by o r before 11 a . m . and 11 p . m .

,

seventy-fif th meridian time , daily . The morning forecast i s madefor a period Of thirty-s ix hours

,and the night forecast f o r a period

o f twenty-four hours . In the discretion o f the Forecast Official forecasts are made for periods o f forty-eight hours . The reg u lar , and

what are termed twenty-four and thirty-s ix hour forecasts , specifythe character o f the weather , such as general o r local and heavy orlight rain o r snow ,

fair or clear weather,higher o r lower te mpera

ture,including terms indicating the am ount o f the anticipated rise

o r fall in temperature , and the force , direction , and shifts of thewind for each S tate o r part Of State east of the Rocky Mountains .The remaining States and Territories

,with the exception o f New

Mexico and Wyoming,are covered by forecasts issued at San Fran

cisco,Cal .

,and Portland

, Oreg . The morning forecasts are o f

special value to outlying o r country districts , a s the messages givingforecasts f o r the following day can be sent t o d ispl aym e n and pointsreferred to ,

and the-

signal s and bulletins displayed w ithout delay .

These forecasts al so appear in all o f the evening papers o f the country . When the morning reports indicate unusual o r dangerous met e o r o l o g ica l conditions , special te legraphic reports are cal led for andsupplementary warnings are telegraphed to threatened districts atthe discretion o f the Forecast Official . The night

, o r twenty-fourhour , forecasts are o f value chiefly in cities and towns where the predictions are dissem inated through the medium of the morning news

papers . The early closing o f telegraph and telephone o flice s in thesmaller towns and villages prevents a prompt transmission o f thenight forecasts to outlying districts .The verification o f forecasts for thirty-six hours shown by the

tables is determined by the conditions presente d by the morning andevening reports o f the day succeeding that for which the forecast i smade , and the verification o f the night

,o r twenty-four hour

,forecasts

i s based upon the data wh ich are shown o n the night charts o f thefollowing day . Cold-wave signals are verified if the required fall intem perature occurs within thirty-s ix hours after the signal i s ordered ,although the order mu st specify the period within which the fal l i santicipated . A forecast o f rain f o r a State requires for a ful l ve r ification that seven-tenths o f the State shall be embraced wi thin therain area . When a sm aller portion o f the State i s covered by therain area the percentage o f verification i s proportional to the area o f


rain . When no rain fall s the percentage of verification is zero . Sim

il a r ly the percentage o f verification o f forecasts of temperature i sproportional t o the area of the district included by the temperaturechanges . Rainfall is considered f o r the twelve-hour periods ending

it at 8 a . m . and 8 p . m .,and verification o f tem perature forecasts i s

determined by the twenty-four hour tem perature changes .The following tables show the percentage Of verification o f rain a n d

temperature forecasts fo r twenty-four and forty-eight hours,and also

the percentage of verification o f cold wave and wind signals during

the last two years

Pe rce n ta g e of ve rifica t io n of r a in f o r eca s ts .

24 ho u r s . 48 ho u r s . 2 4 ho u r s . 48 ho u r s .

0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

O O O O O O O O O O O O O O O O O

0 0 0 0 0 0 0 0 0 0 0 0 0 0

Pe rce n ta g e of ve r ifica tio n of t em p er a tu r e f o r eca s ts .

2 4 ho u r s . 48 ho u r s . 48 ho u r s . 48 ho u r s .

M o n t h .

0 0 0 0 0 0 0

O O O O O O O O O O O O O O O O O O O O O O O O


Per cen ta g e of ver ifica tio n of win d s ig n a l s .

hi o n t h .

Nu m be r . Pe rce n t a g e . Nu m be r . Pe rce n t ag e . Nu m be r . Pe rce n t ag e .

J a n u a ryFe br u a ryDi a rch

Pe rcen ta g e of ve r ifica tio n of co ld-wa ve s ig n a l s .

M o n th .

Nu m be r . Pe rce n t a g e . Nu m be r . Pe rce n t a g e . Nu m be r . Pe rce n t a g e .

J a n u a ry.Fe br u a ryhda rch

O O O O O O O O O O O O O O O O O OOO

O O O O O O O O O O O O O O O O O O O O

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

While a thorough knowledge of general meteorology,and o f the

conditions peculiar to the various districts or local ities for which theforecasts are made is a recognized qualification

,the forecaster must

possess the ripe judgm ent gained by experience,and ‘ the confidence

and aptitude which nature alone bestows , in order to become e m i

n e n t l y successful in the practice o f weather forecasting . The prin

cip l e s upon which the forecasts depend are embodied in the application o f these necessary qual ifications to a solution o f the problem spresented by the weather m aps . The tabulated data Show the degree o f success attained in predicting the several changes o r characterof weather covered by the forecasts . Rain or snow forecasts are a dm it t e d ly the most uncertain and are the forecasts which yield thelowest percentages o f verification ; yet this class o f predictions i s u hdoubtedly of the greatest importance . If

,therefore

,rain can be

accurate ly predicted seven tim es out o f ten it may be safely assertedthat we are able to predict rain . A s regards fair weather forecaststhe percentage o f verification is 15 to 2 0 per cent higher ; we may ,therefore , also claim success in forecasting fine weather . It i s safe tosay , in this connection , that the failu res in rain forecasts are largely


confined to what are termed “ local rains,and that general and heavy

precipitation i s in a m uch larger percentage of instances an ticipated .

The same m ay be said o f the forecasts o f temperature ; in this classo f forecasts marked and general changes are rarely unpredicted . The

success attained in wind signal s is scarcely represented by the percentage of verification

,as in many instances dangerous winds which

out-going vessel s would encounter will prevail within . a few hours ’

sail of the port at which the signal i s displayed without a justifyingvelocity at the station . The success in verifying wind signal s

,as

shown by the table (about 77 per cent ) , p r o ve s ,'

h o we ve r,that w e are

able to predict wind storm s f o r specified points w ith a m arked degreeof success .In fair weather the public desires to know when rain o r snow may

be expected,and when rain or snow is fal ling the demand is confined

to enlightenment as t o the duration of the storm . The public wishes

to know when to prepare f o r cold waves and frost,and requires fore

casts,n o t a statement of existing conditions . The interests reached

and benefited by good forecasts are n o t confined to any business,

profession,o r class . Commerce and agricul ture

,deal ers and handlers

o f al l description o f goods and produce Subj ect to inj ury by the elements , the well and ill , al l have m aterial or personal in t erests ,

which

render valuable a knowledge of the weather o f to-morrow . At thepresent stage o f the science o f m eteorology the character o f theweather o f to-morrow can be regularly and successfully anticipatedfrom the morning reports more than eight o u t of ten times . In caseswhere the changes are m arked the degree o f success i s unquestionably greater .As the forecasts for forty-eight hours are Optional with the Fore

cast Official , the percentages given d o not represent resul ts comparable with the predictions made f o r shorter periods ; neither d o theyshow the degree of accuracy that i s possible in forecasts made regul a r ly for the longer periods .The energies and resources of the Weather Bureau are n ow being

devoted to attempts to determine and adopt m ethods calculated toimprove the forecasts . E ff orts will be m ade not only to obtain thebest forecasting talent

,but to solve by well directed scientific studies

and investigation many of the perplexing problems which n o w con

front the forecaster .— M aj o r H . H . C . Du n wo o dy, U. S . A. , Ch ief of

Fo r eca s t Divisi o n .

S E C T I O N I I .

R IVERS AND FLOOD S .

1.-FLOODS OF THE M ISSISS IPPI RIVER , W ITH REFERENCE TO

THE INUNDATION OF THE ALLUVIAL VALLEY.

WILLIAM STARLING.

The Lower Mississippi derives its main supply of water from threedi ff erent sources— the val leys o f the Ohio , th e Upper Mississippi , andthe Missouri . These regions have many points o f dissim ilarity . The

valley of the Ohio belongs t o an early geological period , the upheava lof the Appalachian chain h aving occurred at the close of the Paleozoic tim e . The valley of the Upper Mississippi w a s partly contem

p o r a n e o u s with that o f the Ohio , while a part i s o f l ater date . The

valley of the Missouri i s of comparatively recent origin . The Rocky

Mountains were hardly in existence when the Ohio was already afull-grown and even an old river . The place o f the Missouri o f to-dayw a s then occupied , first by a prolongation o f the Gulf o f Mexico ,afte rward by a series of great fresh-water lakes . Finally

,o n the

completion o f the elevation of the Rocky Mountains,in the late Ter

t ia ry period , the Missouri assumed som ething like its present shape ,subj ect

,of course

,t o the subsequen t changes o f the Glacial , Cham

plain,and Terrace epochs

,by which its full development as a river

w a s completed,in common with that o f the other tributaries and of

the great alluvial valley o f the united stream .

The physical and climatic features o f these val leys are as variousas their orig in . The Missouri Basin i s much the largest of the three

,

containing about square miles . The Ohio Valley i s n o t half

so great,possessing a watershed of only square miles . Small

est o f all i s the valley o f the Upper Missi ssippi,with an area of only

square miles . These grand basins are Visited by rainfal l swhich are of great diversity . In the Missouri Basin the annual downfall i s given by Humphreys and Abbott as only inches

,and of

thi s it i s estimated by the sam e authorities that only 15 per cent i sactually turned into d r a m a g e . The Upper Mississippi region has a

rainfall o f inches , o f which 24 per cent makes it s appearance inthe channel of its river . In the Ohio Valle y the quantity o f rain

68



comm on basin .They have

,j ointly

,a watershed greater than that

o f the Upper M is s is s ipp i, , a n d annually pour into the great trunkstream more than half as much as the Missouri . The Saint Francisand Yazoo each contribute about o n e -fourth as much . Red River ,while a very im portant stream ,

yet enters the Missi ssippi so near it s

m outh that its floods are of only local importance . Moreover , a co n

s id e r abl e part o f them passes o u t through the Atchafalaya .

The Ohio,Upper Mississippi

,Saint Francis

,and \Vh ite rivers are

n o n -sedimentary streams . They flow mostly through hard forma

tions and stable beds , and at low and m edium stages their waters areclear

.The Missouri

,Arkansas

,Yazoo

,a n d Red rivers are always

m ore or less muddy , flowing , as they d o , in the lower portion Of theircourse

,through beds o f their o wn deposits . The qualities Of the sed

im e n t s which they bring to the great river are diff erent . That o f theMissouri is mostly light and loamy , and thus is capable of beingtransported for very long distances

,as wil l appear from the fact that

the quantity o f matter carried in suspension at New Orleans is muchgreater during floods from the Missouri than at other times . Thesedim ent from the Arkansas is usually of a reddish color

,as i s well

known to the pilots and other river men,all of whom can tell

,by a

glance at the waters,whether the Arkansas is .in freshet .

From causes which will appear hereafter , the western rivers , theUpper Mississippi

,the Missouri

,the Sain t Francis , White and Arkan

s a s usually act together . The Ohio and it s tributaries are governedby different condition s . It has never occurred that al l these stream s

were in extrem e flood at the sam e time . Such an Ominous co n j u n c

tion would produce a discharge o f more than cubic feet,

which would be greater by one-half than has ever been known . But

it has often happened that an extreme flood from the Ohio has met amoderate freshet from the western streams

,or that a high May or

June rise from the latter h a s found the Ohio at more than a medium

stage . In either o f these cases,great floods may and do occur in the

lower river . In fact,it is quite possible for a very great flood in

either the eastern or western system,com bined with a m erely average

stage in the other,t o produce a very high wate r below the mouth of

t h e Arkansas . The usual order o f floods from the several tributariesi s as follows : the Ohio

,Cumberland

,and Tennessee

,al l together ;

then the Upper M issi ssippi ; then the Missouri , with probably theother western rivers .It is Often affirmed that the principal cause o f floods is the meltin g

o f the winter ’ s ice and snow . In the case o f the Ohio this statement,

as a general ity , m u st be denied,and it i s believed that this denial

may be extended t o the other rivers . Certainly,it may to the

southern streams , including the Arkansas . In point of fact, som e ofthe greatest snowfall s ever recorded in the Ohio Valley h ave dis

FLOODS OF THE M ISSISSIPPI . 71

appeared without producing any considerable eff ect,and some o f the

greatest floods have occurred when there was no snow o n the ground .

The m ost potent agent in producing floods has been excessive rainfall . This assertion does not exclude the action of snow where itexists . On the contrary

,it cooperates with it m ost powerfully . A

deep snow,passing o ff with a warm and heavy rain

,o f course greatly

augments the effect o f the latter,and that t o o at a very critical time .

Every fraction of an inch added to a considerable rainfall increasesits eff ective volum e f o r drainage more than proportionately ; f o rinstance

,in a dry season

,an inch o f rain will n o t much m ore than

be absorbed by the ground,but if two inches fall the additional inch

will nearly all turn to drainage,and o f three inches two will be avail

able f o r river water . Thus,a three-inch rain wil l be doubly a s d a m

aging as one o f two inches,and infinitely worse than o n e o f a single

inch .

The heaviest,most sudden

,most violent

,most extensive

,and m ost

dangerous rains that occur in the Ohio Valley are those accompanying

the cyclonic storms which originate o r have their ful lest developmentin the vicinity of the Gulf o f Mexico . These

,as is well known

,gen

e r a l ly first assum e form idable proportions within the territory of theUnited States in southeas tern Texas , and then fol low the Ohio Valley to the Great Lakes

,making their exit at the mouth of the Saint

Lawr ence River . They are usually accompanied by very low barom

eter,southerly winds

,m uch thunder and lightning , and deluges o f

rain,which last is n o t merely caugh t by the mountain ranges that

intercept and,as it were

,strip the clouds

,but al so fall s with equal

impartiality o n al l parts o f the lowlands . A considerable flood has

been known to result from two of these storms,each distributing

about three inches of rain over the Ohio Valley , at an interval o fabout a week apart. The season wa s March t o April . There w a s l ittlesnow

,and the rivers were decidedly low . These storms are supposed

,I

believe (though here I think I am speaking Latin before clerks ) , to follow the direction o f the prevail ing current o f the upper air

,and con

sequently generally travel som ewhat from southwest to northeast,

skirting the western slope of the Al l e gh a n ie s . Their path does notusually include the upper waters of the rivers of the western system

,

though it may encounter their lower portions . They have been

kn own to prevail for weeks,o n e fo l lowing the other at interval s , two

or three in a month,literally drenching the whole country through

which they passed,as probably in the fam ous season of 1867 , and

certainly in the more famous one of 1882 . In the latter year , therainfal l in the Ohio Valley for January was 60 per cent m ore than

the normal f o r the month . In the Cumberland and Tennessee valleys it was th r e e tim e s the average . In the Lower Mississippi Valley it

was more than double the mean . In February, f o r the Ohio Valley


the excess wa s 70 per cent ; for Tennessee , 65 per cent ; for the Upper

Mississippi,50 per cent. The extrem e heigh t o f the flood was reached

at Cairo o n the 2 6t h o f February . In February , 1884 (flood month ) ,Ohio Valley , 70 per cent excess ; Tennessee , 70 per cent . There was

al so m uch snow on the ground . In January , 1890 , Ohio Valley and

Tennessee,35 per cent excess ; in February , 60 per cent ; in March , 40

per cent . There was excessive rain in the valleys o f the Arkansasand White rivers

,especial ly the latter

,keeping it near extreme flood

nearly four months . The flood in the lower river was in March and

April . In 1892 the Ohio was at a m oderate stage throughout . Theflood came from the western rivers

,and culm inated at Cairo o n the

2 7 t h o f April . Before this had tim e to recede it was caught byanother which reached its maximum o n the 2 5—26th o f May . The

rainfall for April in the Upper Mississippi Valley was 80 per cent inexcess of the mean ; for the Missouri Valley , 80 per cent . In May ,Upper Mississippi

,80 per cent excess ; Missouri , 60 per cent ; in the

valleys of the Arkansas,White

,and Saint Francis

,87 per cent . There '

w a s a great flood at Saint Louis,th e highest since 1858, and a very

great one at Fort Smith,the highest o n record

,and at Little Rock ,

the greatest,I believe

,except 1833 .

Thus,there are two classes o f floods which afflict the alluvial plain

o f the Mississippi— the early floods which proceed mostly from t h e

eastern r ivers,and the late

,which owe their origin principally to the

western stream s . The form er usually reach their height in the lowervalley in March

,the latter in June . Som etim es there occur very

high waters of a third class,interm ediate between the two

,in which

a late rise from the Ohio meets an e a r l v freshet from the Upper Miss is s ipp i and Missouri , accom panied , as usual , with sharp ri ses fromthe lower streams . Such floods culm inate in April or M ayf It i s a

question which o f these is most dreaded by the dwellers in the alluvial region . The March floods are frequently very great in volum e

,

com e up very rapidly,and take pl a ce at a very inconvenient and ih

clement season . They are usually accom panied by very high,cold

and persistent winds,often with chilling rain o r sleet and plenty o f

mud and general discom fort . No w,if there i s anything which levee

engineers dread,i t i s storms when the water stands at a very high

stage against their levees . The river at such times i s two m iles,o r

more , wide .

’

There are long stretches over which the wind has full

sweep for eight or ten miles up or down stream,o r in “

o l d river

lakes,

” and a February o r March norther som etimes gets up a form id abl e sea , breaking with great force clean over the levee and men

acing it with speedy destruction . Provision i s m ade against thisdanger , to som e extent , by sodding the slopes o f t h e levees , by giving

them a long and gentle incline towards the water,and in the mos t

exposed places by revetm ents o r breakwaters of plank . It would be


easy enough to make an eff ectual defense,as i s done in Holland

,Ger

many and Italy,by facing the front of the dikes with stone

,but that

is expensive,and the Mississippi levees have not yet reached the stage

o f development when such a course is practicable . They are stil lstruggling toward com pletion . Severe and prolonged storms

,how

ever,will cut through even a long and well-sodded slope

,and plank

are n o t always to be had in an em ergency. Moreover,the levees are

in a state of transition , and frequently the sod h a s not had time togrow

,and it would n o t be proper to erect permanent and costly de

f ens e s when fresh accessions o f earth have to be m ade every two o rthree years . So resort must be had to temporary expedients fo r prot e ct io n against the violence o f waves , and chief of these are what aresometimes improperly called sa n d ba gs . Improperly

,because the less

sand there is in them the better . They a r e strong and new sacksfi lled

,by preference

,with the heaviest and s t ifle s t clay that can be

had,though

,generally

,use must be m ade of the first material that

comes t o hand . These must be filled and placed in the teeth o f a

howling “ blizzard,

” every blast of which cuts through the men,wet

to the skin as they are with spray o r rain . To the novice it alwaysseems wonde rful that earthen dikes can be held at all against suchheavy odds

,and yet it i s but seldom that they are lost from thi s

cause .

The engineer,then

,would rather undertake to hold his levees in

May or June,when the sun is warm

,the air genial and m i ld

,the

earth dry and the days long . Then his men are cheerful and w il

l ing,and

,if they are negroes

,sing all the day as they work . But the

planter dreads a sum mer flood . Should the levees break and he beo ve rflo we d in March , the river will probably subside in time for himto make a crop . But if such a misfortune happen in May or June , itwill be too late to calculate on a full return

,and he must either lose

the season altogether o r incur the expense of replanting with a pros

p e ct of on ly half a crop , o r thereabouts .Overflow is n o t the only calamity that a planter dreads from a

flood . Even with the best constructed levees a vast quantity o f waterleaks through the porous natural soi l beneath the base of the embank

ment,invades the roots o f hi s growing corn and cotton , and even rises

in the furrows o r perhaps above the ridges,thus e fle ct u a l ly drowning

o u t his crops . This seep water , as he cal l s it , is an unmitigated evil .Inundation water brings with it ferti li zing silt . Seep water has been

stripped o f all this by filtration th rough the ground . It is sterile ,stagnant and foul . It destroys cotton or corn ,

and seem s really to

retard the recovery of the land even after it h a s disappeared . Ditching will n o t always get rid of it . Th is damage , too , is far greaterwhen the season is well advanced

,for very sim ilar reasons . Fo r t u

n a t e l y, seep water usually affects only the ground immediately adj a


cent to the levee,and generally only the l o w-lying parts of that.

Moreover,it does not inj ure grass , so the planter may turn his l ow

grounds into pasture .

The highest floods that have ever prevailed in the Mississippi p r o

ce e d e d m ainly from the Ohio , and culminated in March . These were

in 1882 and in 1884 , which are believed to have brought down a

greater quantity o f water at o n e tim e than any of which we have an

accurate account . The floods from the western rivers , while some o f

them have been attended by a discharge n o t very much less thanthese

,have usually been more remarkable f o r their duration .

‘

Of

course,it is a m atter o f the h ighest consequence to forecast, if pos

s ible,the progress of a flood

,and this can best be done by com parison

with the records o f past experience,to see if perchance any analogy

can be traced wh ich may lead to probable inferences f o r the future .

With this view ,the progress of the d ifle r e n t floods h a s been graph

ica l ly indicated by curves o r hydrographs for the different stations,

and much study has been given to them . On the whole they are

very disappointing . As a general ru le there is very little resem blancebetween the various floods o f the Mississippi

,even between those of

the sam e orig in . Wi th a few exceptions there i s hardly any analogyto be traced between the hydrographs o f any two floods at whateverinterval . The most remarkable o f the exceptions al luded to i s thatof the three years

,1882

,1883

,and 1884 , which exhibited a great simi

l a r ity. All cam e principally from the Ohio ; all culm inated at Cairo ,from the 22d t o the 2 7 th o f February

,at a gauge height varying from

to being the highest o n record ; and neither was much

complicated with the western rivers . The floods o f 1892 and 1893

also presented a general similarity,each culmin ating in the lower

river about th e first o f June .

It might be thought that it would be an easy matter to predictfrom a given rainfall in the several valleys the stage t o be reached at

the different points along the course o f the main river,but in reality

it is a very complicated problem . It has already been seen that theratio between downfal l and drainage is very diverse in the severalwatersheds , and even in any o n e region this rate differs from itselfby a very wide range of discrepancy , according to season and cir

cu m s t a n ce s . A rain o f s ay three inches in January does not by anym eans signify the same thing as a rain of three inches in June . In

the one case the ground is frozen,the skies are mostly cloudy

,the

days are short , the a ir is cold , the trees are bare o f foliage , the earth

without any cover o f vegetation . It is well known that taking theaverage o f a whole year there i s no very great di ff erence betweenrainfal l and evaporation ; but the ratio between these two elem entsf o r the several months is widely diverse . Observations made during

a period o f 143 years , at a station in Holland , o n the border of Haar


l e m Lake , show that during Novem ber , Decem ber , and January the

rainfal l is about four times the evaporation . In May and June the

evaporat ion is about double the rainfall . In March the two areabout equal . The course o f the seasons i s diff erent in thi s country

,

and in the Mississippi Valley , at least , would seem to be about amonth l ater .

Now,in June

,the earth is

'

dry,porous

,and receptive

,the sun h o t ,

th e air has greater 1 capacity for m oisture,the days are at their

greatest length,and the whole surface o f the ground

,wild o r culti

va t e d , woodland , field , plantation , o r m eadow ,i s covered with leafage .

It is,therefore

,often found that a storm which would have produced

a calamitous freshet in win ter o r early spring,raises the rivers but a

few feet in summer . Hence it i s,fo r o n e thing

,that spring is the

season f o r floods . It is not that the rain is s o much heavier,o r that

the quantity o f snow on the ground is s o great,but that whatever

does fall is in a great measure converted into river water . Thoseengineers wh o have investigated the question of water supply forcities have found that the proportion o f rainfall which finds it s way

into the water courses i s,in January

,about nine-tenths ; in June ,

four-tenths ; in August , o n e -tenth .

There are many other considerations which greatly influence theheight o f floods . On e o f them is the condition o f the ground as t omoisture when the decisive rainfall occurs . It m ay be that this isalready saturated W i th previous rains o r by melting snows

,s o that it

will not readily absorb any m ore,and thi s independently o f season .

It has already been remarked that a sudden and heavy rain,a “ cloud

burst,

” as it were,produces rises in the stream s o u t of proportion to

the actual quantity o f water which fal ls,and two or more of these

,

with only a short interval between,m ay bring about a disastrous

flood when the sam e rainfal l,spread over a m onth

,would have been

com paratively harmless .

A very important elem ent to be considered,as regards the stage to

be attained,especially in the lower trunk

,is the height at which the

principal rise finds it . It is a proverb among the denizens o f theMississippi Valley that a ful l river o n the first o f January portendsan overflow in the spring . This simply signifies that it i s an uniavo r abl e prognostic

,so far a s it goes

,for the March rise to find the

lower river already occupied by a great volum e of water . Suppose

a heavy storm in the Ohio Valley to produce a rise o f 2 5 feet at

C incinnati,Nashville

,and Chattanooga . It is obvious that it m akes

a great d ifle r e n ce to the people of Greenville , in Mississippi , whetherthis freshet finds the river at that point already standing at 35 feeto r 15 feet .A second element of great consequence is the duration o f the rise .

A freshet which rapidly goes up to 50 feet at Cairo and then as rap


idly fal ls,as in 1886

,is less formidable , and will actually reach a less

height at lower stations than one which i s a foot o r two l ower at thehead o f the alluvial basin at Cairo , but remains at that stage for a

week o r m ore .

A third control l ing elem ent o f a flood in the lower river is the localrainfall in the southern portion o f the basin , which is sometimes sogreat a s to keep that part o f the river

'

, for weeks , higher by severalfeet than the normal relation between the gauges would prescribe .

The rainfall in the lower part o f the valley is sometimes extremelyheavy . In the Yazoo B a sin the rainfall fo r April , 1874 , was 22 inches .

In 1893,the present year

,the rainfal l at Helena , in May , in a single

week,was 144 inches , and the quantity o f water which fell in the

Saint Francis Basin wa s sufficient to maintain the stage at Helenaabove 45 feet o n the gauge f o r about two weeks , when its normal

stage should have been 42,o r less .

The rate o f travel of the flood wave is a subj ect o f much interestand importance

,and it might be thought that it would be very easily

predicted by a reference to the experience o f form er years . It i s

liable,however

,to several perturbations . It is influenced by slope

and by height o f stage,and i s complicated by the intervention of

tributaries and reservoirs . A word as to the latter .

While a great part o f the alluvial plain o f the Mississippi,from

Cairo to the Gulf,i s defended by levees

,yet there are two large ter

r it o r ie s which are nearly destitute o f such protection . These are thegreat bas in of the Saint Francis and the lesser plain which Spreadsout between the hill s and the Mississippi

,at the confluence of the

White and Arkansas rivers with the trunk stream . The upper parto f the latter district wil l soon be shut o ff by levees

,when its disturb

ing influences wil l in a great measure cease,except as complicated

with the action o f the great tributaries which empty into it . The

Saint Franci s Basin is likely to m aintain its present condition forseveral years . In fact

,few active steps have yet been taken f o r its

reclam a tion . There are several such great bottoms in the MississippiValley , defined by the approach o f the tertiary hil l s on either side ,each having its charac teristic tributary stream

,but most of them

have been inclosed by levees,and need not be considered with refer

ence to o u r present subj ect . The Saint Franci s Basin,however

,e xe r

cises a very im portant influence upon the flood stages o f the river inits front and f o r some distance below it .

Th is basin , o r bottom ,is merely a large tract

,some square

miles in area , o f alluvial land,intersected

,however

,by o n e o r two

ridges o f an older period,and bounded by the Mississippi o n the east

,

and by the h il l s cal led Crowley ’ s Ridge o n the south and west . Theminor features o f th is basin were m uch disturbed by the earthquakeo f 1811

, which exh ibite d its greatest activity in the neighborhood of



Yazoo Basin,at the foot o f the latter at Vicksburg , and another at

the m outh o f Red River , acting as the receptacle o f the flood waters

o f the Tensas Basin . The engorgement is only local ; the increasedheight and slope conferring a more than average velocity , which soonflattens o u t

” the flood wave,so that 10 0 miles below Helena it

would signify little whether the Saint Franci s Basin were closed o r

not,were it not for the intolerable retardment o f the flood .

In applying the records of past observations to the purpose o f de

du cin g therefrom the probable stages to be attained by a flood j ustcom ing in sight , it is indispensable to have an accurate knowledge ofthe changes that have occurred and that are continually occurring inthe lower river

,especially in the way o f levee building . During the

period from 1882 t o the present , and especially since 1884 , there has

been tremendous progress made in th is direction,by which the whole

high-water regimen o f the Lower Missi ssippi has undergone a radicalal teration . It i s very diffi cult , therefore , to derive m uch instruction ,in the forecasting o f flood stages and periods , by a direct comparison

o f the records o f years anterior to 1885,and below Arkansas C ity ,

anterior to 1888. This i s the more to be regretted , f o r 1882 , 1883and 1884 were undoubte dly three of the greatest floods that everoccurred

,particularly the first and last . The principal changes that

have occurred are the com plete closure o f the Yazoo front in 1884the rebuilding o f the Arkansas levees o f t h e Tensas front in 1886and the raising and strengthening o f all the different works o f thisclass which has been constantly going o n ever since . It is necessaryt o bear these alterations continually in mind , in attem pting to reasonfrom the older data

,else we shall make serious mistakes . In attempt

ing,again

,to apply analogies based upon any great flood year

,we

must know the circum stances which prevailed in that year,especially

whether any crevasses occurred,and if s o

,when and h o w large

,and

how extensive was their influence . For instance,in 1882 , in some

places,half the water of the river went over the banks . In 1890

there were many localities where o n e -fourth of the discharge was lostin the same way . In 1892 two -thirds of the portentous outpour ofthe Arkansas— more than half that of the Missouri at full flo o d— neverreached the Mississippi at all

,unless it may have been throu gh the Red

River , but went around the head of the Tensas system o f levees . It

is want o f acquaintance with these detail s which has gone so far to

cause the phen omena o f the Mississippi,in tim e o f high water

,to be re

garded as anom alous,and has caused the failure o f many predictions .

It is evident, then , that if any instruction i s to be derived from therecords o f the past

,it m ust be as the result of attentive study

,and

that any conclusion drawn from such records must be in the natureof a calculation in which all the perturbing influences must be takeninto account . There are several m ethods which promise a h ope o f


success in this way . If discharge observations have been sys t e m a t

ica l ly t aken at any point , f o r instance Arkansas C ity , and a certainratio o f progression establ ished between the discharge and the heighto f the gauge , then if the probable discharge can be predicted theflood height may become known . Again

,if w e can find any gauge

which has not been subj ected t o disturbing influences,but h a s r e

mained unaltered for many years,and if we can trace any parallel ism

o r any known relation whatever between it and the gauges below atstages less than the highest

,then by analogy the relation may be ex

tended to extreme heights . Such a gauge is that o f Cairo . If,there

fore,the stage a t Cairo be given , o r can be calculated , the heights of

the lower stations may be estimated . Now , i t is possible from close

observation o f the rises in the great tributaries,or even from the r e

pOr t s of the rainfall , to m ake a pretty fair approxim ation to the stageto be reached at Cairo by any. flood , and the probable d ischarge can

be estimated from the stage and other circumstances .Discharge observations have been taken at several points with co n

s ide r abl e care , and , though scattered and fragmentary , they extend

over a period o f many years . So far a s any hope is concerned o f

deducing a regular relation between discharge and stage,they are

extremely disappointing , f o r they are discrepant and apparently capricio u s beyond measure , and frequently show that the greatest volumepasses at a stage much below the maxim um , even several feet belowSome of the disturbing causes are known and calculable . Others are

stil l involved in much obscurity . In the present state of ourknowledge

,then

,not m uch assistance i s to be derived from this

method .

The relations between the several gauges have been made the s u b

j e ct of study by several engineers , particularly by Colonel Suter andCaptain Rossel l

,o f the Corps o f Engineers

,Un ited States Army

,who

have drawn many interesting conclusions . As an exam ple of theapplication of this method it m ay be said , roughly speaking , thatunder ordinary circumstances a stage o f s ay 48 feet o n the Cairo

g auge corresponds to about at Helena,49 at Arkansas C ity

,44

at Greenville,and at Vicksburg . Of course

,these figures are

subj ect t o m odification in all sorts of ways,by the behavior o f tribu

taries a nd reservoirs,by diversities o f slope

,by duration of flood

,

and by other causes,a l l o f which m ust be taken into account i f an

estim ate is to be at all accurate— and even half a foot i s a matter o fserious consequence at the top o f a great flood . The local “ river

prophets ” have acquired considerable skil l in thi s sort o f prediction ;and when once the extreme height at Cairo has been reached

,or

plausibly calculated,they can foretell the progress o f the flood down

stream within pretty narrow limits . A s to general prognostications ,they also have a number o f saws

,such as the one already quoted ,


that a full river o n the 1s t of January indicates an overflow in thespring . Another is

,that an “ open ” or mild winter forebodes high

water,and a severe winter l ow water . So far as spring floods are

concerned,this view is rational enough . Warm winters imply

southerly winds and Gulf storms— cold ones,high barometer and

winds from the north and west . As to summer floods , they dependon more remote causes .The most wearisom e

,exhausting and dangerous floods are those

which are composed of a succession of rises,each o n e catching its

predecessor before the latter h a s had time to subside . The beginning

o f a flood wave travel s very fast . When the river i s 50 feet at C incin n a t i a n d 25 feet at Cairo the slope is steep and the topmost layersof water move at a great speed . On the contrary , the rate o f recession is slow . When the river is 25 feet at C incinnati and 50 feet atCairo the movement o f the wave is sluggish

,and it i s easily overtaken

by a sharp freshet . In this it i s aided by the tributaries near them ain stream

,fed by the rainfall in the interior valley

,which pour

forth their floods almost instantly,and check the fal l in a very short

time . An inspection of the hydrographs of s ay Cairo and Vicksbu rgshows a series o f elevations and depressions in the former where thelatter exhibits an almost unbroken rise . The great “ num ber of thetributaries o f the Mississippi m akes it peculiarly obnoxious to theseincidents . When the Ohio

,Cumberland

,and Tennessee have begun

fall ing at a good rate , it is pretty hard ,” as Sir Lucius O’

Tr igg e r would

say,if a freshet can not come from the Upper Mississippi or the Ill inois

and keep the water up at Cairo til l the eastern rivers have g o t their seco n d wind

,or fail ing that

,i f theWhite and Arkansas can n o t give it a

fil l ip , j ust t o keep the ball in play . I hope none o f my audience knows

the feeling o f hope indefinitely postponed that seizes upon the suffererwh o has been “ figh ting high water

,

” as it i s very appropriately te rmed,

for two months , and wh o has been f o r two weeks anxiously waitingf o r the fall at Cairo to reach him

,o n hearing that there has been a

rainfal l o f 6 inches at Fort Smith and Little Rock,and a ri se of 14

feet in twenty-four hours .The damage done by overflows In the Mississippi Valley is generally

much overrated . The loss of life i s usually absolutely nothing . Un

less o n e i s unfortunate enough t o l ive actual ly j ust behind a .levee

when it breaks , o r t o be standing upon it , there is very little danger .

The reason o f this is that the drainage is excellent,and al l toward

the back country , and that the fall o f the water surface i s very rapid ,the water spreading in all directions

,and fi l l ing up the swamp s very

slowly .



Wa t e r g a u g e s o n th e M is siss ipp i Rive r— Co n t in u e d .

Dis t a n ce Ga u g e r e a din g s .

L fr o m t h e Ga u g e z e r o

o ca t i o n . Gu If o f a bo ve m e a n

M e xico .Gu l f l e ve l . Lo we s t . H ig he s t .

Be lm o n t , M o

Ne w M a d r id , M o

Co t t o nwo o d Po in t , M o .

Fu l t o n ,Te n n

M e m ph is , Te n nM ho o n La n d in g , M is sHe l e n a , ArkSu n flo we r , M is sM o u t h Wh i t e Rive r , ArkArka n s a s C it ArkGr e e nvi l l e , is s

La k e Pr o vide nce , LaVicks bu rg , M is sSa in t J o s e ph , LaNa tche z , M is sRe d Rive r La n d in g , LaBayo u Sa r a , LaBa t o n Ro u g e , La . 4 .

P la qu e m in e , LaDo n a l ds o n vi l l e , LaCo l l e g e Po in t , LaCa r r o l l t o n , LaNe w Or l e a n s , LaFo r t J acks o n , La

The distances from the Gulf o f Mexico,given in the above table

,are

channel distances above Saint Louis and m id-bank distances belowthat point to the Gulf . The elevations o f gauge zeros are derived

from duplicate lines o f precise level s extending from tide. water of

the Gulf along the river to Saint Paul . The lowest and highest

readings given are the lowest and highest stages , respectively , thathave been recorded from the tim e the gauge was established untilJul y 1, 1893 .

Bank-full stage means that stage of water which reaches the top ofthe average banks in the vicinity o f the gauge .

From the above table the slopes at high and low stages betweensuccessive gauges may be deduced .

Gauges have also been established on all the principal rivers ofthe United States

,and the river-stage bul letin o f 1892

,i ssued by the

Weather Bureau , com prises daily readings at 160 stations .The credit of developing this department of the Weather Bureau

belongs to Prof . Thomas Russel l,and the last bulletin bears strong

evidence o f his energy and good j udgment .

Careful ly kept and accurate continuous records of this kind willbecom e invaluable to the future engineer who takes up the study ofdecrease in flow o f streams . Careless records are m isleading and ”

worse than none,and it i s to be h oped that the Secretary of Agri

culture will see the necessity o f keeping this departm ent in thehands o f a m a n wh o fully appreciates it s im portance , and h a s theskil l and j udgment necessary to secure and digest the desired results .The flood planes o f the river become more and more marked as w e



that concentrate the waters and increase their scouring capacity .

As this effect varies with the magnitude and duration o f the flood ,it i s very diffi cult to measure . The maximum am ount i s reachedwhen the scour reaches bed rock

,which it did at the Merchants Bridge

during the flood o f 1892 . This scour was more than 2 0 feet deep insome places

,and the channel capacity w a s nearly or quite doubled .

This,the greatest known flood near the j unction of the Upper Mis

s is s ipp i and Missouri rivers , does n o t seem to have caused excessivelyhigh wate r in the Lower Mississippi . The maximum stage of thatyear at Vicksburg was several feet lower than other years , and o c

curred several days prior to the maximum stage at Saint Louis ; hence ,could not have been m aterial ly influenced by it .

Referring again to Plate 11,we see that Cairo i s situated at the j unc

tion o f the Ohio and Mississippi rivers . The drainage basin of the

Ohio and its tributaries i s square miles . The M is sis s1pp i and

Missouri basins above Cairo compri se square miles,or a total

above Cairo o f square miles .

During the years 1882 , 1883 , and 1884 the average discharge

amounted to cubic feet per second . The river has an extremerange o f 53 feet between high and low water .

An inspection o f Plate III shows beyond question where the floods

o f the Lower Mississippi come from . The great floods o n the Ohiobegin in February and have passed on down long before the floodsof the Missouri and the Upper Mississippi reach the mouth of the

Ohio . The only exception was in 1875 , when a flood from the Missouri on August 1 wa s j oined at Cairo by a moderate flood from theOhio River . This caused an overflow down as far as Lake Provi

dence . The maximum stage at the latter point occurred , however ,some three months prior t o th e arrival o f the Missouri wave .

The floods o f the Missouri and Upper Mississippi rivers have neverbeen of such volume as to become a serious menace by themselves to theLower Mississippi Valley

,and as they never come in conj unction with

o n e another , o r with .the great floods of the Ohio and its chief tributaries

,they have but l ittle

,if any

,influence on the flood planes of

the Lower Missi ssippi River .

Thus , the startl ing statement that an acre reclaim ed from the ariddeserts of Montana by means of reservoirs wil l reclaim another acre

from the floods in Louisiana is seen to be wholly lacking in theessential elements o f fact.After passing the Ohio the volume o f the Mississippi River at flood

stages is often increased by floods from the tributaries . The Whitein 1892 added cubic feet per second

,the Arkansas

cubic feet per second , and the Red River cubic feet per second .

A coincidence of floods in all of these streams may occur,and the


86

Tabl e of h ig h e s t a n n u a l s ta g e s a n d the ir da te s a n d n u m be r of days a bove overflow s ta g e

Ye a r .

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

Ye a r .

0 0 0 0 0

CHICAGO M ETEOROLOGICAL CONGRESS.

He rm a n n .

du r in g th e ye a r s 1872 to 1893 .

( M is s o u r i Rive r . )

0 0 0 0

0 0 0 0 0

0 0 0 0 0 .

0 0 0 0 .

a 0 0 0

o 0 0

0 0 0

0 0 0 0 0

o O

o 0

. 0 0 0 0 0

0 0 0 0 0 o0 0 0 0 0 0

0 0 0

Pa du ca h .

0 0 0 0 0 0 0 0

(Ohi o Rive r . )

o o o o o

0 0 0 0 0 0

0 0 0 0 0 0

0 0 0 0 0 0

Ha n n iba l .(Uppe r M is s is s ippi Rive r . )

0 0 0 0 0 0

Caa i1o

( M iddl e M i s s is m pp i

0 0 0 0 0 0

0 0 0 0 0 0

0 0 0 0 0

0 0 0 0 0 0

0 0 0 0 0

0 0 0 0 0 0

Sa in t Lo u is( M idd l e M is s is s ipp i Rive r . )

0 0 0 0 0 0

0 0 0 0 0

0 0 0 0 0 0

He l e n a .

(Lo we r M is s is s ipp i Rive r . )Rive r .

Ye a r .

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

FLOOD PLANES OF THE M ISSISSIPPI . 87

Ta bl e of h ig h e s t a n n u a l s ta g e s , e tc— Co n t in u e d .

La k e Pr o vide nce .

(Lo we r M is s is s ipp i Rive r . )

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0 0

Re d Rive r La n d i n(Lo we r M is s is s ipp i Rive r . )

0 0 0 0 0 0

0 0 0 0 0

0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

Ca r r o l l t o n , La .

(Lo we r M is s is s ipp i Rive r . )

0 0 0 0 0

CH ICAGO M ETEOROLOGICAL CONGRESS.88

Pa.3.3an

em

mu33S

.

2.

.

ws.

.

R.

a

s

aw.

fin.

.

mu.

£

3

em.

.

5.

.

R.

a

s

.

em.

am.

.

5.

a

s

aw.

.

R.

i.

as.

a

s

65.

.

R.

.

2.

£

8

.

mu.

.

mm.

.

mu.

.

«uw

.

R.

65.

.

2.

.

unm

8.

am.

.

2.

.

wkw

.

ws.

.

R.

i.

is

as

.

2.

.

fi.

és

3

8.

gm.

Em.

ea.

.

2.

i.

.

2.

saw

A

8.

.

3.

.

2.

is

.

8.

am.

.

ww.

gm.

.

2.

i.

as.

saw

0 0 00 0 0

«mm

gm

cow

Now

com

com

cow

«mm

«mm

«mm

«mm

«mm

Sm

00

N0

Go3

so

~O

mu3

as

m

onFa.

00 00 0O V V)

sc

i

owso

m

53m

55

SixCo

x

3532

.

mh

snm

xo

rw

31

3

853

ace—e

m

Q

EQ

EQ

E

5

0

5858

59

3Sam

”B

E—Em

E.

0

2.

cqa

fiu

o

m


RIVER STAGE PREDICTIONS. 89

Tabl e of h ig h wa te r p r i o r t o 1872 .

Ye a r .

0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Ye a r .

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0

O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O

3 .—RIVER STAG-E PRED ICTIONS IN THE UNITED STATES .

THOM AS RUSSELL.

The river service o f the United States Weather Bureau has 191river gauges

,mostly at large cities on the principal rivers throughout

the country . A record is kept o f the daily stages o f .th e water in theinterest o f low water navigation and f o r flood warnings . Besides theWeather Bureau gauges others are m aintained along the Mississippiand Missouri rivers by the M ississippi and Missouri River Co mmissions in the interest o f river im provements that are being made .

A river gauge consists of a plank about a foot wide and of sufficientlength to include the range Of water from l ow t o high water , and ismarked to feet and tenths o f a foot . It is fastened to a bridge pier ,where On e i s available

,or it may consist o f a narrow strip o f a stone

pier dressed down to a smooth surface to receive the marking andnumbering . Where there i s no bridge pier available , a gauge is made


o f heavy tim bers , 6 by 12 inches,laid along the incline of a river

bank,with a strip o f iron fastened along the t o p for the marking . A

gauge is placed with the zero o f graduation at the level o f the lowestwa ter a s near as possible . The marks indicate vertical heights Of

the water surface above l o w water . The gauge readings are called

stages.The stages read daily at 8 a . m . are telegraphed to various

places interested in information as to the stages o f rivers . At a

number o f the larger c ities throughout the country a river bulletin ,in connection with the weather maps

,i s i ssued daily from the

Weather Bureau offices .

At h igh water when there is danger of a river overflowing its banksthe . Observations o f stages are o f interest to districts liable to beflooded . To places where definite information o f the extent o f a

com ing high water can be given warnings are sen t by telegraph .

The highest water in a freshest,or the crest stage

,occurs first

toward the head waters o f a river . After a flood wave forms there i sa progressive motion of the crest down stream at the rate o f three o rfour miles per hour . Th is renders it possible to form some idea o f

what the highest stages o f water will be along the lower course ofa river from the stages along the u pper course . Better predictions o fhigh stages can be m ade the greater the length o f record o n which

to base a rule for prediction . Accurate predictions of river stagesat low water are not possible . Where the discharges of a river fo r

low stages are known,that is

,the quantity of water passing through

the river f o r diff erent stages near low water,estimates can be made

ahead o f stages below which the river will not fal l ; but the l east fallo f rain after the prediction is made makes the river rise at low stagesvery rapidly .

At high water predictions of stages are m ade in various ways,

depending on the nature o f the rivers .In the case o f two places o n the same stream

,the gauge readings

are more o r less closely dependent,according as the di stance between

them is less or greater . Th e gauge readings follow each other moreclosely in a rise the less the quantity o f water coming into the riverfrom the drainage area between them . Predictions of the lower stagefrom the upper one can be m ade more accurately the less the proportion o f the drainage area between the two places bears to the wholedrainage area above the lower place .

The character of a river varies greatly along its water course as toslope and width . Though two places on the same stream

,a hundred

m iles o r so apart, m ay have nearly the same quantity o f water passin g them , the stage at one of them may be twice as high as at the

‘

other , the difference being made up by the greater width of the riveror the greater velocity of the water at the o n e as compared with thato f the other . At Louisvil le

,for instance

,o n the Ohio River , 132



A rise at o n e o f the u p-river stations has more efle ct in‘

p r o du cin g

a subsequent rise at Cairo the greater the length of tim e the higherstage lasts at the u p-river station . The continuation o f the h igherstage beyond three days has no effect in increasing th e rise at CairoFo r a less continuation than three days , the rise at Cairo is a proportional part o f the greatest rise that takes place for a three-daycontinuation o f the high stage .

From a consideration of the s l e p e s of the river-surfaces betweenCairo and the u p-river stations for different stages at the variousplaces from low to high water

,and the varying cross-sections and

depths f o r the different stages at the various places , the ratio of ao n e -foot rise at the various places to the corresponding rises at Cairois derived . theoretically

,a s far as the data wil l perm it, taking in to

account the extent of cross-section and the velocity o f water as a ff e ct e d by diff erent depths and slopes .

A comparison o f the theoretical rises thus obtained with the risesactually observed in cases f o r which there are records

,gives a factor

f o r each place for certain stages . Only a few of the possible cases

that can occur have ever been observed .

The record o f stages at Cairo and the u p-river stations is too short ,as yet

,to furnish cases o f all the possible variety of combination o f

freshet wave crests from the various rivers which produce a high

water at Cairo . Fo r the possible cases which may occur in the future,

but o f which there are no observations as yet,the theoretical val ue o f

the rise found a s described is multiplied by a factor derived for thestages at wh ich there have been actual observations of rise . In thisway tables are prepared which show the relation of a o n e -foot rise atthe various places to the su bsequent rise in three

.o r four days at

Cairo,that being the crest-wave time between Cairo and the various

places . The rise at Cairo is taken as the s u m of the various co m

p u t e d rises at the five places .In case of a fall at any of the places instead of a rise

,it enters the

sum with a minus sig n .

The stage that will prevail at Cairo can be estimated six to sevendays ahead from the stages at C incinnati

,Chattanooga

,and Saint

Louis , with allowance for the water coming out o f the Cumberlandand Wabash . But as the cross-sections at C incinnati and Chattan o o g a are not known , the rule f o r prediction of stages has to be basedon the com parison o f actually observed rises .In cases where the discharges and cross-s e ct l o n s of rivers at pla ces

are not known,som e idea can still be form ed o f the relative impor

tance o f different tributaries in causing a rise at a point on a mainstream , provided there i s a long record o f stages with rises at thelower point due to rises sometimes in o n e of the tributaries and

RIVER STAGE PREDICTIONS. 93

sometim es in another . This permits o f estim ating the effect o f eachseparately .

A rise at a high stage of a river has more efle ct than an equal rise

at a l o w stage in producing a rise at a point lower down the river.

On the other hand , the higher the stage at the lower point the lessthe eff ect o f a rise at an u p

-river point in producing a rise below .

The products of the rises by the m ean stages during the rises aretaken as comparable throughout the range of stages at the place .

Fo r very high stages this does not hold good . Where a river over

flows its banks and becomes miles in width,very great rises at u p

river stations have very little effect in producing a further rise below,

and it i s impossible to estimate effects in such cases .In some cases the extent o f the drainage areas above u p

-river

gauge stations is taken into account in devising a rule for predictinga high-wate r stage at a lower point . In a case o f this kind at Pittsburg

,the prediction is based o n the stages at the following places

above it :

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

O O O O O O O O O O O O O O O O O O O O O O O O O O

o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o

The effect of a rise at a place in producing a rise at Pittsburg istaken as proportional to the square root of the area above it . Thewhole area above Pittsburg is square miles . The areas above

the six places comprise square m iles of the area above Pittsburg . The rise a t Pittsburg multiplied by the m ean stage during therise

,and by an unknown fac tor f o r a number o f selected cases o f

great rise,are placed equal to the sums o f the rises at the six places

above,weigh t e d according to the square root o f the drainage areas

above them,the unit o f area being taken as square m iles . From

these the value of the unknown factor i s derived . With the facto r atable i s prepared which gives the highest stage at Pittsburg when therises at the upper gauge are known .

Gauge readings at a place made o n successive days,or at interval s

a few hours apart,during a rise are o f som e service as indicating h o w

high the water may be expected to go . The characteristic of a risefor most places i s that the rate o f rise

,small at first, gradual ly

increases as the rise continues,until a maximum rate is attained

,

and then diminishes until it becom es zero at the crest stage . Thecharac teristic variations in rate o f rise vary greatly in differentplaces

,depending largely o n the slopes o f the ground over the drain

age area and o n the customary distribution o f rainfall . As a rule,the

chara cteristics of a rise are more permanent o r more nearly al ike in


different rises the greater the drainage area above a place . At C in

cin n a t i , f o r exam ple , o n the Ohio River, the rate o f rise begins todiminish on the average about three days before the crest stage i sreached . This i s , therefore , a useful criterion in j udging how longthe river will continue to rise . The observed rate o f rise can be used

to estimate a stage f o r some time ahead . Thi s is,however

,mostly

an uncertain method , and only to be used where other methods arenot available .

In som e cases,where there is only a single gauge o n a river and

the drainage area above it i s small,the rel iance in m aking predic

tions must be wholly on observations o f the depth of rainfall over

the area . Definite stage predictions are out o f the question in suchcases

,and the most that can be said is that a very high stage will

prevail when the rainfall over the area i s seen to exceed a certainam ount f o r the average o f a number o f stations .As an exam ple o f thi s it requires , at the least , a rainfall Of 3 inches

in less than three days over the square miles of drainage area

o f the Potomac River'

t o rai se the stage at Harpers Ferry to 34 feet ,which corresponds to 12 feet at Washington

,D . C .

,twelve hours later .

Over the Savannah River drainage area of square miles itrequires a rainfal l o f 5 inches in three days to cause the river to riseat Augusta

,Ga .

,to the highest stage known

,which is feet .

4 .- M ETHODS IN USE IN FRANCE IN FORECASTING FLOODS .

M . BABINET.

His t o r ica l — In requesting M . Georges Lemoine,In g e n i e u r e n ch ef d e s

Po n ts e t Ch a u ssée s,Paris

,to present a paper o n the methods in

'

use

in France in forecasting floods,for the International Congress of

Meteorology to meet at Chicago in August,1893

,the Honorable Chair

man o f the Section devoted to Rivers and Floods kindly remarks thatthe idea o f predicting the level of rivers originated in France

,and

that questions of thi s nature have been treated there with more carethan elsewhere .

It w a s in 1854 that the il lustrious Be l g r a n d organized a network o f

permanent hydrom etric observations in the basin of the Seine . He

derived , a short tim e after , a prelim inary rule for the forecasting o fflood s at Paris ; about the same time similar investigations yieldedappreciable results o n other French rivers better naturally disposed ,without doubt

,to facilitate the forecasts . M . Ca m oy tried to predict

the floods o f the Loire at Orleans and Tours from observations madeat points above and below as far as the confluence of the Allier

,below

which the Loire receives no im portant affluent throughout 30 0 kilometers o f its length . M . Poincaré did the sam e f o r the Meuse , the



region in question correspond to a single continuous great rise of theSeine in the vicinity o f Paris ; it is a multiple wave o f which themaximum does not depend alone on the highest stages prevailing atthe points above . The successive wave s unite below ,

where theirvelocity slackens

,and where there i s thus a very great accumulation

of water . From this has arisen the practice,for the basin o f the

Seine,of considering the relation o f rises

'

Of which the name alone isscarcely a definition . Even in the case o f simple waves , produced bya single group of rains

,the rise ( o r the difference between the level

of the water at the beginning of the rise and when t h e highest pointis reached ) has the advantage of being independent of the stagewhence it starts

,if it i s not artificially influenced by a movable dam

which is finally lowered .

Certain rises m ay, moreover , be taken as representative signs orevidences of the hydrological phenom ena o f which a neighboring

region is the theater . It i s in this way that in the rule for a n n o u ncing floods in the Seine at Paris Be l g r a n d was able to m ake use o f theAi sne at Ste . Menehould and the A ire at Vraincourt, even though thewaters o f these two rivers run into the Oi se and have not , co n s equently

,any actual influence o n the stage as read from the gauge on

the Austerl itz bridge . For a similar reason , in order to take intoaccount the great superficial extent “ of the basin of the Marne

,in

place of double the ri se taken at a single point o f the river co n ve nie n t l y chosen , the formula for prediction

‘

at Paris contains the riseof the Marne at Chaumont and St . Dizier

,o n e of which precedes the

other except for the changes due to the intermediate tributaries .Whatever may be thought o f the principle of this method

,it i s in

any case j ustifiable by the excellence of the results ; f o r the threegreat floods at Paris have been predicted one o r two days in advancewithin a few centimeters o f the correct stage

,notably those of March

,

1876 , and February , 1889 . The same processes have been ‘ employed

elsewhere by M . G . Lemoine in predicting the floods of the principaltributaries o f the Seine

,as m ay be seen in the M a n u a l Hydr o l og iqu e

mentioned above (pp . 51

The prediction of floods by rise is moreover well adapted for taking into account certain necessary corrections due to accessory influe n ce s it i s susceptible of improvement . As

,for instance

,when at a

station for which predictions are made a flood occurs when the stageis fal l ing, that i s t o s ay, fo r a sufficient length of tim e before theriver has returned to the normal level o f the season

,the rise calculate d

by the ordinary formula ought generally to be r e du ce d'

in a certainproportion ; a part o f the water is used in a manner t o overcome thetendency toward lowering

,or is absorbed by the drawing eff ect of the

preceding movem ent .The rise to be predicted at a lower sta tion may moreover be a dis

FORECASTING FLOODS. 97

continuous function of those at points above ; it i s therefore highly‘ probable that when the river rises above the level at which in anumber o f places the wetted perimeter o f a cross-section increasessharply for a slight increase o f height flood level s ) these anom aliesare less appreciable than at principal stations

,where the phenom ena

occurring in a great basin proceed more regularly ; for reasons of thesame nature they are less to be feared than more important floods

.

If the announcement Of slight changes is o f any interest on secondclass rivers , floods m ay be classified by families according to the initial stages o r the magnitudes of the rises and a special formula maybe used for each kind . This has been tried recently for som e stationsin the basin of the Oise .

An n o u n cem e n t of flo o ds by a bso lu te .s ta g e s .

— The study of stages , veryextensive in the basin of the Seine

,o f which the hydrologic compli

cation is sufficiently great,is not s o general ly in use o n the other

French rivers where the situation is d1ff e r e n t .

On the Seine itself where , o n the tributaries , without giving up th emethod of predicting by rises which generally permit o f making

predictions a sufficient length of tim e ahead,i t has been possible

in the last fifteen o r twenty years , especially , to utilize the extension of the telegraph for obtaining in time information as to thestages occurring successively at upper stations s o as to m ake themof use for predictions at places lower down

,as Paris or Mantes

,and

to draw from them conclusions useful for points along the lowercourse o f the river . Inspector General Allard

,former President of

the Com mission on the Forecasting of Floods in the Ministry o f

Public Works,has given

,in a special work,a certain number o f second

ary rules determined in this manner (An n a l e s d e s Po n ts e t Ch a u ssée s ,1889

,l er sem .

,vo l . XVII

,pp . 689 and following ) .

When a sufficiently great distance separates two stations betweenwhich th e course of the water considered does n o t receive any important tributary

,o r if the velocity Of propagation o f floods between

these two points i s small,the prediction can be made by the aid of a

graphical process in which 1) the abscissas are the m aximum stagesoccurring at the upper station in a certain number o f previous floods ,(2 ) the ordinates are the m axima corresponding t o the lower station .

The extrem ities of the ordinates give generally a regular curve whichpermits of finding the highest stage to be predicted from the corresponding abscissa

,the stage at the upper station .

.Ah analogous graphic method was proposed in 1882 by M . La vo in n e,

in a more complicated case,to study the relations between the m axima

o f the Seine at Rouen,that of the Seine at Mantes , and the level o f

the sea at Havre about thirty-six hours in advance . The stage at

Havre w a s taken as abscissa and that at Mantes as ordinate , and alongside o f the point thus located was written the stage that resulted at


Rouen . If there was a constant relation between theseall the points o f equal stage ough t to be o n a regular

j e ct io n o f a line o f level o n a surface conceived to grelation in question . Thi s p r o ce e di

fo r the prediction o f floods o f the

(An n a l e s d e s Po n ts a n d Ch a u ssée s , 1

i s modified somewhat : one o f

height of water actually observed o n

the mean o f the maxim um stages incertain num ber o f tributaries wh ose relative ininto account by the aid o f proper coefficients .

In the tw o cases in question , the graphical processes can be repl

bytables of single or double entry

,as has been done by M . Jo l l o i

the upper Loire . (An n a l e s de s Po n ts e t Ch a u ssée s , 1881, 1e r sem .,

I,pp . 2 73 toIt seems useless t o dwel l here on the investigation o f formulas

unknown co eflicie n t s (by trial or by the method of linfin e , o n the graphic representation of the relationbers by means o f the processes indicated in

d ’Oca g n e (Gauthier Villars , 1892 , pp . 65 to T

in particular,i s n o t yet completely studied , and i

cation in hydrometry can not be pronounced upon immethese investigations the important point i s to be s a t isficonditions indicated for great floods ; it i s rarely that thean equal interest for dwellers along rivers .Pr e dictio n of flo o ds fr om r a in s — At the present tim e

speak affirmatively as to the possibil ity of effectivelytions o f rainfall for predicting the level of the water icertain regions ; they have served asbut do not appear to be easily usable . (M a n u a l Hydr

de l a Se in e, p .

In very impermeable basins with small extent o fslopes the detail s o f observations o f rainfall permimore readily than el seflow . It i s precisely incult to procure in timealong the upper courses

,which perm it

the nature and configuration o f the ground,its dryness

,its

ature , o r o ther perturbing action .

In studying the floods Of a very little river in the north o f

( the Liane ) which empties in the straits o f Calais at Boulogne-surMer , M . Voisin has shown that

,in certain cases at least

,having r e

gard to these influences,o n e can

,without passing through the inter

mediary of river gauges at upper stations,give approximate estimates

of stages in time to be of u se . (An n a l e s de s Po n ts e t Ch a u ssée s , 1888,



water courses have only been made at a small number Of places andhave n o relation to each other and are , in general , only sl ightly comparable . To extract anything from data of this kind

,one is especially

embarrassed by the manoeuvering of movable dam s on the numerousrivers where navigation has been improved by artificial m eans . Final ly ,there does not exist o n any French water course anything that canbe compared to the great work done for the E lbe and its tributariesin Bohem ia under the direction o f Prof. H a r l a ch e r of the PolytechnicInstitute at Prague . It was only after eighteen years of patient inve s t ig a t io n that he succeeded , in 1881, in p eprediction by discharges

,published over his

Prof . Richter in the m onth of December,1886

There is often difficulty in choosingalong head waters

,such that the sum of their

a certain proportion to take into account th eary streams below

,produce with sufficient

in,

a :

determ inate time at the station for which the predictionsare to be m ade . If this correspondence is attained , might not onefind an approximate relation by a sim pler application of the abs o

lute stages o r the rises,without passing through the interm ediary

of the gaugings? To obtain the greatest chances Of success,the dis

charges ought to be well determ ined in connection with the observation o f levels at upper points ; there ought t o be a great va r ia t ioheight without any notable change o f volume in the wat er pasper second ; this supposes that the valleys are embanked . Atsam e time an appreciable error in the discharge ought not t o in vat the lower station any great uncertainty in the corresponding heo f water ; . this case is presen ted only in a flat valley with a large

,

broad bed . These conditions are n o t found together in France , wherethe water courses have fo r the most part a r eg im e n too variable topermit of the m ethod in question being applied advantageously .

GENERAL ORGANIZATION FOR THE PREDICTION OF FLOOD S IN FRANCE.

Co n clu sio n — In what precedes there has been no question as to thedissemination of warnin gs

,for transmission i s n o t directly bound up

with the technical work of prediction . Scientific progress cannot

obtained by decree,adm itting that in many other cases it i s possi

realizable .

The important results obtained byBe l g r a n d for the basin ofSeine preceded all corresponding administrative organthe investigations to which local services ought to devotegies in order to m ake useful predictions are often long anThe persons charged with them give them special attentionexpect to find a way to derive satisfactory rules through sk

bin a t io n s,and when the labor and ingenuity best

RIVERS OF SIBERIA . 10 1

do not risk passing unnoticed . The Co m mission on Prediction ofFloods

,instituted in 1875 under the Ministry o f Public Works at

Paris,has certainly played from this point o f view a very im portant

rOl e . It has organized in a permanent , definite manner the serviceof Observation

,the preparation o f warnings

,and their distribution

throughout almost the whole of France . The results obtained in

France up to the present time have been attained without very greatmeans which have recently per

o r o l ogi e u n d Hydr o g r ap h ic of the

e the recent magnificent publica

tion o n the Rhine and it s a ffiu e n t s . With more modest resources,the

h ydrometric services o f the different French basins have probablynot yet said the last word .

5 — THE FOUR GREAT RIVERS OF S IBERIA .

FRANZ OTTO SPERR.

The whole northern portion o f Asia that bears the name o f Siberia,

extending from the Ural Mountains in the west t o the waters o f thePacific Ocean in the east , i . e .

,approximately from the fiftieth to the

o n e hundred and fortieth m eridians east o f Greenwich,covers

(according t o the computation o f Mr . St r e l bi t sky) an area ofsquare miles

,o r square kilometers

,not including the i slands .

This vast territory i s in tersected by four enormous river basins .Three of these rivers

,the Obi , the Yenisei , and the Lena , receive their

waters from the Altai and Saian mountains and their numerousspurs

,and empty into the Arctic Ocean ; while the fourth , the great

Amoor River,discharges its waters into the Gulf o f Tartary of the

Pacific Ocean . I shall not speak of the rivers o f the extreme northof Siberia

,such as the Pia sin a

,Khatanga

,Olenek

,Yana

,In d igh irka ,

Kolyma,An a d e e r , a n d others , although some o f these are very large .

The dimensions o f the four principal river systems o f Siberia are

as follows : (1) The river Obi has a length o f miles,o r

kilom eters ; its basin com prises an area o f square

miles,o r square kilom eters . (2 ) The river Yenisei , with

the Angara,Lake Baikal

,and the Upper Angara River , has a length

o f miles,or kilom eters ; the basin extends over an area

of square miles,o r square kilometers . If

,how

ever,the Selenga River is taken as the beginning of the Angara ,

both the length and the area o f the basin would be considerablyincreased . (3 ) The Lena has a length of miles , o r kilo

meters ; the area o f its basin i s square miles , o r

!Th e Pr u ssia n m ile s e e m s t o be u s e d t h r o u gh o u t t h is p a pe r . Th e En glish e qu iva

le n t is sta t u t e m ile s . Th e t e m p e r a t u r e s give n a r e in ce n t ig r a de de g r e e s— EDITOR.


square kilom eters . Fin a l ly,r (4 ) the Amoor River is 60 3 miles , orkilometers

,long ; and the portion of its basin belon ging to the

Russian Empire covers an area of square miles,o r

square kilometers .There can be no doubt as to the great importance o f these rivers fo r

the climate o f the country . Thus the annual covering of these riverswith ice and their delivery from the ice must exert a powerful influence o n the temperature o f

’

the locality,for the formation o f the ice

absorbs a large amount of heat,and the cover o f ice over th e rivers

changes the conditions o f evaporation and radiation ; again , the thawing o f the ice appreciably lowers the air temperature of the springm onths . Thus

,while the rivers are

,so to speak

,a product of th e

climate of the country,they may themselves serve

,apart from direct

m eteorological observations,as an indication of the greater or less

am ount of precipitation,and their -changes o f level may allow infer

ch e es as to the annual distribution Of the precipitation . The existence of the m ighty rivers o f Siberia , with their numerous powerfuland extensive tributaries

,shows clearly that Siberia i s n o t deficient

in precipitation ; and the considerable changes of level taking placein these innum erable rivers a t diffe r e n t tim e s indicate that this ratherlarge amount of precipitation i s not uniform ly distributed over thevarious parts of Siberia and throu gh the seasons . In general

,pre

cip it a t io n is greater in the west in sum mer , in.

the east in winter .

Unfortunately,the questions that m eteorology m ight raise concerning

the Siberian rivers have,as yet , hardly been seriously considered .

Besides the work o f Dr . Ryka t sch e w“On the Opening and Freezing

of Rivers,

” and that o f Dr . Stel ling On the Di scharge of Water o fthe Angara

,

” some scanty material,not yet scientifically elaborated

,

is to be found in various works concernin g th e discharge of rivers and

allied problems .

The present brief sketch is based on all the material I have beenable to collect

,and o n my personal observations and recollections

from a n

'

eighteen years ’ residence in Siberia .

I shall begin with the freezing and Opening of the rivers,these

phenom ena accompanying the transition from summ er to winter,and

again from the cold t o the warm season . While the times o f thefreezing and opening o f the rivers vary within rather wide limits ,these lim its are more narrow in Siberia than in European Russia .

I give below a t able derived from the resul ts of Dr . Ryka t sch ew ,and

exhibiting the m ean and maximum deviations from the general meansfor the separate periods



S iberian rivers begins about the 2 l st o f April in the southern portionOf west S iberia

,between the cities o f Semipalatinsk and Barnaul .

From there it moves to the upper part o f the Yenisei,beginning later

toward the east,o r

,i f simultaneous

,farther south . About the 1s t

o f May the opening o f the rivers already extends to the m iddle of thecourse o f the Amoor . On the 2 1s t o f May the opening , beginning in thewest near Berezov

,extends eastward

,bends at Yakutsk somewhat t o

the south,and turns o n the shores of the Okhotsk Sea abruptly south

west towards Niko l a if sk on the Am oor . The Yenisei , in the lower

part o f its course , between To o r o okh a n sk and the m outh of the river ,throws off its coverin g o f ice only at the beginning of June ; the sameis true o f the rivers Yana and Kolym a in their lower course . Toward

the end o f June the mouth o f the Lena becom es Open,and partly that

o f the Yenisei . But even in the month o f July some smaller rivers

can be found covered with ice o n the Ta im yr Peninsula . Accordingto the m eans deduced from the Observations

,the opening o f the

more im portant Siberian rivers takes place sixteen days after theoccurrence o f a mean temperature of the air o f zero ; for the smallerrivers the in terval is twelve days .

The gradual onward motion.o f the boundary of the ice covering

from the south to the north goes o n more rapidly in the east . In

eastern Siberia,in N . it takes place in a month

,the motion

northward being at the rate of 10 ° in nineteen days,but o n o n e and

the sam e parallel the opening is considerably retarded toward theeast in the more southern parts o f the country in comparison withthe northern parts

,and it i s evidently connected with the direction

o f the isothermal lines and depends on local topograph ical conditions .

The covering o f the rivers with ice,i . e .

,the process of the form a

tion o f the ice o n the rivers,is very interesting in Siberia

,but

,

unfortunately , it has been investigated very little . It is known from

the investigations of Messrs . Sh ch ukin and Schwarz that the formation o f ice in the Angara

,and al so in some other rivers

,for instance

in the Ol okm a, takes place not only on the surface but at the bottom

a s well .

Special mention should here be made of the Angara,as this river

i s distinguished from other Siberi an rivers by several peculiarities .

Forcing it s way through the rocky shores of Lake Baikal , the Angaracarries i t s waters

,cold and clear as crystal

,in rapid flow from the

lake t o the city o f Irkutsk ; and on this distance of only about 70kilom eters it s bed has a fall of 30 meters

,in other words

,it has an

average grade o f meters per kilometer . A long this whole di stance the Angara does n o t receive a single im portant tributary

,and

represents a pure type of a lake river . From this it might be ex

p e ct e d that it s level would vary but‘

slightly . It appears,however

,

as we shal l s e e later , that the height o f its water level i s subj ect to


very considerable and abrupt variations . Another peculiarity o f the

Angara is it s late freezing, which takes place about eighty days afterthe beginning of frosts

,at a time when the cold reaches n o t less than

it occurs m ore than a month and a half later than in otherSiberian rivers situated in the sam e latitude . A third peculiarity o f

the Angara lies in the fact that it s overflow takes place not in spring,

sum mer,o r fall

,as i s the case with other rivers

,but in win ter

,at a

time o f the s e ve r e s t s frosts , wh en the river is freezing . Dr . Stellingbelieves that the overflow o f the Angara at the tim e o f its freezing i sdue partly to the dim inution of the velocity o f the current arising

from the friction o f the wate r o n the ice crust,and partly t o the nar

rowing o f its bed through the ice . The latter cause i s probably theprincipal o n e . It i s to be regretted that no exact data are availableas to the distance over which the Angara below Irkutsk is coveredwith ice at an earlier period than at this city . The obstruction byice in these portions o f river

,which freeze at an earlier time

,are the

main cause o f the overflows .

I shal l n o w give a brief extract from Dr . Schwarz ’ s observations o nthe freezing of the Angara at Irkutsk in the ‘winter o f 1856—1857 :On Decem ber 14

"

ice began f o rm in g ,a l o n g the banks ; o n the 18th ,

when the tem perature of the air fel l to while that o f the

water was +0 0 3 , the whole river began to be covered with ice flo e s ,and the ice o n the banks extended to a considerable distance into theriver . At the bottom o f the river could be noticed num erous ice

crystal s which would,from tim e to tim e

,break loose from the bottom

and rise to the surface . On December 19 , the tem perature of the air

bein g the ice crystals at the bottom disappeared , and the rivercontinued to carry ice flo e s . Beginning with January 15 , 1857 , the

water began t o rise in the river,the temperature of the air being

On the 18t h the river o ve rflo we d all the l o w bank near thecity ; o n the 19 t h the water rose t o a height o f 3 m eters , and o n thesam e day

,the temperature being the m ain channel o f the

Angara became completely covered with an ice sheet . Nevertheless ,the overflow o f the river continued to increase up to January 24 , andonly o n the 25t h the water began to fall . It must also be noticedthat the fal l of the water

,after the overflow h a s reached it s greatest

height,takes place very gradual ly

,the water continuing to fal l for a

m onth . Thus it appears from the observations on the freezing o fthe Angara that ice crystal s form at the bottom , that this goes on ,

with interruptions,in spite o f the increasing cold , and that these

crystals rise to the surface in the form o f laminas which freeze o n t o

the ice forming on the surface .

I add the following details as t o the tim e o f freezing o f the AngaraIn the year 1739 the river becam e completely covered with ice onJanuary 9

,simultaneously with Lake Baikal . This i s a rare occur


rence,as the Baikal usually freezes earl ier . In 1751 there wa s a very

heavy inundation at the time'

o i freezing,January 8. In the winter

o f the year 1755 the river began being covered with ice only on Feb

r u a ry 2 ; the ice , however , wa s carried away seven times , and there

w a s hardly any period o f complete covering with ice . In the year

1870 , when the river became covered with ice on January 15 , a large

part o f the city o f Irkutsk and m any settlem ents along the Angarawere inundated . The same thing occurred in 1887 , when the riverfroze on January 18 a n d 19

,and the overflowing waters carried large

masses o f ice with them . It i s al so worthy o f notice that,from the

very beginning o f frost,i . e .

,from the month of October

,a heavy fog

w a s constantly hovering over the Angara River . This fog disappeared

only when the river becam e covered with ice .

Returning again to the results obtained by Dr . Ryka t sch ew ,we find

that the covering of the rivers with ice is subj ect to gr eater variationsthan the Opening

,and that i t proceeds in the Opposi te order , i . e .

,from

the northeast to the southwest . First o f all becom e covered with ice

the small rivers o f the Ta im yr Peninsul a , as early as in Septem ber .

Next,about two weeks later

,such rather considerable rivers as the

Pia s in a , In d igh irka , Yana , em ptying into the Arctic Ocean , begin t ofreeze alm ost simultaneously

,and the bou ndary of the ice covering

advances pretty rapidly,forming two bends toward the south and

southwest,one in eastern Siberia along the Amoor the other in west

Siberia toward the upper course o f the rivers To m and Om i ; an u p

ward bend occurs along the valley o f the Yenisei .

The form ation of these bends in the progress o f the ice sheet o n therivers depends both on the distribution of the air te m peratures ando n the slow cooling o f the large mass o f water in the rivers . To the

same cause is due the late freezing o f the northern parts o f the otherlarge rivers

,viz .

,the Obi and the Lena

,In its progress from east to

west the covering o f the rivers with ice i s retarded about ten days f o r

every 24° o f longitude,while the southward march o f the boundary

o f the ice sheet,from the polar circle to the fiftieth parallel

,i s a c

com plished,o n an average

,in thirty-one days .

As regards the duration o f the ice covering,i t varies between very

wide lim its . Thus,at the mouth o f the Pia s in a the ice sta ys three

hundred days,while in southern Siberia the duration i s not over o n e

h undred and sixty days . An exception is m ade by the Angara,which

,

after leaving the Baikal,for a distance of -7 kilometers , never freezes

at all , owing t o the rapidity of i t s current , and this in spite o f tem

p e r a t u r e s o f and o f the fact that f o r a period of one hundred and

seventy days the temperature always rem ains below zero . The same

phenomenon occurs in the course o f the Yenisei,in the narrow rocky

passes o f the Saian Mountains,and in a large number o f small moun

tain rivers , which in some parts do not freeze at all , owing either t o



those for the rivers o f European Russia and o f west Siberia . Thus ,in 1887 , the mean level of the Angara at Irkutsk ( in meters ) was as

followsJanuary

,February

,March

,April

,May

,

June,

July,

August,

September,

October,

November,

December,These figures indicate by how many meters the water level of the

Angara wa s below the mark established at the entrance to theMuseum ; they are daily means from three-hour observations taken

7 a . m .

,1 p . m .

,and 7 p . m .

The daily observations showed that,beginning with the principal

maxim um which occurs toward the end o f January,the wa ter level

o f the river fall s pretty uniformly to th e second half o f March ;then

,at the opening of the river

,which takes place very rapidly and

without overflow,a stil l greate r fall occurs ; but from the beginning

o f April to the beginning of June the water rises slowly . In July

the rise becom es more pronounced,and in September the level

reaches a second smaller maximum,arising from the great amount

of precipitation in the Baikal r egion . At the expirat ion of the rainyperiod the water level of the Angara begins to fall slightly

,continuing

to do so with sl ight oscillations until the period o f freezing . It thusappears that in spring

,i . e . ,

at the time o f the greatest overflows ofthe rivers of Russia and west Siberia

,the height o f the water in the

Angara is usually least . This is largely due to the sm a l l a m o u n t ofsn ow in the Trans-Baikal and o n the mountains near Lake Baikal

,

which supply the a ffiu e n t s o f this lake with water ; al so to th e s l owth a win g of th e sn ow during the cold and dry spring, and to th e h e a vywin ds which produce intensified evaporation accompanied by dryness

o f the a ir . On the other hand,during the latter part o f the sum mer

and the beginning o f fall,when in Russia everyb ody complains of a

lack o f water,the Trans-Baikal cdu n t ry i s visited by frequent rains .

This abundance of precipitation is due to the monsoon which,in

som e years,extends into a portion of the Province (Gu be rn ia ) o f

Irkutsk . Al l the rivers then begin to rise,not excepting even the

Selenga , if we m ay j udge from the scanty information obtainable f o rthis river . Even the level of Lake Baikal

,in spite o f the e nbrm o u s

extent of it s surface will,in some years

,rise appreciably .

Of the changes o f level of other Siberian rivers,and o f the times of

greatest height o f the water,w e can j udge only from the available

data a s t o overflows and inundations caused by such overflows . Thus,

beginning in west Siberia,in the basin of the Obi

,the overflows of

the rivers are usually observed in the spring . What contributes mostto the intensity of the overflow is the early opening Of the rivers

,the

great am ount Of the snowfal l in winter,in particular if the snow fall s

o n a previously frozen soil,the rapid approach o f warm weather , or ,


what i s called , a kind spring,” which causes the simultaneous open

ing o f many rivers . In the east,o n the other hand

,owing to the long

and uninterrupted prevalence o f the anticyclone in winter tim e,the

winters are marked by an exceedingly small am ount o f snowfall,

while in summer , during the reign o f the summ er monsoon whichcarries moisture from the Pacific Ocean , there is abundant rainfall ,causing heavy overflows o f the rivers

,especial ly in the Am oor country .

It is,however

,not yet decided how far the rains caused by winds

from the Pacific extend into the interior o f eastern Siberia . But theheavy inundations occurring sometimes in the Province of Irkutsk insummer would seem to indicate that the influence o f the m onsoonoccasionally reaches thi s territory .

To characterize the distribution o f the precipitation over Siberia,

I give the fol lowing results derived by Dr . Wo e iko f for the m ean

precipitation as percentage of the total annual amount

P l ace .

Se m ipa la t in s kBa rn a u l

Ne rch in s kNico l a ifsk o n t h e Am o o r

It appears from these data that the amount of precipitation i s distributed over the year more uniformly in western Siberia than ineastern Siberia . The lack o f uniform ity is particularly striking in

the Trans-Baikal country .

For the Province of Irkutsk we have observations f o r a considerable period at the city of Irkutsk ; these give for the annual d is t r ibution o f the precipitation ( in m il l imeters ) the following figures : Ja n

uary,

February,

March,

April , May ,June

,July

,August

,September , October ,

November,

Decem ber ,But from year to year the m onthly amount o f moisture varies very

much . Thus,in June the greatest m ean precipitation was mm .

( in in'

July,

m m . ( in in August , m m . ( inwhile in win ter the m aximum precipitation reached only

mm . in January , 1882 . But there are years in which the precipitation for January and February is equal to zero .

The mean amount o f precipitation at Irkutsk is as follows : Winter ,mm . ; Spring, mm . ; summer , m m . ; autumn ,

mm . ; in th e driest summer there w a s m m . and in thewettest mm . Besides

,the a m o u n t

'

o f precipitation in

a ny given year is distributed very differently over the territory o f


the Province of Irkutsk, as will appear from the following tablefor 1887 :

Place . W in t e r . Spr ing . Su m m e r . Au t u m n .

O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 2 5 . 9

In the Amoor territory the want of uniform ity in the distribution o f

the precipitation in the course'

o f the year is stil l greater . Thus in

1878 the amount in millimeters was as follows

P l a ce . W in t e r . Spr in g . Su m m e r . Au tu m n .

Ne rch in s k M in ing Wo r ksBl ag o ve ch e n skK ha ba r o vkaVla d ivo s t o kIlg a Ha rbo r ( 1880 )

To il lustrate the non-uniformity o f the distribution o f precipitationduring the summer months in different years

,I give the fol lowing

two years

P l ace .

Au g u s t .

0 0 0 0 0

o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o

The abundance o f moisture in the Amoor territory , accompanied atthe same tim e by cloudy weather

,low temperature

,and reduced e va p o

ration,i s the cause o f heavy inundations in sum mer . Although the

ratio o f the am ount o f precipitation in the wettest to that in the drie s t summer month is considerably less at Irkutsk than in the TransBaikal

,yet it i s m uch greater than at Yeniseisk and other localities

situated farther west and agreeing more closely in this respect withRussia .

I n o w proceed to give a brief account of the data concerning inundations availabl e for Siberia . I begin in the west. An unusuallylarge inundation took place in the spring of the year o f 1857 alongthe western a ffiu e n t s of the Obi . It began o n the river Vagai andits tributaries . In

’

the same spring an unusual increase o f water wasalso noticed in the rivers Irtish

,Tobol

,Toora

, Om i , and others . Thus,



rapidity and such incredible power that an oboz (row o f wagons

for transportation ) laden with tea , standing near the bank , did nothave time t o escape and was carried away by the flood . On July 8 a

heavy overflow began al so in the valley of the Biriusa,an affluent of

the Oo d a .

The Angara,as mentioned above , has its overflow near Irkutsk in

winter . This winter high water is an annual phenomenon,though

varying in intensity more o r less from year t o year . There are no

exact data as t o how far down the river this winter overflow exte nds .It i s known

,however , that at the vil lage Br a t sky-Ostrog, Situated o n

the Angara,310 kilometers below Irkutsk , the river overflows at the

tim e of the breaking up of its ice in the spring,and inundates the

settlem ents si tuated on it s banks .

The difference between the summer and winter water level o f LakeBaikal does not generally exceed 1 m eter

,though sometim es it

amounts to as m uch as 3 m eters . During the very rainy su m

m e r o f the year 1869 the waters o f Lake Baikal rose meters above

the usual level . Considering the great dimensions o f thi s lake (aboutsquare kilom eters ) , thi s enormous increase o f water gives an

indication o f the im m ense quantities o f water which the Baikal

must receive from the rivers em ptying into it . According to the Irkutsk records

,there were in that summer twenty-four c l oudy days in

July and twenty-one in August . Now,o n an average

,there cor

responds t o every rainy day the followin g am ount o f precipitationIn Jun e

,m il l imeters (m aximum ,

in July, (m axi

mum,

in August, (maxim um ,

and in September,

(maxim um ,It follows that the rise o f the water level of

the Baikal w a s directly dependent upon the great amount of p r e cip it a t io n received in the surrounding country at the time of the sum

mer monsoon ; th is precipitation being carried into the Baikal bythe numerous rivers emptying into it

,o f which there are as m any as

three hundred and thirty-s ix . Am ong these afli u e n t s there are threeo f considerable size

,viz .

,the Selenga

,the Upper Angara

,and the Bar

g o o ze e n . The only outflow for the waters o f the Baikal i s furnishedby the Angara . The a rea from which th e a ffiu e n t s o f Lake Baikal

col lect their wate rs h a s been com puted as square kilometers .

The river Lena,in i t s upper course

,does not overflow when the ice

breaks ; a small increase of its water level occurs in the second halfo f May

,when the Snow melts in the m ountains . The time o f great

est increase o f water i s usually in the m iddle of July ; it i s due t othe heavy rains occurring at thi s tim e . Destructive inundations are ,however , rarely caused by this high water ; those best known are theinundations o f 1816 and 1864 . In the latter year the water of theLena , near the city of Verkholensk began to rise rapidly o n July11 ; o n the 13th it o ve rflo w e d the ban ks and inundated the m e a

RIVERS OF SIBERIA .

dows,islands

,and the lower end o f the city ; o n the 17th the river

was again confined t o its banks . The whole month of July o f theyear 1864 w a s rather cold for sum m er weather ; the same is true o f themonth of August, which had a m ean temperature o f only o n

September 3,at 5 a . m . , the therm om eter stood at There

were twenty-two rainy days in July ; in August it rained twelve timesand snowed twice

,the water in the Lena increasing again rapidly

from August 8 to 14 .

In the m iddle course of the Lena,in the Kirensk district

,we find

,

besides the sum m er h igh water,overflows at the time of the opening

o f the ice . Thus in 1870 the ice on the Lena near the city of Kire n sk began moving on April 30 ; the river o ve rflowe d its banks andinundated part o f the city ; and when , o n May 1, the ice began also t omove o ri

’ the right-hand affluent,Kirenga

,which empties into the

Lena near the city,the water rose sti l l higher

,doing a great deal of

' damage in the city . The overflow was especially great in the villagesof the Vit im sk district

,situated along the Lena

,below the city o f

Kirensk .

The farther w e go eastward into Siberia the less frequent anddestructive are the spring inundations

,owing to the sm al l amount o f

snow that fall s dur ing the winter months in the Trans-Baikal andthe Amoor te rritory . In the latter country deep snow is found onlyin the lower course o f the Amoor . But

,at the tim e o f the sum m er

monsoon,the abundant precipitation causes heavy summ er inunda

tions . During the years 1855 to 1882 eight great inundations wereobserved in the basin o f the Amoor ; the most destructive of thesewa s the o n e in 1872 , which cam e from the upper course o f the Amoor .

At St r e t e n sk the water level reached i t s greatest height on July 9 , atBl a g o ve ch e n sk o n the 15th and 16th . At the latter place the waterrose 10 meters above the ordinary level , in spite of the fact that theriver valley is there rather Open and the banks of the Amoor , in part icu l a r the left-hand bank below the City , are lowlands . But , in theMaly-Khin gan m ountains the water in the Amoor rose 16 metersabove its mean level . On July 19 the water began to fall nearBl a g o ve ch e n sk. Ou t o f twenty-seven settlements (s ta n i ts a ) , situatedo n the left bank of the upper Am oor , ten were carried away by theflood . A second

,and very considerable

,rise o f

’

the waters of theAmoor occurred in the m onth o f August o f the same year , comingfrom the rivers Argoon and Zeia . Thus on August 9 and 10 thewater of the Amoor rose meters ( above the ordinary level atthe Pokr o vsky settlement ; at the city o f Bl a g o ve ch e n sk the greatestheight o f the water occurred o n August 15 and 16 , but the height

'

reached w a s less than in July .

Heavy floods were al so observed in the river Zeia in 1861, when the

water o f this river rose three tim es in the course o f the sam e sum8


mer— from June 10 to 13 , from July 2 0 to Augu st 2 , and from August

31 to September 10 .

In t h e southern portion o f Us s u rysk territory the heaviest inundations occurred in the years 1861

,1863

,1868

,and 1873 . The overflows

o f the Amoor begin usually after more o r less prolonged rains . Butas the amount o f precipitation is not uniformly distributed , the in

u n da t io n s are most heavy som etimes in'

one,som etimes in another

part o f the Am oor and Us s u rysk territory .

As regards the current velocity and the discharge o f t h e Siberian

rivers,w e have , besides the scientific determination o f these quanti

ties m ade by Dr . Stelling f o r the Angara , some observations on smallerrivers

,vi z .

,the Tobol and the Toora . These Observations were made

with a View t o decide the question as to the sufficiency of the supply

o f water in these rivers for navigation purposes . Observations were

also made on some other small rivers,viz .

,the Ozio r n a ia

,Lo m o va t a ia

,

Ja ze va ia , Maly K a s,and Bolshoi Kas

,in connection with the investi

g a t io n s for the purpose o f connecting the Obi with the Yenisei .The observations o f Dr . Stell ing at Irkutsk showed that the mean

wate r level o f the An gara at this point h a s an elevation o f

meters . He also found the results tabulate d as follows

At t h e Zn a m e n skyCo nve n t , be l o w t h e m o u t h o f t h er ive r Irko o tAt t h e Tr o i t skyFe r ryAt t h e s a m e p l ace , wh e n t h e wa t e r wa s h igh e r , Ju ly

0 0 0 0 0 0 0 0

Sa m e p la ce , u n de r t h e ice

At the Tr o it sky Ferry , where the current velocity and the deptho f the river are less

,while the area o f the cross-section i s greater

,

m ore water flows by than farther down the river , at the Zn a m e n skyConvent . When the water-level i s higher

,the dim ensions o f the

river , its depth , are increased , and so is the velocity o f the currentand the quantity of water flowing by (discharge ) .

It i s to be hoped that,in connection with the proposed railway

line through Siberia,the Ministry of Roads of Com m unication will

at once institute investiga tions o f a similar kind for t h e other important rivers o f Siberia .

The annual discharge of the Angara shows a maximum in the falland a m inimum in the spring .

The am ount o f water carried by the Angara in the year 1887 was ,according to the results obtained by Dr . Stel ling , in cubic kilometers ,as follows


CH ICAGO M ETEOROLOGICAL CONGRESS.116

A

l0

3ova

m

5a

temu

mum

o

fRH

0

9won

QB

0

8m

9w

?anA

a

we

om

DonDo

n2ow

«a32

AN.

9

0

0

o

wm

.

“o

o

8.

80

as

.

Eom

w.

a

o

c

.

wH

.

g

o

o

.

2

80

.

ma

Go

o

.

R.

0

0

0

8seem

.

H

u

.

u

o

o

nu

xx

;

2

52

v

9.

34 .

ww

ha

«

a

o

wo

ofl

g

<

83

54

a

n

A

?

3a

v

20

9emu

«mu

«m:

was

no”

no”

as.

no“

2

$0

2

on

.

>o

Z

.

e

.

>o

Z

.

9.

u

o

o

.

on

.

30

.

S

55

.

.

S.

a

o

o

2.

>o

Z

A

l

.

ma

.

a

o

o

.

m

.

>o

Z

.

Sha

s

.

mu

.

.

w

he

S

.

0«

ha

s

2

me

g

.

u.

.

a<

8

32

.

0m

v

a<

.

own

ae .

.

v

en

s.

.

ou

0en”

m

:Ova

ewe

a«mu

m

?onS

”

we

ova

mm

“NH

mm

on

”

we

ow“

no

mom

OH

Sou

we

n

o

av

n

o

us

n

o

mm

00

“v

mm

x

0

mo

mm

9

on

mm

a

w

H

o

3

3mm

Dan

an

mm

3on

ma

em

5we

on.

mmO

MQMM

Q~OMW

Z

s

o

o

o

o

o

o

o

o

.

xmc®-0®>omfiw—m

amcos

oes

m

a

$552

82

5

MGO

5MQ

>

Mmao

tvm

3

H

EEL

:a

x

G

ET

S.

Mae

gocm

fim

m

5§5<

3

Sach

e

m

8

BO

THE RH INE. 117

6 — REG-IM IEN OF THE RHINE REG-ION : H IGH-WATER PHENOM ENA AND THEIR PRED ICTION.

M . VON TEIN.

The regim en o f a river , as well as the appearance and progress o f

high-water phenomena,it i s known

,bear a close relation to the physi

cal characteristics o f the region,particularly the relief o f the ground

,

the extent o f the drainage area and the clim atic elements dependenton the form er

,especially the distribution of the rainfa l l . A con

sideration of the hydrological phenomena o f a region requires,there

fore,a t first a glance at it s general physical characteristics .

The Rhine,although it drains scarcely an area o f square

kilometers with it s tributaries , extends diagonally through the princip a l parts o f middle Europe , the Alps , the mountain region o f centralGerm any

,and the Netherland lowlands . Its drainage area has there

fore a n uncommonly great variety o f relief . From the chain o f thecentral Alps

,towering to a height o f meters

,form ing the south

ern boundary,the region fall s toward the Swiss and upper Swabian

highland,o f which the average elevation is 50 0 meters

,and ri ses

again to over meters in the Jura,the Black Forest

,and the

Vosges,the most considerable elevations o f which bound it o n the west

and north . Sunk between the Black Forest and the Vosges l ies theupper Rhine lowlands traversed by the Rhine

,while o n the verge o f

the mountain range roundabout spreads extensive terrace and the flatvalleys

,forming the region o f the entrance of the great central moun

tain rivers— the Ne cka r,Main

,and Moselle .

On the north the mountains of the lower Rhine , traversed by theRhine in a deeply eroded val ley

,shut o ff the central m ountains

toward the north German highland to which the lowest parts o f thedrainage area belong . The Rhine flows here in its upper course

through high mountains,a high tableland

,and the central m ountain

range ; in its middle course through a low plain and then , for thefirst time

,after breaking through the great chain of the central

mountain range,begins its lower course with its entrance in to the

north Germ an lowlands .The grouping o f these surfaces forms a many-branching river

system am ong the streams of the Alps,of the central mountain

region , and the lowlands , a distinction which , f o r the regim en of thedrainage area

,is of the greatest significance .

The Rhine assembles the stream s o f the Alpine country and theborders where it leaves the Swiss highland ; here the Alpine part o fthe tributary region ends

,which

,with considerable rainfall

,has also

great capacity for retaining water.The precipitation in the Alps for the greater part of the year being


in a solid form,i s stored up in theWinter and in the warm season is

given up a s mel ted snow more o r less freely , and the run-off is modi

fie d by a considerable number o f lakes around the border of the Alps .The region o f the central mountain rivers begins with the breaking

o f the Rhine through the Jura o r with its entrance into the uppeRhine lowlands at Basel . At first the river receives only sm allstreams

,f o r

,o n both sides parallel t o i t and only a short distance

from it,the Black Forest range and the Vosges turn their steep flanks

toward it,while o n the other side they drain off toward the Neckar

and the Moselle . The Rhine in th is , the second principal part o f itscourse

,traverses a distance of more than 30 0 kilometers before the

gre at central mountain river , the Neckar , empties into it, and then ,in relatively rapid succession in a distance o f about 150 kilometers

by the river,there come into it the Main , the Nahe , the Lahn , and

the Moselle . In the region o f these central mountain range rivers,

the Black Forest range,and the Vosges

,the rainfal l i s stil l very con

s id e r abl e . The snow covering in the course o f the winter,from

repeated thaws and according to its height and density and theaccompanying circum stances o f its melting , occasionally producesgreat floods in the streams in a short tim e . Persistent rain andthunderstorm s often carry. the wet period into m idsummer

,and in

autumn in the high central mountain region the precipitation reachesits maximum ; the feeding out o f water to streams dim inishes veryconsiderably in consequence o f the great loss of water by evaporationand absorption by plants . The third division o f the Rhine drainage

area,com prising the streams o f the lowland from about the mouth

o f the river up to the lower Rhine Mountains , is relatively sm all andnot o f much im portance as regards the water which flows from it intothe Rhine . The region is under the m oderating influence o f the

ocean climate,so that sudden

'

changes of weather and their co n s e

qu e n ce s are scarcely perceptible in the rivers .

The changes in river stage along the Rhine occur as follows alongthe upper course and down to the Neckar the flow from the BlackForest range and the Vosges is

,a s a rule

,n o t considerable enough to

make any appreciable change ; it i s completely controlled by thewater from the high mountain region ; very little water comes induring the win ter ; there i s considerable addition of water in thespring with the disappearance o f the snow o n the lower mountains ;there i s an increased flo w at the beginning of summer a s the meltingadvances up the high mountains ; and finally

,again a diminishing

supply o f water,until toward spring

,in spite o f the continuing addi

tions of glacier water . There i s,therefore

,a tolerably steady increase

from winter,not, however , without this regular course being varied

and broken into at times by heavy floods due to great rainfall suchas often accom panies the FOh n wind . In the section o f the Rhine



the preceding remarks,that the Rhine region , as compared with the

neighboring regions,is naturally protected in a higher degree against

the frequent occurrence o f disastrous high waters,because many Oppo

site conditions m ust act together t o produce such a h igh water .

1 Never

th e l e s s , the occurrence o f such an event in such a closely cultivatedregion a s this i s , means , in every case , a great dam age to agriculturalinterests . Ten years ago , after the mem orable high water in thewinter o f 1882 the German Imperial Government established acommission to investigate the Rhine and consider the best methodsof artificial protection against floods . This com mission reported in

1891 to the Chancellor o f the Empire the results o f its eig ht years’

activity . In the report the fact is emphasized that , in view o f thedense population and m inutely cultivated parts , i t i s o u t o f the ques

tion t o try com bating the might o f high waters by measures calculatedto restrain the waters o n a large scale , aside from the fact of the enorm ous expense of such a proceeding— o u t of all proportion t o the a d

vantages t o be derived— and other vested interests o f the populationo f the region would be thereby greatly dam aged . High water pr o t e c

tion,in addition t o measures of protection by a suitable treatment of

the course o f the river , and by the leveeing o f frontages threatened by

floods,must rem ain l im ited t o deriving a correct knowledge as to

what m ust be withstood from the beginning and throughout the

course o f dangerous rises . Th is knowledge consists , up to the presenttim e

,in a carefully organized service f o r the dissemination o f in f o r

m ation regarding high water , by telegraph principally . What stil lrem ains in this dom ain worth gaining, and which the com m ission in

investigating the Rhine proposes t o attain , i s the num erical d e t e rm i

nation o f water”

heights to be reached during ri ses along the m iddleand low er courses o f the Rhine

,and along the upper course and its

larger tributaries,and al so the establ ishm ent of a system o f high water

predictions . This,for the population interested

,would doubtless be

more valuable than sim ply information a s to the stages o f water at

places on the river above . Considering the extraordinary difficultiesin the w ay o f forecasting

,in view o f the changing regim en o f the river

just described and the com plex phenom enon o f a high water,and the

lack o f previ ous hydrological investigation (which heightens the difficulty ) , high water predictions have n o t yet been attempted . In theyear 1886

,o n the suggestion o f the Im perial Comm ission before men

t io n e d,the Ce n t r a l Bu r e a u fu r M e te o r o l og i e u n d Hydr o l og i e o f Baden

,

l Th e r e g im e n o f th e Rh in e r e g io n a n d th e be h a vio r o f t h e rive r du r in g h ig h w a t e r

is t r e a t e d o f ve ry fu lly in t h e w o rk D e r Rh e in s tr om u n d s e in e wich tig s te n Nebe nf l itss e ,Be rlin , 1889 , issu e d fr o m t h e Ce n tr a l Bu r e a u f u r M e t e o r o l o g i e u n d Hyd r og r ap h ic in

t h e Gra n d Du chy o f Ba de n . Th e h ig h wa t e r s o ccu r r in g in t h is ce n t u ry a r e t r e a t e d

o f ve ry fu lly in th e fir s t vo l u m e o f Er g ebn iss e d e r Un t e r su ch u n g de r Hochwa s s e rve r

hc'

i l tn iss e im D eu tsch e n Rh e in g ebi e t , Be rlin , 1891.

THE NILE. 121

at Carlsruhe , was designated as the proper institution to be entrustedwith the investigation of hydrological phenomena during the in ce ption and progress of great rises in the Rhine and its tributaries

,so

as to lay a firm,scientific foundation

,on which

,perhaps

,satisfactory

high-water predictions may in the future be based .

7 — THE NILE.

W . WILLCOCKS,M . I. C . E.

The recent explorations of Lu g a rd and Baumann have com pletedthe work originated by Burton and carried on by Speke

,Grant

,

Baker , S tanley , Gordon , Junker , and Schweinfurth , and we can nowfol low the course o f the Nile from its springs far south o f theequator to its termination north o f the thirtieth parallel o f latitude .

A river so regular and gentle in its m ovem ents a s the Egyptian Nileca n only be understood after a study of its sources o f supply

,and the

early part of this paper must,therefore

,be devoted to the hydrology o f

the Nile Valley . On the accompanying plan and longitudinal sectionare detailed the observed times and heights of h igh and low supply

,

and the tim es and proportions o f rainfall . From the sea to IVa dyHalfa the Egyptian Irrigation Service has supplied the figures ; fromWady Halfa to Khartoum Sir John Fowler ’ s surveys and levelshave been used

,with a correction of 13 meters t o suit the figures o f

the irrigation departm ent at Wady Halfa ; while t o the south ofKhartoum the distances and h eights have been taken from the obse rva t io n s o f Gordon Pasha ’ s staff when he was governor o f the Soudan .

To Lake Victoria a m ean level has been applied . I take this o pp o r

t u n ity o f acknowledging my thanks to Bo n o l a Bey,the Secretary o f

the Khedivial Geographical Society o f Cairo , for the assistance Ihave received from him . I a m at the present moment engaged in

making a study o f the Nile f o r the Egyptian Governm ent . This

study will n o t be com pleted before the end of Decem ber , but on theinvitation .o f your com m ittee

,with the consent of Mr . Ga r s t in , Under

Secretary of State for Public Works in Egypt, I have collected al lthe inform ation which is at my disposal to-day and have embodiedit in this paper .

Th e Nile drains nearly the whole of northeastern Africa , an iareacomprising square kilometers . It s main tributary , theWhite Nile , has its source to the south of Lake Victoria and has trave r s e d over kilometers before it i s j oined by the Blue Nile atKhartoum . From the j unction onward the river i s known as the

Nile,and after a farther course o f kilometers flows into the

Mediterranean Sea by the Rosetta and Damietta m ouths .

Lake Victoria,covering an area of square kilometers is the


first reservoir o f the N ile . The equator passes through this lake,

which lies in the region o f almost perpetual rains and receives anexcessive supply o f water from its western tributaries , from subsoil

springs,a n d heavy rainfall . Stanley considered the discharge of the

Wh ite Nile as it left Lake Victoria as one-third greater than that ofthe Ta n g o u r ie , the principal affluent o f the lake . Judging from r e

corded observations farther down the river,the mean discharge o f the

l ake is probably 750 cubic meters per second . Shortly after leavingLake Victoria , the White Nile descends the Ripon Falls o n a widthof 40 0 m eters and a drop of 4 meters . Lake Victoria l ies aboutmeters above sea level and is 50 0 meters higher than Lake Albert .

Between these lakes,o n a distance o f 480 kilometers the White Nile

(known here a s the Som erset ) traverses at first the succession o f

swamps known a s the Ibrah imia Lake , and then taking the character

o f a mountain torrent precipitates itself into the southeast corner o fLake Albert . The survey o f Lake Albert

,which has an area o f

square kilom eters , wa s made in 1877 by Mason Bey , and he recorded

the fact that the lake was meters below its high-water level .The rainfall of that year was deficient in the whole o f the Nile Valley

,and the sum m er supply of the Nile wa s the lowest of which there

is any record . In July , 1892 , Capt . Lu g a rd noticed that Lake Victoria w a s 2 m eters above it s normal level after the heavy rains o f thatyear

,and the sum mer supply of the Nile in 1893 i s so high that i t

has only once been ekce e d e d , according to our records . Lake Edward ,with an area of square kilometers and a height above sea level

o f 880 m eters , is a feeder o f Lake Albert . After leaving Lake Albertthe White Nile flows for 2 0 0 kilometers in a deep

,broad arm with

scarcely any slope and scarcely any velocity as far as Du ffl é,and then

after a short,troubled course tosses over the Fola rapids on a width

o f 90 m eters,and continues as a torre nt for another 2 0 0 kilom eters to

a short distance south o f Gondokoro . At Gondokoro the river is 2m eters deep at l o w water

,and only meters deep in flood

,the d is

charge ranging between 50 0 and cubic meters per second . The

regulating eff ect of the great lakes is well fel t here . We are indebted

to Em in Pasha for this inform ation . It i s o n e o f the keys for understanding the flo w o f the Nile

,and wil l be dwelt o n later in this paper .

At Gondokoro the river is at the lowest in winter,it begins to rise

about April 15,and reaches its maximum between August 15 and 30 .

From Gondokoro to Be r,a distance of about 12 0 kilometers

,the

river keeps in o n e channel and has a rapid fall,while from BOr to

the n i o u t h o f the Gazel le River,o n a farther reach o f 380 kilometers

,

the river divides into num erous channel s and has a very feeble slope .The m ain channel i s known as the Bahr e l Gebel ( the m ountain

stream ) , and is the o n e always used f o r navigation . In this reachare the “

s a dds ” or dams of living vegetation which,at times

,are



At a distance Of 90 kilom eters down stream o f Khartoum is thesixth cataract . Here the Nile descends 6 m eters o n a length ofmeters . At a distance o f 32 0 kilometers from Khartoum the Nile i sj oined by the Atbara River . This latter i s another stream fed by th eAbyssinian torrents

,and though dry in summer is a considerable

river in flood . Heavily charged with volcanic detritus it provides thegreater part o f the rich , fertilizing mud which the Nile carries

'

in

flood . The Atbara has a range of 8 meters , and from calculationsand comparisons I estim ate that its floods range between and

cubic meters per second,with a mean high flood o f cubic

meters per second . It i s in flood from July to October,with its ordi

nary maximum ih August . Below the Atbara j unction the Nile hasno tributary

,and flows throughout its kilometers to the sea a

solitary stream . Traversing o n e of the greatest deserts on the globe,

it is the sole source o f l ife and vigor to whatever exists o n its banks .Twenty-four kilometers downstream o f the Atbara j unction i s Ber

ber,and 45 kilom eters downstream of Berber is the beginning o f the

fifth cataract,which has a length of 160 kilometers and drop of 66

meters,with three principal rapids

,the So l im a n ia

,Baggara

,and

M o g r a t . The village o f Abu Hamed is situate d at the foot of thiscataract . Between Abu Ham ed and Dongola is the fourth cataract

,

which begins at a point 10 0 kilometers downstream o f Abu Hamed,

and has a length o f 110 kilom eters,with a drop o f 49 meters . In

this series of rapids are the Um Deras and Gu e r e n did . Between the

fourth and third cataracts is a reach o f 310 kilom eters on a slope o f1 On this reach i s the town o f Dongola . The third catar a ct h a s a length of 70 kilometers and drop o f 11 m eters with theHannek and Kaibar rapids

,surveye d and leveled by De Gottberg in

1857 . Upstream o f the Hannek rapid o n the left bank o f the Nile

i s the termination of the long depression in the deserts,which goes

by the name o f Wady e l Kab,and is considered by many as lower

than the Nile valley . Between the third and second cataracts i s anordinary reach o f 130 kilom eters . West of this part o f the Nile arethe Sel im a Well s

,and

,according to some travellers

,an o l d abandoned

course o f the Nile slightly above the present h igh level of the river .Thi s waterless river terminates in the Oasi s of Be rys , which is separ

ated from the Kh a r g e h Oasis by a l imestone ridge .

The second cataract,known as the “ Batn e l Haggar

,

” has a lengtho f 2 0 0 kilom eters and drop o f 66 m eters with the rapids of Amara

,

Dal, Sem na , and Abka . At Semna are the rocks where Lepsius dis

covered the Nile gauges cut by o n e o f the Pharoahs some years

ago . The N ile flood then was 8 meters higher at this Spot than whatit i s to-day . The erosion which h a s taken place here is very excessive

compared with that between the second and first cataracts and at thefirst cataract . At Wady Halfa

,near the foot o f the second cataract ,

THE NILE. 125

a masonry gauge , divided into meters , h a s been erected and read since1877 . Between the first and second cataracts the Nile has a lengthof 350 kilometers and slope o f 1 The mean width of theriver is 50 0 meters , and the mean depths in flood and sum mer are 9and 2 m eters . The velocity in summ er fal l s to 50 centimeters persecond and rises to 2 m eters per second in flood . The river in thisreach i s generally within sandstone

,and the greater part is provided

with gigantic spurs o n both banks . These spurs perform the doublework of collecting soil o n the sides in flood and training the river in

summ er . They were probably put up by the great Ramesesyears ago

,as som e o f the most massive of them have evidently been

constructed to turn the river o n a curve o u t o f its natural channelon to the opposite side in order to secure deep water in front of thetemple o f Jerf Husain Jerf means steep

,scoured bank ) . The

spurs have been constructed with care,and

,as the courses o f roughly

dressed stone can be examined at fairly l ow water I have never seenthem at absolutely low water ) , it i s evident that there has been nogreat degradation o f the bed during the last or years .

The first,or Assuan

,cataract has a drop o f 5 m eters o n a length of 5

kilometers .From Khartoum to As s u ft n

,on a total length of kilom eters

,

there are 563 kilom eters o f so -called cataracts with a total drop o f20 3 meters

,and kilometers of ordinary channel with a total

drop of 83 metersAt the foot o f the first cataract

,opposite the town o f Assuan , on

the island o f Elephant ine , h a s stood a Nile gauge from very ancienttimes . An officer belonging to the -Rom an garrison in the time o fthe Em peror Severus

,marked an extraordinary high flood o n the

gauge . The maximum flood mark at the tim e of the visit o f Napoleon ’ s French savants was

,however

,meters higher than the above

As the middle of Severus reig n w a s A . D . 2 0 0 , and the visit of the

French savants A . D . 180 0,they concluded that the bed and banks o f

the Nile had risen m eters in years , o r meters per 10 0

years . The new gauge,divided into cubits and twenty-fourths , w a s

erected in 1869 and has been recorded daily since then (a cubit 54

centimeters ) .From Assuan to the Barrage the length o f the river is 970 kilome

ters,and the slope 1 : while the mean fall of the valley is

1 from the Barrages,at the head of the Delta proper , the dis

tance to the sea down either branch is 236 kilom eters , with the sameslope as before

.The mean width of the m ain Nile is 82 0 meters , and

the m ean depth in flood meters . On the Rosetta branch the meanwidth is 50 0 meters and depth 8 meters , while the Damietta branchhas a mean width of 350 meters and mean depth of 7 .5 meters . Themean velocity in floo d is between and meters per second . As


the Nile in these reaches is in soil,it i s evident that

velocity of meters per second scours o u t a channel

is ninety tim es its depth,while a velocity o f over o n e me

has a w idth som e fifty times its depth . The natural canal

take o ff the river and never silt,have a mean velocity o f

centimeters per second while the proportion o f width toabout 12 to 1. Artificial canal s of thi s section do n o t siltvelocities are 70 centim eters per second

,while silting takes

readily when the velocity is greater a s when it is less than thIn muddy stream s

,like the Nile in

'

flo o d,certain velocities dem and

certain proportions of width to depth,and if these are not given t o

i t they wil l m ake it f o r them selves by eating away the sides if they

can,o r by silting up and raising the bed i f they can n o t eat away the

sides .

On Rhoda Island opposite Cairo has stood a gauge from veryancient tim es . It has been frequently reconstructed . The present

gauge was erected in A . D . 861.

-It i s in cubits and half cubits o nsome very arbitrary scale . When the gauge was constructed a read

ing o f 16 cubits on the gauge meant the lowest level at which flood

irrigation could be insured everywhere . In 1887 the Egyptian Gove r n m e n t called upon all the Inspectors of Irrigation to report on them inimum gauge for perfect flood irrigation

,and they reported

cubits o n the gauge . The difference between 16 and cubits o n

the Rhoda Island gauge is m eters,and as years had elapsed

since the construction o f the gauge,it meant a ri se o f m eters

per o n e hundred years . This is slightly under‘ the rise calculated atAssuan , but then the river is muddier at Assuan than at Cairo . The

Rhoda Island gauge has been read since A . D . 641,with interrup

tions,and the following table gives the mean readings p e r century of

maxim um flood and minimum l o w supply

No . o f e a r sYe a r A. D. Fl o o d . Lo w s upp ly .

r e co r e d

110 0 -12 0 0

130 0—140 0

The low-level gauges have been vitiated during the last f e w yearsby regulation at the Barrages . A s the flood and low-level gauges inthe above li st have no accord with o n e another they are probablyincorrect

,l ike all other ancient records in this country . Even in the



M e a n a n n u a l r a inf a l l in th e Ni l e Va l ley.

Sg‘

g gfeg" M e t e r s . Cu bicm e t e r s .

1. 260,0 0 0 x 1. 5 390

,0 0 0

2 . x 2 . 0 0 0 0,0 0 0

s . 60,0 0 0 x 0 0 0

4 . x 1. 5 285,0 0 0

,0 0 0

,0 0 0

5 . 2 20,0 0 0 X 0 . 75 z 165

,0 0 0

,0 0 0

,0 0 0

6 . 340,0 0 0 X 0 . 2 5 z 85

,0 0 0

,0 0 0

,0 0 0

7 . x 2 . 0 0 z 2 60,0 0 0

,0 0 0

,0 0 0

8. 32 0,0 0 0 X 0 . 30 96

,0 0 0

,0 0 0

,0 0 0

9 . 310,0 0 0 X l . 70 r: 52 7 , 0 0 0 , 0 0 0 , 0 0 0

10 . 240,0 0 0 X 1. 40 z 336

,0 0 0

,0 0 0

,0 0 0

11. 910,0 0 0 X 0 . 12 2: 10 9

,0 0 0 , 0 0 0 , 0 0 0

We have next to consider the times o f rainfall . In the great lakeregions the rainy season lasts from March to December

,with a max

imum in Au gust . At Gondokoro the rains continue from April toNovember

,with a maximum in August . In the valley o f the Sa u ba t

the rainy season is from June to Novem ber,with a maximum in

August . It rains from April to Septem ber in the valley of the Gazelle River . From July to September i s the rainy season at Khartoum

,and from July to August in Kordofan and Da r ffir

,In Abys

sinia there are light rains in January and February and heavy rainsfrom the middle o f April t o September

,with a m axim um in August .

August is the center o f heavy rainfal l everywhere .

The time it takes the water to travel down the different lengths ofthe river m ay be found from discharge

,velocity

,and slope calcula

tions,and from comparisons between the fluctuations o f the Ge n

dokoro,Khartoum

,Assuan

,and Cairo gauges . I calculate that it

takes the water eight days to travel from Lake Victoria to Lake

Albert and five days from Lake Albert to Gondokoro . There is n o tmuch d ifle r e n ce between hig h and l o w supply in these reaches . Ittakes the water thirty-six days to traverse the distance between Gondokoro and Khartoum in low supply and twenty days in flood . Be

tween Khartoum and A ssuan the times are twenty-six days in lowsupply and ten days in flood . Between Assuan and Cairo we havetwelve days in low supply and six days in flood

,while between Cairo

and the sea w e have three days and two days,respectively . It takes

eighty-four days for the water in low supply to reach the sea fromLake Victoria

,while in flood it takes fif ty-one days .

The Blue Nile traverses the distance between its sources and Kh artoum in some seventeen days in l o w supply and seven days in flood .

The Atbara takes five days in flood,and the Sa u ba t can not take a

much longer time .

Referring to the map and keeping al l the above facts in mind,an

average year in the Nile Basin may be thus described : The heavy

THE NILE. 12 9

rains near Gondokoro begin in April and force down th e green waterof the swamp regions . About April 15 the White Nile at Gondokorobegins to r ise , and by September 1 has reached its maximum . In

this interval the discharge has risen from 550 cubic meters per secondt o cubic meters per second . This rise is felt at Khartoum aboutMay 2 0

,and at Assuan about June 10 . The green water announcing

this rise i s seen at Cairo about June 2 2 . In an average year o n May

20,the White Nile discharge of 30 0 cubic meters per second at Khar

toum begins to increase,and goes o n gradually increasing to Septem

ber 15 o r 2 0 , when the maximum floods o f the White Nile and Sanbatreach Khartoum and attain a discharge of cubic m eters per

second . The low-water discharge o f the Blue Nile is 160 cubic m etersper second , a n d about June 5 it begins to rise fairly quickly andreaches its ordinary maximum o f cubic m eters per second byabout August 25 . Owing to the two floods rarely being contemporary

the ordinary maximum flood o f cubic meters per second is gen

e r a l ly on September 5 . The red,muddy water of the Blue Nile reaches

Assuan about July 15,and Cairo about July 25 . Once the red water

begins to appear the rise is rapid,f o r the Atba ra i s in flood shortly

after the Blue Nile,and its flood waters rise with great rapidity . The

’

Atbara wou ld come down much earlier th an it does were it not thata whole month is expended in saturating the desert and it s own drysandy bed . The Atbara flood begins in the early part of July and isat its highest about August 2 0 , reaching an ordinary maxim um o f

cubic meters per second,and occasionally an extraordinary

maximum o f cubic meters per second .

It i s owing to the earl iness of the Atbara high flood and the latenesso f the White Nile high flood that the ordinary m aximum dischargeof , the Nile at Assuan is only cubic meters per second . This

is generally on September 5 . When the White Nile is weak the maximum at Assuan is reached before o r o n September 5 ; when theWhite Nile is strong the m axim um i s reached about September 2 0 .

An early maximum at Assuan is always followed by a low sum m er

supply,while a late maximum is nearly always followed by a high

summer supply . Only once has this rule been broken and that wasin 1891

,when there were two m axim ums , one o n September 4 and

another o n the 2 7 th . In this year there must have been an extraordinary fall of rain in Abyssinia in September , f o r the flood o f September 27 was very muddy

,while as a rule the river at Assuan is very

m uddy in August,less so in Septem ber

,and very m uch less so in

October,when the White Nile i s the rul ing factor in the supply o f

the river .

Appendix I contains discharge tables of the Khartoum ,As su it n

,

a n d Cairo gauges . The zero is everywhere mean low-water level .Appendix II g ives five daily gau ges and discharges of the Nile at

9


Khartoum during flood . Appendix III gives five daily gauges anddischarges o f the Nile at A ssuan throughout the year . Appen dix IVgives five daily gauges and discharges o f the N ile at Cairo .

If the White Nile happens to be in very heavy flood late in Sep

tember,and the September rains in Abyssinia are also very heavy ,

an extraordinary flood passes Assuan at the end o f Septem ber and i sdisastrou s for Egypt, This happened in 1878. Appendices III andIV contain detai l s o f this flood , of the minimum flo o d-year

,1877 ,

and the mean of the twenty years from 1873 to 1892 .

At Assuan the Nile enters Egypt,and it now remains to consider

it in its last kilom eters . The ordinary minim um discharge atAssuan i s 470 cubic meters per second , and is reached about the endo f May . The river rises slowly til l about July 2 0

,and then rapidly

through August,reaching its m aximum about September 5

,and then

falling very slowly through October and Novem ber . The tables in

Appendix III give every detail of a maximum,m inimum

,and mean

year . The deep perennial irrigation canals take water al l the year

round,but the flood irrigation canal s are closed with e a r th e r n banks

til l Augu st 15, and are then all opened . These flood canal s, o f which

there a ré som e forty-five,are capable of discharging cubic m ete rs

per second in an ordinary year,and have an im m ediate effect on the d is

charge Of the Nile . The channel of the Nile itself and it s numerousbranches and arms consume a considerable quantity o f water ; theperennial canal s take 2 0 0 cubic met ers per second

,the direct irriga

tion from the Nile between A ssuan and Cairo takes 10 0 cubic m etersper second

,and 10 0 cubic meters per second are lost by evaporation

o fl the Nile . Owing to al l these difle r e n t causes there i s the netresul t that from Augu st 15 to October 1 the Nile is dischargingcubic meters per second less at Cairo than at Assuan . During October and November the flood canals are closed

,and the basins which

have been fil l e d in August a n d September discharge back into theNile

,and from October 5 to November 15 the Nile at Cairo is dis

charging cubic meters per second in excess o f the discharge atAssuan . An examination o f Appendices III and IV will Show thisvery clearly .

The ordinary minimum discharge at Cairo is 370 cubic meters persecond

,and is attained on June 10 ; the river rises slowly through

July,and fairly quickly in August

,and reaches its ordinary maximum

on October 1, when there is no irrigation in the basins and the discharge from the basins is j ust beginning . The ordinary m axim um

discharge at Cairo i s about 7,70 0 cubic meters per second . Through

October the Nile at Cairo is practically stationary , and fal l s rapidlyin Novem ber .North o f Cairo are the heads of the perennial canal s which irrigate

the Delta proper . These canal s , with their feeders lower down , dis



stations in the valley o f the Nile , we ar e always ignorant o f the coming flood

,though famine years in India are generally years of l o w

flood in Egypt . If, however , the summer supply o f the Nile has been

exceedingly low and exceedingly late,we anticipate a high flood fol

lowing it,as the drought in the valley of the White Nile must create

a powerful draugh t o n to the Indian Ocean . Again,as to the summer

supply,we generally anticipate a poor volum e in the river at that

season if the Nile flood at Assuan is an early o n e,and a good sup

ply if the Nile flood at Assuan is a late one . Appendix V,which

contains numerous statistics of tim es and proportions of flood andlow supply for the twenty years from 1873 to 1892 , fully bears o u tthis statement . Between As su it n and Cairo

,previous to 1890

,w e had

little control over the flood,as the canals and escapes in upper Egypt

had n o masonry-regulating works , and the Nile in high flood did verymuch what it l iked . Since 1890

,however

,the Public Works Depart

ment has constructed ninety important regulating works,and by

proper manipulation we can now fairly control a high flood by using

the canals and escapes so as n o t to let the Nile at Cairo rise above 8meters

,which i s the maximum gauge the banks o n the Rosetta and

Damietta branches can support with any degree of security . It was

mainly owing t o thi s power of control that the excessive flood of 1892passed through Egypt without causing any real damage . The Egyptian Governm ent t o -day is very seriously considering the qu estionof flood control and increase o f summer supply

,and we hope to find

a solution for the former by escaping excess flood water in to some Ofthe depressions which border the Nile Valley

,and a solution f o r the

latter by the creation of reservoirs either in the deserts and the channel of the Nile itself north o f Wady Halfa

, o r by regulating works atthe sources o f the rivers themselves

,o r perhaps by a combination of

both .

When we consider the energy and the self-denying labors of themen wh o achieved the great discoveries of the sources o f the Nile

,it

seems but a poor compensation to them to know that these sourcescan now be depicted on the plans . It would be -a trium ph indeed

,and

a real compensation,if the resources o f modern science could be e m

ployed to utili ze these great lakes,and by the construction of suitable

works to insure a constant and plentiful supply of water to the NileValley during the summer months when water is scarce and as valuable as gold . Both the Victoria and the Albert lakes lend themselvesto be util ized as reservoirs as they have rocky sill s at their outlets

,

while the Albert and Tsana lakes,by their convenient size

, a r e emiu ently suited f o r regulating basins . The day these works are carriedout at the sources o f the Nile the lakes will take their proper placein the economy of the wate r supply , and we shall be able to s ay o f

them in their entirety,as we can say of them t o -day in their degree ,



Disch a r g e t a bl e f o r th e Assu an g a u g e— Co n t in u e d .

Ga u g e . Discha rg e . Ga u g e . Discha rg e . Discha rg e : Ga u g e . Discha rg e .

Disch a r g e t abl e f o r th e Ca ir o g a u g e .

!Ga u g e s i n m e t e r s . Discha rge

;in cu bic m e t e r s p e r s e co n d. Ze r o o n t h e g a u e is m e a n l o w

6wa t e r l e ve l o r R. L . g u l a t io n a t t h e Ba r r a g e s vi t ia t e s t h e d ischa rg e s l o w a t t h e

pr e s e n t t im e . ]

Ga u g e . D ischa rg e . Ga u g e . D i scha rg e . Ga u g e . D ischa rg e . Ga u g e . Disch a rg e .

II .— KHARTOUM .

Five daily gauges and discharges for the maxim um , minimum , and

m ean years,from 1873 to 1882 , during flood .

Zero on the gauge i s mean l o w-water level , R . L . meters , or

meters on the old gauge at Khartoum , on the Blue Nile .

THE NILE. 135

Ga u g e s in m e t e rls a n d disch a r g e s in cu bic m e t e r s p e r s eco n d of th e Ni l e a t Kh a r to um

f o r th e m a xim u m , m in im u m , a n d m e a n ye a r s be twe e n 1873 a n d 1882 .

!Ze r o o n t h e g a u g e is m e a n l o w-wa t e r l e ve l .]

M a xim u m , 1878. M in im u m , 1877 . M e a n o f 10 ye a r s .

Da t e .

Ga u g e . Discha rg e . Ga u g e . Discha rg e . Ga u g e .

Au g . I

Se p t . 1

Oct . I

15J u n e 15 t o J u ly 15Ju ly 15 t o Au 0 'u s t 15Au g u s t 15 t o e p t e m be r 15Se p t e m be r 15 t o Oct o be r 15

IH .

— ASSUAN.

Five daily gauges and discharges for the maximum ,m inimum , and

mean years from 1873 to 1892 .

Mean monthly and yearly discharges for the m a x1m u m, minimum ,

and m ean years from 1873 to 1892 .

Zero o n the gauge is mean low-water level , i . e ., R . L . meters

above sea level .

Ga u g e s in m e te r s a n d discha r g e s i n cu bic m e t er s p e r s econd f o r th e m a xim u m , m in i

m u m , a n d m e a n ye a r s be twe en 1873 a n d 1892 .

!Ze r o o n t h e g a u g e is m e a n l o w-wa t e r l e ve l .]

M a xim u m , 1878 M in im u m , 1877 M e a n o f 2 0 ye a r s .

Da t e .

Ga u g e . Discha rg e . Ga u g e . Discha rg e . Ga u g e . Discha rg e .

O O O O O O O O O O O O O O O O O O O O O

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0


Ga u g es in m e ter s a n d disch a r g e s in cu bicm e te r s p er s eco nd , e t a— Co n t in u e d .

M a xim u m , 1878 M in im u m , 1877

Da t e .

Ga u g e . Discha rg e . Ga u g e . Ga u g e .

Au g . 15 .

se pt . 0 0 0 0 0

M a r . 1

Apr il 1

M ay 1

M o n t h .

J u n eJu ly0 0 0 0 0 0 0 0

Au gus e p t e m be rOct o be r . 0 Q 0 O 0 0 0 0 0 0 0 O 0 0 0 0 0 0 0 0

J a n u a ryM a rch 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ‘

M e a n o f t h e



Ga u g es in m e t e r s a n d d isch a r g es in cubicm e te r s p e r s eco n d , e ta— Co n t in u e d .

M a xim u m , 1878 M in im u m , 1877 M e a n o f 2 0 ye a r s .

Da t e .

Ga u g e . Ga u g e . Discha rg e Ga u g e . Discha rg e .

M ay I

M o n t h

M e a n o f t h e ye a r 3 , 9 2 0 I, 9 10

V.

— TABLES AND STATISTIC S .

Miscellaneous tables o f times and heights of high flood and low

water level,disposal o f the water o f the Nile between Assuan and

Cairo,proportion o f rainfall discharged into the s e a

,and approxim ate

quanti ty o f solids discharged into t h e s e a and deposited on the soil of

Egypt in a n average year at the present time

Ta ble g ivin g d a t es a n d h e ig h ts of th e r e a l m in im u m a t Assu an .

!Ze r o o f t h e g a u g e m e a n s m e a n l o w-wa t e r

O O O O O O O O O O O O O O O O O O O O O O O O O O O O O

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

o o o o o o o oO O O O O O O O O O O O O O O O O O O O O O O O O O O O O

O O O O O O O O O O O O O O O O O O O O O O O O O O O O

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

W o r s t l o w s u pp ly.

The river was below the mean low-water level thirteen years , andabove seven years . The m ean o f the m inimum 1s

'

meters .In order to find the date o f real minim um at Cairo add twe lve daysto the above dates . The Cairo gauge at thi s stage o f the river i svalueless as the regulation at the Barrages affects it .Very h igh floods scour o u t the deepest parts o f the river ’ s . bed

,and

a gauge of — O.60 meters in 1878 and 1889 , after the poor floodsOf 1877

M e t er s .

0 . 37 1883 , Ju n e 2 21884 , M ay 2 7

I7 1885, J u n e 20 . I3 I886 , Ju n e 3

1887 , M ay 8

7 1° 1888 , Ju n e 8

1. 88b 1889 , J u n e 2 40 . 82 1890 , Ju n e 8

189I, M a y I90 . 55 I89 2 , Ju n e 18

bBe s t l o w s u pply.

THE NILE. 139

1888, gave a discharge 2 5 per cent less than a gauge ofe t e r s in 1892 after the good and late flood o f 1891.

Ta bl e g ivin g da t e s a n d h e ig h ts of th e m a xim u m flo od l eve ls a t Assu a‘

n .

!Z e r o o n t h e g a u g e m e a n s“m e a n l o w-wa t e r l e ve l .

8M e t e r s .

e p 1883 , Se pt . 1Se pt . 6 1884 , Se pt . I7

Se p . xx 8 36 1885 , Au g . 2 8Se pt . 7 1886 , Se pt . 2 2Au g . 2 0 6 . 4o

u 1887 , Se p t . IOct . I 9 . 15

" 1888, Au g . 2 4Se pt . I3 8 . 59 1889 , S e p t . 2 .

Se pt . 4 189 0 , S e p t . 2Se pt . 4 8 . I4 189 1, Se p t .

Au g . 2 8 I80 2 , Se pt . 2 0Th e po o r e s t flo o d .

bTh e h ig h e s t flo o d .cSe p t . 2 2 , d Se p t . 10 ,

The late and very high flood o f 1892 i s being followed by a sum

mer supply in the Nile nearly as high as that o f 1879 after the veryhigh and late flood o f 1878 .

A mean high flood is meters ; the mean of the maximum is

meters .

Ta bl e g ivin g th e d a t es a n d h e ig h ts of th e m a xim u m flo o d l eve l s a t Ca ir o .

!Ze r o o n t h e g a u g e m e a n s“ m e a n l o w-wa t e r

M e t e r s.

1873 , Se pt .1874, Oct . 6

1875, Oct . 181876 Se p t . 2 71877 Au g . 2 7IS78 Oct . 11

1879 , Oct . I

1880 , Oct . 2 6

ISSI, Oct I31882 , Oct . 2 8 .

a Sho u l d be

In 1874 and 1878 the gauges were incorrectly recorded at Cairo .

Corrections have been applied by calculations from the Barragegauges .The m ost serious flood of the century was that of 1878, but what

it s height might have been at Cairo will'

never be known as the Nile

banks were s wept away on October 11, while the Nile w a s rising.

The mean o f the maximum is a m ean high flood is

Tabl e g ivin g app r oxim a te d a te s o n wh ich m a xim u m a n d r e a l m in im u m g u a g es we r e

r e a ch ed a t Ass u a‘

n .

Nu m be r o f t im e s m in im u m . Nu m be r o f t im e s m a x im u m .

1 Au g . 2 0

Ju n e 1

2 o ct d 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 .

1883 , Oct . 11

1884 , Oct . 2 51885, Oct . 18

1886 , Oct . 41887 , Se p . 2 51888, Se p t . 151889 , Oct . 16

1890 , Oct . 2 5189 1, Oct . 2 51892 , Oct . 7

bSh o u l d be


Ta bl e g ivin g app r oxim a te da t es o n wh ich m axim u m a n d r e a l m in im um g a u g es wer e'

r e a ch ed a t Ca ir o .

Nu m be r o f t im e s m in im u m . Nu m be r o f t im e s m a xim u m .

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 0

The date o n which the real minimum is reached is the last day Oflow supply before the final rise begins ; occasiona l ly the actual minimum

,a few centimeters below the real minimum

,precedes the latter

by many days .

Ca lcu l a tio n exp l a in in g th e co n su mp tio n of wa t er in a n a ve r a g e yea r betwee n Assu an

a n d Ca ir o .

!An Egyp t ia n a cr e e qu a l s squ a r e m e t e r s ]

Eva po r a t io n o f?th e ba sin s Cu bicm e t e r s .

a cr e s X X '

. 0 0 8 da ily X 45 days e qu a ls 2,

0 0 0

Eva po r a t io n o f?t h e Nile it se lfm e t e r s X 70 0 m e t e r s x m e t e r s e qu a ls 1

,330

,0 0 0

,0 0 0

Ir r ig a t io n o f a cr e s o f la n d p e r e n n ia lly ir r ig a t e dX x m e t e r s e qu a ls

Esca pe s dir e ctly in t o th e Ro s e t t a br a n ch .

To t a l e xp e n dit u r e 9,642

,0 0 0

,0 0 0

M e a n disch a rg e p e r day a t Ass u ’

an,fro m App e n dix III

M e a n disch a r g e p e r day a t Ca ir o , fro m App e n dix IV

Ba la n ce spe n t

To t a l qu a n t ity o f w a t e r e xpe n de d be twe e n Assu an a n d Ca ir oin o n e ye a r e qu a ls 365 X 42 0 X a m o u n t in g t o . 13

Th e r e fo r e th e_qu a n t ityOf wa t e r abs o rbe d in t o th e so il p e r a n n u m e qu a l s13 , 245 , ooo , ooo cu bic m e t e r s , a n d

a s th is is a bs o rbe d o ve r a n a r e a o f a cr e s , th e de pt h Of

w a t e r a bso rbe de qu a ls 0 m e t e r s .

Ta bl e s h owin g th e a m o u n t of wa te r wh ich r e a ch es th e s e a in a n a ve r a g e yea r .

Fr o m Appe n dix IV, t h e m e a n disch a rg e a t Ca ir o cu bic m e t e r s p e r se co n d.

Fr o m t h is de du ct th e wa t e r w ith dra w n fr o m th e Nile by t h e De lt a ca n a ls n o r th o f

Ca ir o , vizCm . pe r s e c. Cm . pe r s e c.

Ja n u a ry 80 0 Au g u stFebr u a ry 30 0 Se pt e m be rM a rch 30 0 Oct o be r

30 0 No ve m be r350 De ce m be r40 0

50 0 M e a n fo r th e ye a r



To t a l qu a n tity o f s o lids ca r rie d pa st Assu an in a n a ve r a g e ye a r e qu als365 x x cu bic m e t e r s .

Q u a n t ity o f so lids de po sit e d o n th e s o il o f Egypt e qu a lso r Cu bic m e t e r s p e r a n n u m .

As th e a r e a o ve r wh ich th is is de po sit e d is a cr e s , th e de pth de po sit e d p e r10 0 ye a r s e qu a ls

X

Be fo r e ba sin ir r ig a t io n wa s ch a n g e d in t o p e rin n ia l ir r ig a t io n o ve r two -t h irds th e a r e a

o f Egypt th e m e a n de po sit m u st h a ve be e n co n side r a bly g r e a t e r .

8 .—THE BEST M EANS OF FIND ING RULES FOR PRED ICTING

FLOODS IN WATER COURSES .

M . BABINET.

It does n o t appear to m e possible to exhaust in a few pages the

subj ect proposed to M . Lem oine , In g in i e u r e n Ch ef d e s Po n ts e t

Ch a u ssée s at Paris , in charge o f the Central Hydrom etric Service inthe Basin o f the Seine , by the Honorable President of Section II Ofthe International Congress Of Meteorology , held at Chicago in them onth Of August , 1893 .

Very happily,the task to ‘ be fulfil led i s wel l defined by a very clear

programme,to which we reply with our best eff ort

,but of which the

last two questions apparently ought to be interchanged for c l earn esso f exposition .

Q UESTION I.

Wh a t o u gh t we to p r op o se t o o u r se lve s in th e m a t te r of flo od p r e dicti o n s ;o u gh t p r e dictio n to be g e n e r a l o r sp ecific a s r e g a rds th e s ta tio n s a n d th e

l eve l s wh ich we a r e t o exp ect th e r e 27

It i s n o t very difficult to predict that a flood river i s going to risewhen one has knowledge Of abundant rain s fal len over its basin ; theconsequences are particularly grave in the higher portions Of thecountry when the impermeabil ity o f the soil i s there very marked ;the slopes of the land t here facil itate in every instance the superficialdraining o r deter evaporation o r absorption by the soil . So

,then

,one

may rely o n a few rain gau ges j udiciousl y distributed and connectedwith a central station as sufficiently adequate to organize a system o f

flood predictions .Da t e o f m a xim u m — This first resul t which

,moreover

,must not be

considered as al together negligible,may be improved if we strive to

forewarn the inhabitants of a determined locality,in place of simply

predicting a rising of the level of the water in a whole region .,We

shall soon be led to state precisely the epoch when the level wil l

RULES FOR PREDICTING FLOODS. 143

attain its greatest heigh t at the place considered,from similar phe

nomena’

Observed up the stream and signaled by the telegraph .

The im portance of the principle Of the tim e o f propagation of themaxim um is thus m ade evident .Imp o r ta n ce of th e flo o d — We can n o t admit that a flood has beenthoroughly announced if we content ourselves with indicating m erelythe time o f it s passage at such and such a point without troublingourselves with the level it is to attain . Even when the insufficiency

of previous investigation and verification does n o t allow o f detailed

predictions,if the service is content with sim ple warnings

,there

can generally be added to the word flood a quali fying word,such as

feeble,mean

,o r strong . This is a first approximation ordinarily

realizable everywhere from the recollections of the people o f the

country,o r from the mean o f a sm al l number o f observations .

Nu m e r ica l p r edicti o n s .—_TO g o further and hazard the indication o f

a stage Of determ ined height o n an invariable river gauge,it is

certainly preferable to be able to m ake num erous com parisons between a large number of occurrences of h igh water . However , the

il lustrious Be l g r a n d , founder o f the Hydrometric Service in the Basin

o f the Seine in 1854,whose researches are appreciated tod ay in the

entire world,did n o t take more than two years to establ ish a formula

for predicting three days in advance the total rise of the Seine at

Paris from those o f the upper tributaries , a s shown by eight wellchosen stations toward the l imit o f the m ost dista nt im permeablelands drained . In Spite o f the apparent complication of the basinof the Seine

,where the water courses o f equivalent importance are

numerous with their superficial drainage converging o n the outskirtsof Pari s

,the same principles applied with perseverance by M . G .

Lem oine,pupi l o f Be l g r a n d , and by his co -workers , have allowed of

predicting for the past twenty years the probable levels o f the important floods on the Seine and principal tributaries to within 30 or 40

centimete rs , o r better .Sim ilar studies inspired or not by the sam e principles have

succeeded as well el sewhere . They have been instituted in France

( l st ) on the Loire at Orleans , (2d ) the basin of the Meuse alm ostexactly at the tim e when Be l g r a n d drew from the first results a new

science o f hydrology o f which the announcem en t o f floods i s only anapplication

, (3d ) on the SaOn e , the Garonne , the upper and lowerLoire

,and more recently in several parts o f the basin o f the Rh ine .

They did as much in Italy f o r the basins o f the Arno and the Tiberabou t 1866 , a little later on the E lbe in Bohemia , and finally on the

Ohio and Mississippi in the United States since 1884 . Without co n

s id e r in g the mo st convenient processes for each basin , according toits configuration and the lands that compose it , we can affirm t h e

possibil ity o f predicting floods almost everywhere outside o f the


mountainous regions where they are formed,and where their ravages

are less to be dreaded than in the fertile plains menaced by the over

flow o f wate r .

QUESTION 11.

Wh a t a r r a n g em en t of hydr om e tr ic s ta ti o n s o n th e p r in cip a l r ive r s a n d

o n th e ir tr ibu t a r i e s is th e m o s t a dva n ta g e o u s f o r p r e dictin g l eve ls .5’

The choice of means to be employed for announcing floods,and

particularly numerical predictions o f heights o n certain gauges,de

pends essentially o n the time available for concentrating the in f o rmation as to the stages at points along the upper courses and the timerequired for giving information of their results to points lower down .

The number o f hours depends materially on the rapidity of propagation of the wave

,but the facility of transm i ssion Of the warnings

principally by telegraph plays also an important rdl e and at timespreponderates in the question .

A. Lo n g-tim e fo r eca s ts by m e a n s of up p e r tr ibu ta r ie s

— However per

f e ct may be t h e com munication from o n e poin t t o another alonga water course

,it i s always advantageous t o note the indications as

far up the river as possible . In thi s method we are only limited bythe multiplicity of nearly equal influences of which it is then meces

sary t o take account .

Thirty o r forty years ago the French telegraph system w a s yet

l ittle developed . Warnings were mostly sent by post . In order tohave time to receive data f o r th e problem Be l g r a n d had taken typicalstations in the circumference o f the Seine basin at a convenient distance from the water-parting , approximately o n the arc o f a circle o fwhich Paris i s almost the center . He remarked , moreover , that inthese regions

,as everywhere el se

,the water courses from impermeab l e

land carry Off alm ost all the rain that fall s over their draina ge areas ,consequently

,they determine the beginning and maximum of floods .

In the basin o f the Seine the permeable lands cause the level to besustained for a period of greater or less length

,accordin g t o the

duration o f the rain , after the com plete saturation of the soil ; theirlevel s rise and fall slowly

,and their influence i s more Often negligible

o r very secondary .

B. Sh o r t-tim e f o r eca s ts bym e a n s of seve r a l upp e r obse rva tio n s ta tions .

The Observation stations used fo r the announcement o f floods atParis

,so reduced to eight

,are yet sufficiently numerous for their

influence o n the resul t to vary a little from o n e flood to another,

according as t h e rise i s earl ier o r later at o n e point or the other .Thanks to the actual rapidity o f the telegraphic transmission

,the

indications that one receives thus from the most distant observersserve to establ ish three days in sidva n ce at least a first approximation ,subj ect to correction by subsequent warnings from nearer stations .



twelve hours interval before us after the maximum occu rs up theriver in order to give warnings to points below ; thi s i s even in su fficien t i f the ri se i s rapid and occurs in a night .D . Imp o r ta n ce of th e de t e rm in a tio n of th e m a xim u m .

— S e lf-r eco rdin gapp a r a tu s

— Whatever m ay be the arrangement o f the observing s tations

,according t o circumstances , we can not insist t o o strongly o n

the necessity of knowing exactly the maxim um,its height

,and the

exact m oment it occurs at each gauge in order to m ake precise comparisons . The natural law ,

according to which a variable quantitychanges slowly in value in the neighborhood o f a maximum or a

minimum,and more rapidly in every other case during the interval s

between them prevents making any very great errors in this matterin tranquil rivers whose regim en i s perm anent .The difficul ties are much greater on certain torrential a fflue n t s where

the observations at fixed hours , however close together they may be in

times of flood,allow the m ost interesting heights to be missed and also

the moment of their occurrence . In order to guard against thi s in co n

ve n ie n ce,self-registering apparatus o f various kinds is beginning to be

be employed in France,not only in Pari s and its outskirts

,but partie n

l a r l y in the basin o f the Durance , th e principal aflfl u e n t of the Rh6ne ,where the slope o f the water courses and the rapidity o f their floods

are quite exceptional . This means of investigation can not be passedover in silence t o -day , and wil l probably permit of pursuing investi

g a t io n s , which otherw i se would not give any important result .

QUESTION 111.

Wh a t a r e th e be s t m e th ods f o r fin din g r u l e s f o r a n n o u n cin g flo ods

According to the difficulties considered in the above paragraphs,A

,

B, C ,

and D,two general processes are recomm ended for forecasting

the levels of rivers at stations : ( 1) by utili zing the risings o f theprincipal water course and its a ffiu e n t s , that i s to say , the diff erences

between the m inimum,where t h e rise o f water begins ( initial stage )

and the maxim um,where i t ceases . 2 ) By com paring the highest

absolute stages that the water reaches successively at diff erent pointsduring the considered flood .

1. An n o u n cem e n t of flo ods by r ise s — The first process is alone applicable t o the lon g-time predictions Of paragraph A this i s whatBe l g r a n demployed fo r announcing the floods o n the Seine at Pari s in 1856 . It

el iminates an im portant source o f error by taking count Of the ine

qualities o f the initial stage o n the gauge for which the predictions arem ade . To this stage (variab l e according to the circum stances whichhave preceded the flood considered ) we add a probable rise calculatedby the aid o f the actual rises at the stations of observation above ; itwil l generally be a function o f the first degree if the development of

the series which corresponds t o the influence o f each gauge up the

river allows o f considering them as converging rapidly enough .

RULES FOR PREDICTING FLOODS. 147

“

The study o f the rises i s particularly indispensable if one considersmultiple flood ; that i s to say , i f a continued great elevation o f

rater on the gauge i s produced by m any distinct oscil lations of therater up stream . The case is presented frequently at Paris

,and in

h e outskirts , where the waters supplied t o the upper a ffiu e n t s unitero m successive fall s of rain . The case in point i s due to the peracable lands from which the waters arrive late after those o f the imle rm e abl e lands and sustain their floods . The highest stages attained'

y the a ffiu e n t s do not then perm it Of foreseeing directly the maxi:Iu m down the ri ver a s one would hope if there was only a single

ise at each station up stream .

We try,Often with success

,to take for the probable rise at the sta

io n a s im ple m ean of Observed rises o n the upper secondary basins'

y taking count o f the inequality o f the surfaces and o f the p a r t ic

l a r degree of im perm eabil ity of each o n e of them by appropriateo e fficie n t s

,o r by the choice of m any stations in the sam e basin

,as

l e l g r a n d has done by taking at t h e same time the rises of the Marne

t Chaum ont and at St . Dozier f o r announcem ent at Paris .If the hyp othetical relation between the rises up and down streamo e s n o t appear to be sufficiently sim ple , the proceedings indicated in

h e outlin e sketch by Mr . D . De a g u e (Gauthier Villars , 1892 , pp .

5 will permit o f giving to i t a graphic representation ; if

.e ce s s a ry, o n e can state the equations o f relation by means o f u h

nown coefficients,and determ ine those co eflicie n t s by the method

f least squares,so a s to satisfy in the best way a certain num ber Of

.bs e rve d cases ; but this method , very laborious , has the in co n ve

iie n ce o f not wel l taking account o f the natural circum stances bywhich the floods are distinguished .

Som e rapid “ trial s,

” without direct solution , led to the same re

u l t s , as h a s been demonstrated by Inspector General Allard (An n a l e s

fe s Po n ts e t Ch a u ss ée s , 1889 , 1e r s e m .

,Le S e in e , Tome I , p .

Nothing proves,moreover

,that the relation above m entioned should

»e a continuous function o f the variables it represents , and there are

ven some chances that it m ay be otherwise when the real rises sur

ia s s certain critical points at som e stations , such as certain level s o f

u bm e r sio n beyond which the wetted perim eter o f the bed o f the river

hanges quickly .

On e is thus led to form categories o f Similar floods for a given sta

io n and to m odify the formula o f prevision according t o the ca t e g o

ies . A trial o f this kind h a s been made quite recently f o r many

;auges in the basin o f the Oise , but in such cases there i s great dan

;e r o f multiplyin g t o o m uch the particular cases so that they are not

e a d il y recognizable .

By taking as abscissas a conveniently chosen function o f the rises.t certain o f

“the upper stations and as ordinates a similar function

O


o f all the others,and writing by the side o f the point thus deter

mined the actual corresponding rise observed a t the station for whichthe prediction is to be made , the locus o f the points o f equal stagem ay be considered as the proj ection o f a curve o f level s o f a surface

which wil l give som e idea of how the rises in question are related .

2 . Pr e dictio n s by a bso lu te s ta g e s — Thi s representation o f a law establ ish e d between many variable quantities is util ized by M . M a zoye r f o r

the announcem ent Of the floods Of the Loire at Nevers (An n a l e s d e sPo n ts e t Ch a u ssee s , 1890 , 2d sem . ,

Tome xx,pp . 441-511) but in place

o f rises the highest levels attained at each point are considered . Asimilar graphical process had been in use since 1882 f o r studying therelation between the maximum Of the Seine at Rouen , that Of theSeine at Mantes

,and the level o f the Open s e a at Havre

,about thirty

six hours after this latter .

If w e consider the simple floods in which the elevation Of t h e waterobserved at a station arises from a single similar movement Observedat a distant station up stream without any interm ediate affluent

,the

comparison o f rises is no longer essential . The highest levels attainedat both places are then generally in direct relation ; in taking the

first as abscissas,the second as ordinate s

,we o ften find that the ex

t r e m it ie s Of the latter depart but little from a regular curve which isuseful in making predictions . The Hydrometric Service of the Basin

Of the Seine has established many graphics o f thi s kind which it

uses to great advantage .

It goes without sayi ng that the above curves m ay be replaced by

tables o f single o r double entry ; this latter process has been in preference employed by M . JOl l o is for the floods of the upper Loire

(An n a l e s d e s Po n ts e t Ch a u ssée s , 1881, l er s e m .,Tom e I

,pp . 2 73

CONCLUSION.

The rules j ust considered for predicting floods are quite simpleenough ; f o r finding them o r m aking application of them

,it suffices

to observe exactly the heights o f the water,either by the eye directly

or by self-registering apparatus well maintained . It seems certainthat we m ay Obtain thus in m ost cases satisfactory predictions

,par

t icu l a r ly by carefully studying the conditions supplied by the greatestknown floods .

Pr e dicti o n s by m e a n s of disch a rg e s — A m ethod rather m ore co m

plicated,o f which the principle is due to M . H a r l a ch e r

,Professor at

the H igher Technical School at Prague,gives

,it appears

,good results

o n the E lbe in Bohem ia . It might be recom mended in analogouscircumstances

,though it can n o t be made use o f very easily in many

other cases such as described above . It presupposes essentially that

the progression of the floods at a station depends exclusively o n theheights observed at several stations up the stream sufficiently distant


S E C T I O N I I I .

M AR INE M ETEOROLOGY.

1. THE FORECASTING OF OCEAN STORM S AND THE BEST

M ETHODS OF M AK ING SUCH FORECASTS AVAILABLE TO

COM M ERCE.

WILLIAM ALLINGHAM .

Every seafarer will very readily adm it that the forecasting o f suchdread meteors as ocean storms i s a far easier m atter in theory to thefew than in practice to the m any . Hence

,I approach a considera

tion-o f thi s intensely interesting and highly im portant su bj ect with

a feel ing o f diffidence verging o n despair . The inte rval al lotted for

reading the paper is necessarily lim ited,the field for discussion s o

vast and fertile,that for mortal to com mand success in his venture

is impossible , however much he may strive to deserve it . Nauticalm e n there are

,under every sky in the wide world ’ s navies

,whether o f

peace or o f war,thoroughly competen t to treat the vexed question o f

ocean storms from a higher plane than I . The arduous duties o f o u r

noble but neglected profession,however

,too often preclude close

application to clerkly work of this nature,and m ankind is thereby

a decided loser . I have,therefore

,acc epted the invitation which you

have done me the honor to give,as an earnest that

,in the words of

the il lustrious Maury, a seam an is fit f o r other things than tackingship or washing down decks ; and in the sincere desire to arouse sea

farer ’ s of every nation not only to assist in weather work by recordingobservations at sea and in unfrequented ports

,but al so by taking a

far more active part in conferences at which nautical matters arebrought forward for detailed discussion .

The forecasting of ocean storm s i s of great util ity,both to those

that go down to the sea in ships and t o those wh o prefer to gazeupon t h e ‘

m igh ty ocean from dry land . I have,consequently

,deemed

it necessary to deal with the subj ect chosen for me from both pointso f view . A navigator

,rem ote from the land and the electric tele

graph,i s perforce his o wn forecaster o f ocean storms ; and the gravest

responsibility attaches to h is decision , inasmuch as a m i sinterpretstion of the scanty data at hand may tend to the total loss o f h is

150

FORECASTING OF OCEAN STORM S . 151

devoted bark and all her crew . He wil l rely upon such signs as skyand sea afford to men whose l ives are spent in continual conflict withthe elements ; wh ile , at the same tim e , not unm indful of instrumentalindications and the published deductions from the experience o f

navigators wh o , in som e instances , will long since have passed awaydown the dim corridors of time . An overwhelm ing torrent of literature re lating t o the law o f storm s has flooded the m arket since itwas first formulated , and the ebb is n o t yet . It would

,however

,be

utterly unsafe to assume that increased certainty has been borne o nward by the turbulent, frothy stream o f words

,either as to the law

itself o r the deductions therefrom embodied in so -cal led rules fo rhandling a ship th a t sh e m ay altogether avoid , o r partially utilize ,the winds Of a cyclonic storm . Despite the imm ense am ount of

labor bestowed upon tracking these meteors by the aid o f synch r o n o u s charts , I a m reluctantly compelled to confess

,without

reserve , that navigators have not been supplied with m uch informa

tion o f really practical value with respect to ocean storms subsequentt o the discovery that they are

,generally speaking

,circular whirl s of

varying size and energy ; m oving onward , n o w fast , n o w slow ,over

the waste of waters . The sailor i s unable to depend implicitly o n

the curiously contradictory conclusions o f m odern professional and

amateur wea ther w orkers . He n o t infrequently finds that his o wn

watchfulness and faculty for generali zation are m uch more essentialt o safety than all the drawing-room storm maneuvers in existen ce .

Forecasting Of storms at sea involves a rapid approximation to thevalues o f several variable quantities ; a n d

,having regard to the

indisputable fact thatweather workers o n shore,although assisted as

far a s p o ss ibl e by electric communication with outlying districts ,occasional ly forecast a storm which fails to put in an appearance , o r

let o n e slip in o n them unwittingly,there is matter for congratula

tion that navigators com e out o f th e ordeal by wind and wave so

well . To insure an exact result to any given prediction of an ocean

storm the anxious but self-reliant m ariner must know the bearing o fits center

,its distance from the ship , the direction whither it is

traveling,and its rate o f m otion onward . Need I say that the

modern book com piler,in a hurry

,has only helped to m ake confusion

worse confounded as regards our knowledge o f these points . I most

heartily agree with a statem ent referring to ocean storms made by a

well-known navigator,Capt . S . T . S . Lecky , R . N . R.

, that“ we ca n

not but feel that t o a great extent their origin , shape , and movements are

,as yet

,purely matters Of speculation . SO much that i s

contradictory is daily appearing,and such various plausible theories

are being propounded,that it i s most d ifficult to arrive at any safe

and practical conclusion .

”

Probably no great discovery has ever flashed upon the world unless ,


and until,a path had been cleared through a dense growth of rank

weeds of em piricism,and doubtless many a o n e came within almost

m easurable distance o f the law o f storm s prior to t h e advent o f Redfie l d . The first hurricane on record is perhaps that which ChristopherColum bus and his hardy toilers on an unknown

,awe-inspiring sea

endured for three days and nights Of leaden-footed hours in their tiny

craft near the Azores in February,1493 . It is

,therefore

,peculiarly

appropriate for prom inent mention in this paper when all the world

and his wife have set their faces toward the Columbian Exposition atChicago . Even five centuries ago seafarers noticed that the stormwind did not blow unceasingly from o n e direction only

,but from

several points o f the compass in succession . The Ph i l o sop h ica l Tr a n s

a cti o n s of 1698 contain a clearly-drawn word picture of West Indian

hurricanes by a Capt . Langford,who w a s evidently intim ately a o

qu a in t e d with some o f these u ndesirable visitors . This o l d-time navigator pointed o u t that a West Indian hurricane is a whirlwind , inwhich the gale commences from the northward

,gradually changing

through west to south and southeast,which point being attained its

fury forthwith abates ; o r , as the modern mariner , even of the mostslender experience

,would say

,cyclon e centers travel westward t o the

northward of the West India Islands . On e page o f nature ’ s entranoing book lay wide spread before the observant eyes o f that m erchantshipm aster , yet he failed to decipher it s crabbed characters by theimperfect light which then prevailed in the world of science

,even

though he quaintly relates that storm warnings sent to more westernislands from Dominica and St . Vin cent

,ten days in advance

,were

generally correct . He used to get under way and run o u t before the

northerly gale in order to obtain the necessary and sufficient searoomto keep clear o f the land when the wind should shift to southwest .Three centuries ago

,then

,seamen were wel l aware that West Indian

hurricanes are whirlwinds o f com paratively insignificant diameter

bu t awful energy , and that they might be fallen in with m ost oftenfrom July to September . Little if anything , however , w a s known asto their direction and rate o f travel . The full

,change

,

'

and quarters

o f the moon were considered critical periods , especially if the sun

were exceptionally red,the stars with halos

,the hill s unusually free

from cloud and mist,the northwest sky black and foul

,o r the sea

smelling m ore strongly than its want . Franklin,in a letter dated at

Philadelphia,July 16

,1747 , wrote that the air i s in violent m otion

in Virginia before it moves in Connecticut,and in Connecticu t before

it moves at Cape Sable,

” thus foreshadowing the result arrived at by

Re dfie l d,a naval architect of New York

,to whom the world is deeply

indebted for the very first reliable enunciation of the law of storm s .He gathered together ships ’ l o g books , laid down the data thus obtainedin their respective geographical positions o n simple synoptic char ts ,



that a ship would sail away from the cente r by keeping the wind on

the starboard quarter in the northern hemisphere and o n the portquarter in the southern hem isphere ; provided , sh e would st eer satis

f a ct o r il y, and not broach to , a fact only known to those conversantwith her sail ing qualities . In 1872 Capt . Wales , harbor master atMauritius

,appears to

,have arrived independently at a similar rule

,

and this maneuver is now given in the text books . It i s to Pidding:

t o n that w e owe the term cyclone,as applied to revolving st orms ,

which he derived from 16110 9 ; not as some assert as affirming a true

circle,but merely a closed curve

,fo r in the Greek that word represents

among other things the coi l of a snake . There is a serious difficul ty

in the way o f und erstanding exactly what Piddington and h is contemporaries meant by incurving spiral s and cycloidal wind

system s .

Modern weather workers have introduced so many tantal izing exce p t io n s to the law of sto rm s that a seaman aware of them would bebewi ldered . A ship at sea

,in a cyclone

,i s not a fixed Observatory .

Hence,if this fact be ignored

,it follows that arithmetical exercises

relative t o the angle of indraft will prove exasperatingly misleadin g .

Fo r practical purposes the circular theory i s not more uncertain thanany other . Blanford asserted that a cyclone center may be from 1 t o

5 points before the port beam when running with the wind righ t aftin the Bay Of Bengal ; F . Chambers concludes that the indraft variesfrom point to point around the whirl

,increasing from zero to 35° as

the observer recedes from the storm center ; Capt . Toynbee found thatthe indraft increases as the center is approached and i s more markedin front Of the storm ; Capt . Whall i s firmly convinced that with agood o fli n g the wind blows directly for the storm center in the rear ;Ferrel proved mathematically that indraft varies not only with thedistance from the center

,but al so with the latitude . Many other ex

amples might easily be given o f confl icting estimates for finding t h ebe aring o f a storm center ; but enough h a s said to show that theproblem is

,so far

,an indeterminate o n e in a great measure . Even

the term center has n o t been sati sfactorily defined . Granted thato n synchronous charts the shape of a cyclonic di sturbance i s e l l iptical

,with the maj or axis in the direction of travel

,then i s the

s o -called center a physical point or an area at On e o r other Of thefoci

,or at the intersection o f the axes? Occasionally a cyclone ex

tends right across the North Atlantic from America to Eu rope , andthe question arises as t o the bearing Of t h e center of such a systemat positions along the closed curve . Abercromby does not help meto form any defini te conclusion when he says that the center o f a

cyclone is displaced toward one side o f the oval and may move from

o n e side to the other!Yet the center is the first requisite in forecasting a storm . Comment i s superfluous from a nautical point ofVl eW.

FORECASTING OF OCEAN STORM S . 155

The average tracks of storm s have been approximately known formany years , but even a cursory glance at the 1892 North AtlanticPilot Charts , published by the U . S . Hydrograph ic Office

,shows that

com plicated and unexpected d ive rg e fice s from the usu al routes occur

at times . Similar instances are al so noticeable in the erratic behavior o f storms over other oceans which would upset the best laid

plans o f experienced storm forecasters . The storm tracks o f 1883,

determined by the U . S . Signal Service,clearly indicate that the route

most affected by Atlantic cyclones runs from a position south o f

Newfoundland to the north o f Scotland . They drift eastward directly

along the track o f the Gulf Stream . Som e,however

,which start well

,

either die o u t altogether o r proceed due north in m id Atlantic . Others

form closed curves a n d defy prediction . In March one apparently

broke up into two dis t inct cyclones , o n e of which made the Bay ofBiscay , and the other Va l e n t ia , on the west coast o f Ireland . In

April one which had reached N . W broke Off to southeast,

east,and northeast

,eventually passing over Brest instead o f Aberdeen

,

as a well-r e g u l a t e d cycl o n e would have done . Another in mid-oceantraveled east

,north

,southwest

,south

,and northeast . In November

a cyclone moved eastward to the southward Of the Azores for threedays ; and another in Decem ber moving southeast in N . W.

turned east and northeast to N . W. thence north,west , south ,

and southeast to N . W . where it apparently j oined forceswith another

,which

,three days later had followed i t over Halifax ,

N . S . The rate o f travel i s also very variable . On e o f the above

mentioned storm s moved over 2 0 ° Of longitude during each of two

consecutive days,but only 10 ° during the followin g forty-eight hours .

Occasionally a rapidly-moving storm comes to a halt for a few daysand then takes up the running again like a giant refreshed . Mr .

R . H . Scott, and others , have referred to this fact at various times .Hence

,there i s l ittle cause for surpri se that the public -spirited

attem pt o f the New York Herald to forecast storms bound across theNorth Atlantic w a s n o t so successful as it deserved to be . Theaverage direction and rate of t r a ve l .f o r cyclones over a given oceanavail but l ittle when tracks are not infrequently looped and theonward motion anything up to 70 miles an hour .

The electric telegraph has done much to make easier the lot o f astorm forecaster o n shore , working in a snug room far distant froman approaching disturbance . Re dfie l d , in 1847 , seems to have sug

gested that this means of conveying storm intelligence between oneplace and another would be u seful . The late Adm iral F itzroy may ,however

,be regarded as the pioneer o f storm forecasting based upon

actual ob servations transm itted by wire from remote stations to acentral weather ofli ce , to be deal t with there and warnings issued tothe seaports when necessary . His predictions were not always in


agreement with the results , but it m ust not be forgotten that they weretentative

,and even now certainty is denied . The idea of warning the

coasts Of Europe by telegram from ships anchored in the ocean towestward h a s frequently been

.

m ooted,and warnings from North

Am erica are,and have been in favor . The Anglo-Am erican Telegraph

Com pany sent messages without cha rge from Heart ’ s Content,New

f o u n d l a n d , to England , but the place of Observation.

w a s unsuitablef o r the purpose and they were discontinued in 1871. James Gordon

Bennett Obtained better resul ts in a similar w ay at his own expense ,and France has not lost all faith in th is m ethod

,a s sh e stil l receives

inform ation from Wa s hington o f storms encountered by steam shipsbound westward . The U . S . Hydrographic Oflice is in possession o f

many records o f the weather experienced by the Atlantic grey

hounds,

” and an examination of these passages would perchance de

termine whether,and how often

,a Ca mp a n ia or a Pa r is might give

rel iable storm warnings at either end o f the j ourney,provided every

effort were made t o obtain such information im m ediately upon arrival

at Queenstown , Southam pton , and New York . Her Maj esty ’ s ship

Br iskanchored f o r s ix weeks at the entrance t o the English Channela s a stationary storm -warnin g vessel

,but she proved a failure . It

may be that those respo nsible did n o t have their hearts in the work ;for Capt . Wharton , R . N.

, Hydrographer to the Briti sh Adm iralty ,h a s said that anchoring a t s e a i s n o t such a phys ica l impossibil ity assome shore folk believe . Morse thought that simple automatic registering buoys might be dotted over the ocean

,and Capt . W . Parker

Snow h a s been bold enough to indicate a cordon of ships at intervalso f 50 0 m iles anchored between North America and Europe , in elect r ica l communication with each other and with the land .

If storm warning be worth doing at al l i t is worth doing well,and

money should no more be begrudged in prom oting the safety of lifethan it is .to the invention o f m eans f o r the more expeditious destruotion thereof . There should be educated observers

,n a u t ica l

i

m e n by

preference,familiar with weather indications at their several stations

,

at Martinique,St . Thom as

,Habana

,Nantucket

,Cape Sable

,Cape

Race,Valentia

,Iceland

,Bermuda

,Madeira

,and Flores

,al l

‘

in com

m u n ica t io n by submarine cable with the United States and Europe .

Science i s catholic,and each m aritim e nation might be required to

support this international system of storm warning . If this be impossible

,then I would suggest th at synchronous charts for the whole

globe be undertaken,o n which would be carefully laid down the

requis ite inform ation , but free from an undue striving after artisticeffect wh ich adds nothing to their utility though much to their cost .

They should be international in every w ay, and , after the weather

workers of e ach nation have drawn isobars,etc .

,o n precisely the same

data,a critical comparison should be carried o u t by a truly r e p r e



of cloud observations : “ Probably no prognostic i s so valuable to asailor as that afforded by clouds

,particularly those o f the cirrus

formation ; and while their val u e as prognostics has been recognizedfrom the most ancient times

,i t is only rarely cultivated , and yet I

have no hesitation in saying that there i s no weather warning for anisolated Observer that can in any way compare with them .

” My o l d

m aster,Capt . Henry Toynbee , whose nam e i s a household word am ong

Officers o f the Briti sh mercanti le marine , Ensign Everett Hayden ,U . S . Navy , and other observers have al so mentioned the same fact f o rthe benefit of navigators . Capt . A. G . Froud

,R . N . R.

,has j ust sent

me an interesting letter o f Vice -Consul Ramsden , at Santiago deCuba

,explaining the method of Padre Vifie s at Habana

,a wel l-known

authority o n West Indian hurricanes , and stating that in Cuba cirrusgives the first indication o f the position o f a hurricane

,and that the

clouds “ enable o n e to say whether the low barometer is due to a circu l a r storm o r n o t .

” Nevertheless,it must not be forgotten that

cloud observation requires careful training,and schools for teaching

the elements o f weather work are conspicuous by their absence o n

this side of the North Atlantic . Many leagues away,I can not but

await your reception o f m y paper with some trepidation , mindful ,however

,Of the fact that the m a n of science “ loveth truth more than

his theory,and that the subj ect i s of itself far more important than

the m ann er of explanation .

2 .— THE CREATION OF M ETEOROLOGICAL OBSERVATOR IES ONISLAND S CONNECTED BY CABLE W ITH A CONTINENT.

ALBERT,PRINCE OF M ONACO .

During the long periods o f time spent o n the North Atl antic,o n

board of my schooner,L’H ir o n de l l e , devoted to investigations touching

oceanography,and after a careful study o f the im portant labors of

Am erican oceanographers and m eteorologists,I remain convinced o f

the utility attached t o the creation,upon the scattered i slands be

tween Europe and America,o f posts Of Observation dai ly, reporting

the state of the atmosphere t o a central bureau,as sOo n as they shall

be in direct communication by m eans Of telegraphic cables witheither o f these two continents .By m eans o f the m eteorological observatories n o w on the continent

we are perm itted to forecast,in a general m anner

,the approach of

certain tem pests . But what resul ts would be Obtained if those pert u rba t io n s were studied upon the very spot where formed

,inasmuch

as the surface of the waters gives origin to most of the phenomenawhich break the equil ibrium o f the atmosphere .

Meteorological observatories o n the ocean,allowing us to trace in a

OBSERVATORIES ON ISLANDS. 159

'

e g u l a r manner the mutation o f minima and maxima,the' variations

bf the temperature and winds , would aff ord the means o f distinguishn g the principal whirlwinds and the secondary depressions

,and

enable us t o trace the zone of influence o f each of them .

The follow ing is my schem e com prehending,in its broad lines

,the

)rg a n iza t io n of North Atlantic Observatories :The expenditures for the creation and keeping up of these estab

.ish m e n t s should be supported in common by the governments ofEurope and that of the United States . Individual donations

,besides

,

lh O l l l d be accepted .

The points which I consider as most im portant f o r the m eteoroo g ica l observatories of the North Atlantic are the Cape Verdeislands , the Azores , and the Berm udas .The Cape Verde Islands are situated in a region in which

,accord

n g to the Pilot Chart, many storms originate , thence going toravage the West Indies and the coasts of the United States . Those

e lands are , at the sam e time , situated along the outer border Of the

zircu l a r movem ent of the North Atlantic waters,Of which m y

researches upon the currents have shown the existence and theourse .

The Azores are situated near the cent er of this circulation,o n

which account they deserve special attention , forasmuch a s an inter

esting coalescence occurs between that center and the center o f thearea of high oceanic pressures when the m aximum is bearing west

wa rdly t o coincide with that o f the Bermudas .The Bermudas are situated near the western border o f the circula

tion o f the waters,not far from the Gulf of Mexico

,which plays an

important part in oceanic m eteorology ; m oreover , they are under

th e influence o f the Gulf Stream .

With these three points at o u r com mand an efficient supervision

could be exercised over the North Atlantic . So m uch the m ore as atth e Cape Verde Islands and at t h e Azores the height o f the m ountains m . and m . ) would perm it com plete Observations

to be made by means o f neighboring posts for the Observation Of the

u pper regions o f the atm osphere . But such supplem entary observa

t ions would,fo r the present, be o f secondary importance in the pre

vision o f weather , inasm uch as the inquiry thus far m ade into the

m aterial s collected at the Observatory o f Ben Nevis shows that Oh

servation o f the inferior layers is the most advantageous to such a

prevision .

It will oftentim es happen that the Observatories , if placed at St .

Vincent f o r the Cape Verde Islands , and at San Miguel for the Azoresand on the principal island for the Bermudas , will be in a position ,by ships putting into port

,to add to the local Observations , Observa

tions m ade at sea one o r two days previously . Thus , we should pos


sess,at a given time

,a sm al l network o f Observations covering a su r

face o f several degrees .

Fo r a lon g time I have been meditating upon the programme o fwhich the broad lines are indicated above . But ere entering into itsexecution I m ust wait unt il the center Of Observations , the mostinteresting f o r Europe

,the group o f Azores

,shall be Connected with .

the continent by a cable . Judging from the actual agitation aboutthat undertaking

,it is allowed t o hope that the laying o f the cable is

a mere question o f time ; the m om ent seem s , therefore , opportune toprepare the desired scheme .

I thought it useful t o bring before the intellectual assem bly whichm eets thi s day in Am erica ( scientific representatives o f the whole

world bestowing a new im petus upon the activity o f human intellect )thi s question of oceanic obse rvatories which

,by the manifold services

they should return,would soon be m ultipl ied o n the surface o f the

globe .

This question w a s brought by m e l ast year before the Academy o f

Sciences of Paris,and before the British Association at its session in

Edinburgh o n both oc'

casions the meteorologists and oceanographers

Of Europe agreed com pletely upon the desirabil ity Of establishing the

aforesaid observatories . I am”

convinced that the American sa va n ts,

always practical and stout hearted concerning enterprises o f great

scope,wil l l ikewise j oin their ef forts to mine in hastening the e xe cu

tion o f my plans . Is North Am erica n o t interested in the same degree as Europe in the possession o f advanced information of atmos

ph e r ic perturbations originating upon the ocean , and which exercise

so considerable an influence upon the m eteorology o f both continents?Unquestionably it is a great progress Wanting realization for the a d

vance of m odern civilization .

Again , what will be , fo r powerful na tions , the pecuniary sacrificesinvolved in the aforesaid schem e compared to the ruinous preparations f o r w a r which seem rath er contemplated t o thrust back thehum an race into barbarism .

At least when these edifices shal l arise in the m idst of the seas,far

from the turmoil of politics and war,will it n o t be a legitim ate com

p e n s a t io n t o wise m e n,thoughtful Of labor

,progress

,and peace

,and

jus tly alarmed in viewing people arm ed f o r destruction? Most cert a in ly s o

, and the good parole o f science announcing new discoveries

shall attenuate the voice o f cannons .I com e with s o m uch the m ore j oy to lay before you m y proj ects ,

a s I am certain to find with yo u a sim il ar thought , for you are thedescendants of the sturdy m e n who fought ofyore for life and knowledge ; yo u are already the men o f the future

,contemptuou s o f the

vain glory Of conquests .



about which rotates a system Of whirl ing winds .'

We mustguish between this center o f winds and the barometric stormwhich latter is defined as the center o f the smallest isobaric

“

c

ellipse . This latter is the storm center of modernology ; the former is the storm center o f the mariner and Of the Older

cycl o n o l o g is t s . These centers are often identical , but not necessarilyalways so . Mechanical principles have of late years required us

to study the relations o f the winds to isobars and is abn o rm a l s,but

having done this we must n o w return to the older problem and forthe use of the mariner

,must apply our increased knowledge to the

study o f the simple geometrical problem that he has to do with,

namely,the relation between the movement of the wind and the bear

ing and distance of the storm center .As the direction o f the wind is so minutely observed by navigators

who understand h o w to determine its true direction,notwithstanding

the m otion o f the vessel o n which they are sail ing', it should be easily

pos sible by the accumulation of weather maps to determine the direction and incurvature o f the wind o n all sides of the storm centersat sea . Therefore , up to the present tim e the navigator has reliedupon the wind and its changes to indicate t o him the location andm ovement of the storm center that he wishes to avoid .

The progress Of o u r knowledge of the motions o f the upper and

lower currents o f a ir in the neighborhood Of a well-defin e d hurricanecenter has made it apparent that we may improve u pon the old ruleof the earl ier cycl o n o l o g is t s who assumed that the wind blew in acircle around a hurricane cente r and wh o

,therefore

,stated that if in

the northern hemisphere the navigator stand with his back to thewind he will have the center on h is left . hand .

This rule was always recognized as rather crude,yet for a long time

nothin g better wa s Offered for the use o f mariners,notwithstanding

the fact that the charts o f Re dfie l d , Espy , Loomis , Lloyd , and Lever

rier all showed that the rule is not a law of nature . The f a ct'

th a t

the winds are inclined inward,as compared with the path required

with the truly circular theory,was stated very emphatically by Red

fie l d in 1846 , and he adds that in his charts o f storm s the engraver

had sometimes drawn the winds in accordance with the Old theory,

contrary to Re dfie l d ’ s better j udgment . He states that the averageinclination of the wind to the circular tangent rarely exceeds twopoints of the compass

,and is never so much as w a s often claimed by

Espy ; but it seems to me that the fact should not be lost sight ofthat the land storms studied byEspy and the ocean hurricanes studiedby Re dfie l d are two modes o f m otion in the atmosphere that are oftenessential ly different from each other .The rules f o r locating the center of a hurricane and for determining

the direction Of its motion,hitherto used by navigators , have been

THE M ARINENEPHOSCOPE. 163

of the wind,but this i s

the mariner is in thehood o f any land ; moreover , the incl ination o f the wind to the

ius from the storm cen ter varies largely with the latitude and theposition with regard to that center . Numerous studies

,especially

those of Broun , Ley , H il d ebr a n d ss o n ,Ekholm

,and C layton

,have

shown that the movement Of the wind is subj ect to considerable ir r e g ul a r ity and if the navigator ca n avail himself o f the direction o f m o

tion of the clouds he m ay locate the storm center with much greateraccuracy . The m ost extensive series o f observations o f upper andlower clouds Is that publ ished by Broun in the annual volum es Of

his Observations at M ake r s t o u n, Scotland , for 1843 which form a

part of the Transactions o f the Royal Society Of Edinburg . As theresult of about 3 ,0 0 0 observations Broun found that the lower cumulusscud Is incl ined outward to the winds by 14 . the next layer above

,

o r cirro- ,stratus incline s outward 2 2 . the highest layer o f clouds,

or true cirri , i s incl ined outward These observations were formany years overlooked until

,in 1871 both Clem ent Ley and m y

self,by the study o f .En g l ish and American observations , respectively ,

independently announced the general rule,alm ost in the words that

Broun h ad used twenty-five years before,that a s we ascend in the a t

m o sph e r e the angle by which the m ovem ent o f a given layer differs

from the movement o f the lowest wind deviates m ore and more to the

right . As a result o f the work that has hitherto been done on thissubj ect I think w e may f o r the present adopt the general rule thatbetween th e winds that blow spirally inward and the upper clouds that

.blow spiral ly outward there i s an interm ediate layer o f the s o -called

lower clouds whose motion is very nearly along a circular arc and

that the mariner may more safely locate h is storm center as being ina line perpendicular to the motion o f the lower clouds rather than t o

rely entirely upon the surface winds . If he observe the angle be

tween the m ovements Of the wind and the lower clouds and again

between the lower and the upper clouds,he has a further means o f

determining even the distance o f the storm , although the definite rules

fo r s o doing need n o t now be given .

Assuming,therefore

,that the storm center bears at right angles to

the direction o f m ovement o f lower clouds and is o n one’ s right hand

when he faces the direction from which these clouds are com ing, i tremains only to show how t o use the nephoscope in order to Obtainthe direction o f cloud movem ent .The accompanying diagram s (Plate VI ) present both a horizontal

proj ection and a vertical section o f Ritchie ’ s Patent Liquid Com pass ,as used o n American naval vessel s , as also a similar proj ection andvertical section of my marine nephoscope . The com pass proper maybe described as a heavy bowl mounted o n gimbals and so adjusted as


t o its axis of gyration that its time o f vibration i s rather long, namabout o n e -half second . The lower half of the bowl i s ballasted

,

its upper hal f constitutes a closed receptacle ful l o f l iquid ,by the circular plate of glass . Within the liquid floats thecard and needles ; the compass card shows notquarter points

,but al so every degree o f azimuth .

ing down upon the pla

just below it,and also

r im . As the vessel roll s or phorizontal position fairly well up t o a limiting roll o f aboutdegrees . In the standard compass o f the U . S . Navy the upp

o r flange o f the compass bowl , EE,i s neatly turned to an e xac

concentric with the pivot and about 95 inches in diameter .for the purpose o f setting thereon

,at any moment

,the al idade and

sights fo r Observing the sun and stars , or otherwise d e t e rm l n in g thetrue azimuth and the magnetic variation s and deviations at any time .

Ordinarily , this apparatus i s not in place on the com pass , and , therefore

,without disturbing the regular work of the ship

,we m ay set the

nephoscope o n the compass in place of the astronomical apparatus .The thin circ

‘

ular vertical flange of the nephoscope is shown in section RR,

and it fits snugly over EE. The nephoscope consi sts essen

t ia l ly o f thi s circular flange RR, whose upper horizontal surface i s

the ring o n which appear the graduations f o r every five degrees,num

bered from 0 around to 360 in the direction in which azimuths areordinarily measured . In order to revolve this ring horizontally

, two

small handles , PQ , are provided . Within the graduated ring the ciro nlar area i s covered by a single plate of thin mirror glass of excellent quality

,silvered o n the lower side . But as it is necessary to look

through at the compass card below,the silvering has been removed

in a broad circular band ; there i s also a smaller circle of half its

size,as shown by the heavy black l ine ; the outer and inner bounda

ries of these circles are made quite exactly smooth and concen tri cwith the cen ter of the small bl ack spot

,0,which is immediately over

the compass pivot .

When this silvered m irror is in place upon the compass i t r e pr e

sents a horizontal plane,and it preserves its horizontality with

remarkable persiste nce,notwithstanding the ordinary roll ing and

pitching o f the vessel . In the absence Of any convenient method o f

exact m easurement,I have been able only t o estimate that, in the

case o f the compass used by me o n board Of the U . S . S . Pe n sa co l a ,

the inclination o f the mirror plane to the horizon was rarely morethan two degrees

,and to thi s extent

,therefore

,an uncertainty is

introduced into al l our measurements ; bu t , a s the inclination is pe r

p e t u a l ly oscillating from positive to negative , we have , therefore , onlyto take the average Of a few observations in order t o obtain results


THE M ARINE NEPHOSCOPE. 165

that are apprecia bly free from this source of error.The observer

must, however, be careful to keep the com pass in such adjustm ent

that the bowl shall n o t have any constant error in th is respect ; o f

course this sam e adj ustm ent is also Iie ce s s a ry in connection with theobservation of the the sun or stars

.

In s o far as the mirror is horizontal,therefore

,a l ine drawn per

p e n d icu l a r to it at i t s center is an approximate realization of astandard vertical line o n shipboard , and o u r Obj ect n o w is t o determ ine the m otion Of the clouds with reference to the zenith and hori

zon o f this mirror . When the Observer looks into the mirror he seesreflected therein n o t only the masts

,and rigging

,and pennants

,but

the clouds , and even the sun , moon , and stars . The apparatus is a

simple , crude , but convenient al titude and azim uth instrument,and

with it we can perform all the operations o f determining altitudes,

latitude , tim e , longitude , and azimuth with a very surprising degreeo f accuracy . 1 have many tim es had occasion to set up this n e ph oscope on shore , a n d

, besides observing the clouds , have determ inedthe altitude and azimuth o f the sun ; the probable error o f a single

measured altitude o f the s u n o r m oon is about one-quarter of a degree , and could be m ade stil l smaller by appropriate changes in the

construction . In order to measure the apparent altitude and a zi

muth o f clouds by a . method sufficiently expeditious,sim ple

,and

accurate for use at sea I devised the hollow tube SS ,and the sliding

r o d which fit s within it with friction,and which carries at it s end

the small knob , K . The tube has a motion in a vertical plane a boutthe hinge , S , and when elevated to any altitude is held there by the

friction Of this j oint . The vertical plane through the tube,the knob

,

the central spot , 0 , and the hinge , S ,corresponds with the zero o f the

graduation of the horizontal rim . The num bering Of the degrees

i s from 0 t o 360 . The knob and the central spot, 0 ,have the same

diameter so that in whatever position the knob may be placed (byelevating the tube and sliding the rod in o r o u t ) the Observer canbring his eye to such a position that he will see the knob reflected in

the mirror and exactly covering the spot . Let us suppose that theObserver has done this and that he al so sees reflected

,at C

,a sm all

bit of cloud o r a point in a large cloud,then if he continues to bold

his eye in such a position that K always falls upon 0 the cloud wil lseem to move away from the center of the m irror . But he m ay, if he

choose,so move his eye that the im age of the knob shall continually

cover the selec ted point of cloud,and if he does this , then both cloud

and knob will appear to move together away from the center of themirror . This latter is the m ethod Of observation that is always to berecomm ended

,and if o n e

‘could keep the cloud ' and knob togetheruntil th eir reflections simultaneously reach the edge o f the graduatedr im

,he could then read on the rim an angle that represents the


azim uthal direction Of their m otion relative to the zero of that circle .

The position of this zero with reference to the lubber line,FA

,Of th e

vessel is given by taking from the sam e circle the reading corresponding to the forward end of the line

,F ; the relation of F to the magnetic

meridian is given by taking from the compass card,as seen through

the unsilvered glass,the angle corresponding to the same forward end

of the line , AF ; the relation o f the magnetic to the true meridian isknown from the tables o f deviations and variatious . These four

angular readings,when added together

,give the true azimuth of the

apparent motion Of the cloud .

Inasmuch as we do not Often care to wait a s long as is necessaryfor the image o f the cloud and knob to move from the

'

center to theedge Of the mirror , and especially since it continually happens that

the cloud disappears o r becom es unrecognizable in the m idst of an

Observation,it i s necessary to provide for that class o f Observations

which real ly occurs most frequently,namely

,where the cloud is fol

lowed only o u t to the first small circle whose radius in the presentapparatus is exac t ly o n e inch ; I have , therefore , provided a blackcopper wire or silk thread that stretches entirely across the circular

mirror and is attached to a rather heavy wire form ing a circle adj acent

to the inner edge Of the rim . As th is circle with its wire must beeas ily turned in az imuth , there a re provided two small ha ndles , h and

h ; by taking hold o f these the observer easily brings the thread intosuch a position that both cloud and knob traverse it together as theym ove across the m irror , and n o m atter how short their path may be

,

the azim uth o f their motion is easily read at the end Of the thread .

We thus provide all that is necessary in order to Obtain either thetrue o r the magnetic bearing o f the movement o f the cloud . It i seasy to see how one may utilize the same thread to determine the azim u t h a l trend Of the trail o f sm oke which a steam er leaves in its wake ,o r the trend o f the streamers and pennants seen reflected in themirror

,and

,as all these depend upon the combined motion of the

wind and vessel,they have been subj ects o f regular observation by

'

myself o n the U . S . S . Pe n sa co l a . Moreover , when one wishes to Observe the trend o f the troughs and ridges o f waves

,or o f the foam

that flecks the water with white streaks during high winds,he has

here an apparatus more convenient and accurate than the estimateso f any but the most skilful navigators , a s I can testify from co n s id

e r abl e personal experience . Not only the motions o f the clouds , butgeneral trend

,or the vanishing points o f special formations in the

cirrus clouds,the boundaries of cloud roll s

,th e location Of the

zodiacal light,and the dimensions o f halos and rainbows , are easily

determined .

By determining the apparent angular altitude and the apparentvelocity per second o f the cloud under observation

,when a vessel i s



wa s superseded by,

the m eteorological log book,which requires but

On e regular observation to be made daily . The hour fixed is noon,

Greenwich mean tim e,so that n o m atter in what part of the -ocean

the observations are being m ade the observers are acting sim ultaneo u s l y. This simplified l o g was found to be much m ore desirable thanthe j ournal

,as Observers were induced to continue the work

, .wh o ,

after fi l l ing one j ournal,were apt to decline keeping another o n a c

count o f the labor involved . The result has been that the number of

observers has increased nearly eigh t fold , so that the oceans are nowdotted with many interested workers .

Another valuable feature is the indenture o f the leaves of the newlog book

,which enables the observers to remove the pages as fast as

they are fil led up and forward them to the Hydrographic Office,there

to be utili zed in current work . The m aster o f a vessel,o r the Observer

,

wants to se e the results of his observations while th e facts are stil lfresh in h is m emory and while he i s yet interested in what has r ece n t ly taken place . The Pilot Chart of the North Atlantic attemptsto satisfy this want by placing before the mariner in a graphic formsuch matters as are deem ed o f interest o r importance to h im . Fo r

his time and trouble the Observer wants a ready return if possible .

The accumulation of several years ’ data in the home meteorologicaloffices

,there t o be com piled at lei sure , then t o appear in a volume

t o o bulky f o r reading and too scientific for the ordinary navigator,

with too m uch attention paid to minor detail s,is a danger which

should be carefully guarded a gainst .

The past h a s shown that the above cause has d riven from the fieldm any good observers wh o were once interested

,and kept out Of the

field m any m ore whose co -operation would have been Of the greatestvalue . The loss of their services i s a direct loss t o the science o f

marine meteorology,but

,let u s hope

,it is not too late to again stim

u l a t e them to further eff orts .It goes without s ayin g

‘

t h a t m ercurial barometers are the best andm ost reliable

,but

,unfortunately

,a good mercurial instrument i s an

expensive one . . Fo r this reason many sea-going vessel s a r e su pplied

with aneroids only . Some are supplied with both,but generally the

m ercurial,the reliable one

,is placed in the captain ’ s cabin

,where h e

alone has access to it . On other vessel s it i s often placed t o o high, .

where the light i s not good,o r more with regard to i t s safety than its

accessibil ity,so that on a dark night

,d u ring heavy weather , the

Observer experiences no little d ifl‘icu l ty in getting even a n a pp r o xi

mate reading . Generally speaking,the placing Of many barometers ,

especially in merchant ships,is in the interest of the vessel and its

owners and n o t in the in terest Of science . We must a ccept the situation as we find it

,and deduce from the data furnished the best results

we can .

THE BAROM ETER AT SEA. 169

First o f .a l l , the m ost important thing in considering a set of barometer readings is to determine the reliability of the instrument andObserver . To do this frequent com parisons with a standard barometerare necessary Of readings recorded by th e obs e rve r h im se lf .

A sim ple plan for obtaining these comparisons has been in use bythe U . S . hydrographic offices fo r the past three years

,and the results

Obtained have been m ost satisfactory . The credit Of the plan is dueto the force employed in the Meteorological Division o f the Hydro

graphic Office at Washington , which plan was arrived at after the

mistakes and difficulties o f form er methods in use had been clearly

dem onstrated . I can n o t do better than give an account o f the plann ow in use . Although simple in the extreme

,it answers all practical

purposes .

On the arrival of a vessel in port the meteorological reports are

forwarded immediately t o the nearest branch hydrographic office .

Accompanying the acknowledgem ent o f the receipt o f these reportsare two o r m ore franked postal barom eter cards

,o n the back of which

are brief instructions showing how the colum ns should be.

fi lled .

When in ports o f the United States o r Canada,observers are requested

to record the readings o f the barom eter used for observations at seaat 8 a . m . o r

'

8 p . m ., seventy-fif th meridian time , as at those hours

the U . S . Weather Bureau observers record their observations . Ifthe vessel s are in those ports where branch hydrographic offices are

located,readings at other tim es will answer

,as a record is kept Of the

hourly readings o f the standard in each office . When the cards havebeen properly fil led o u t they are mailed by the observer to the branchhydrographic o flice

,where each reading i s com pared with that o f the

standard instrument for the corresponding time . A copy of thesecomparisons i s immediately furnished the observer . The originalcards are forwarded to the Hydrographic Office at Washington , wherethe comparisons are examined and copied

,after which they are r e

turned to the branch Offices whence they came , there to be filed away ,so that any master o r Observer can readily find out how h is barometerh a s been acting from month ' to month o r from year to year . Inmaking these comparisons it has been found best to take the absolutedifference between the reading Of an aneroid and the corrected reading o f the standard as the total correction to be applied to all thereadings o f the aneroid for that pressure . With mercurial ba r o m e

ters the reading is first corrected fo r temperature ; the diff erence ,then

,between that result and the corrected reading o f the standard

is the correction to be appl ied to the reading o f the mercurial . It isevident that these total corrections are but the algebraic sum o f theinstrumental error

,correction for altitude

,and personal error o f the

Observer.This last error is Of no little consequence , for if the Ob

server i s not faithful in recording the observations at the proper time ,


his work is n o m ore reliable than that of a faithful observe r with aninferior instrument .It might be contended that the corrections Obtained from compar

ison e made in port when the vessel i s light would n o t answer when

she was at s e a deep laden . Supposing this di ff erence in heights of the

barom eter to be 15 o r 2 0 feet the diff erence of correction would be

only o n e o r two hundredths , an unnecessary refinement when it isrem embered that in the height of the storm ,

o r when the m ercury is“ pumping ” considerably

,an approximate reading i s al l that can be

Obtained .

Another m ethod Of obtaining comparisons,which has proved quite

satisfactory,i s by making use of the i sobars o n the U . S . Weather

Maps and the readings at the stations along the coast . As the mornin g readings are taken at 8 o

’clock,seventy-fif th meridian time ( o r 1

p . m . ,Greenwich mean time ) , there i s only an hour

’ s diff erence

between the shore readings and the readings at sea . Under normal

conditions this i s not Of much consequence,especially when the pres

sure changes only a few hundredths in as many hours . Use is made ,also

,o f the 2 p . m . readings o f the Briti sh DailyWeather Report . It

w il l be seen that checks ” obtained from the readings of a vessel ’ sbarometer while in the vicinity of Key West

,Jupiter

,Hatteras

,Block

Island,or Nantucket on thi s side

,and again near the outer stations ,

such as Moville, Valentia , Bishop

’ s Rock,o r Dungeness o n the other ,

would determine pretty wel l whether or not the readings fo r the voyage should be rej ected .

These comparisons are often the only ones obtainable,as the many

duties o f the Officers while in port leaves them little or no time forfil l ing out blanks . Hence

,the importance Of these shore readings

when vessels are adj acent to the stations . It might be mentioned inthis conn ection that the readings recorded o n the vessel s at the timeare obtained under the same conditions

,most l ikely

,as those recorded

for the previous o r subsequent part o f the voyage , which fact lendsvalue to .the comparisons .This second method o f Obtaining comparisons i s confined at present

t o vessel s approaching or leaving the east coast o f the United Statesor the coasts o f Europe . It i s to be hoped

,however

,that in the near

future reliable readings from standard instrum ents for noon,Green

wich tim e,will be promptly furnished from the Azores

,Canaries ,

Cape Verde,and West Indies

,so that corrections for vessel s ’ barom

etere can be obtained in much the same manner that a navigatordetermines hi s chronom eter error when in the vicinity o f a place , thelatitude and longitude Of which have been accurately determ ined .

Without a good idea o f the approximate correction to be applied tothe readings fu rnished

,the investigator will find it quite difficul t to

harmonize ocean barometric data .



reliable . At 10 a . m ., Greenwich mean tim e , February 1, 1892 , the

British steamship Be l l in i,in N . 59° W . 7° had a barometer

(aneroid ) reading o f inches . Applying the correction, +0 15 ,

for th is instrument,the corrected reading would be inches . As

the reading recorded by the observer o f the Briti sh Weather Service

at Sumburgh Head at 6 p . In . Of that day was inches,with steep

gradients to the westward , and as the sto rm center passed to the northwest o f the Shetland Islands , it i s not unlikely that the correctedreading o f the Be l l in i ’s barometer , 180 miles west o f Sumburgh , and

near the storm center , was not far from a correct pressure . The cor

rection applied in this instance ,+0 15, was Obtained from comparison so n this

,the preceding , and the subsequent voyage .

The next lowest reading plotted is that o f the Dutch steamshipWe rke n da m in the cyclone of December 22—2 3 , 1892 . At 2 a . m .

,

Greenwich mean time,December 23

,in N . 49° W . 30 ° the

corrected reading o f thi s instrument (mercurial ) was 70 4 mm . ,or

inches .

The highest reading so far plotted is 790 mm . ,o r inches

,the

corrected reading o f the German s teamship Fn lda ’s mercurial barometer

,at noon

,Greenwich mean time

,January 14

,1891

,in N . 49°

W . 14° These readings would indicate that at sea the greatestrange o f the barometric column occurs in the high latitudes duringthe w inter m onths , the same as o n land , and is about 84 mm .

,or

inches .

The importance o f frequent comparisons can not be overestimated .

To illustrate , the m ecurial barometer o f a well-known trans-Atlanticl iner after being quite regular for two years suddenly changed so

that a correction Of +0 7 8 o f a n inch was necessary . The observer

being notified o f the fact began using an aneroid . The latter in s t rument wa s found t o be of an inch t o o high . Here , then , were twobarometers on the same vessel with a difference o f inch in theirreadings . Thi s is

,perhaps

,an exceptional c’ ase

,but it shows that

each and every barometer should be carefully checked ” before thereadings are plotted as final . Ou t o f the vessel s previouslymentioned the author has taken 70 that had good barometer records .The records show that o f these barometers 60 were mercurial and10 aneroid

,and that the average variation o f the 70 barometers over a

period o f twenty-five months w a s only o f an inch . Only thoserecords were taken where the variation in the correction applied wasless than Of an inch

,and where the barometer had been in use

more than a year . The superiority of the m ercurial barometers ishere a gain shown by a ratio of 6 to 1. Of the 60 mercurial barometere it was necessary to apply a plus correction with 50 and a minuscorrection with the remaining 10

,wh ich fact might indicate that

even with the good Observers the tendency is in r e a din g m e rcu r ia l

ba rometers to move the vernier too far down and thus read to o l ow .

THE BAROM ETER AT SEA. 173

An intelligent interpretation o f the prevailing conditions as indica t e d by the barometer , direction and force of wind , state o f sea , andatmosphere , with a view to not only the present

,but the future action

of his vessel , should be the obj ect of every mariner .At the approach

o f a cyclone , or even when the storm is on , the action Of the barometer together with the shifts Of wind will determine the all importantpoint o f which tack to lay the vessel o n . This done

,and the storm

passed , the next thing is to take advantage o f the future shifts by solaying the course , when first able t o proceed

,that the diff erent sh ifts

will be provided for beforehand and the ve s s s l allowed to continue

on her way without the probability o f being headed off.Good judg

m ent in this direction , based upon the knowledge we already have o fthe general laws o f atmospheric movements

,will often serve to shorten

the passage and bring the vessel into port without much working.It

is not only in bad weather,but in good weather a l s o t h a t the m aster

should be o n the alert . The approach of a “ h igh,

” with successive

shifts o f wind due to that circulation,should be as well understood

and maneuvered for as the approach and shifts of a l o w,

” and forthe same reasons as given above . This important subj ect is worthy

of the fullest investigation and should be thoroughly m astered byevery navigator .

In conclusion,I would beg t o submit for your consideration the

following suggestionsThat the members o f this Congress im press upon ' their respective

governments the desirability and importance of a least o n e set o fsimultaneous Observations taken daily ; that the hour be noon , Greenwich tim e , for reasons previously mentioned ; that all barometer readings be “ checked “ by frequent com parisons before being used ; that

a uniform and simple system o f recording Observations by m ariners

be adopted ; that the recording o f observations be encouraged am ong

shipmasters and officers,and also the study o f ocean m eteorology by

putting before them from time t o tim e , and in as graphic a manner

as possible,the explanation o f the general laws o f atm ospheric move

ments and such other m atters as would be beneficial to them ; and

finally,that al l t h e data collected be used in an exhaustive manner

to the end that from a thorough investigation o f the results obtainedour knowledge o f the subj ect o f ocean m eteorology m ay be consider

ably increased .


5 .— THE SECULAR CHANGE IN THE D IRECTION OF THE M AG

NETIC NEEDLE ; ITS CAUSE AND PERIOD .

G. W. LITTLEHALES.

A freely suspended m agnetic needle i s Observed to be in a state o fcontinuous tremulous m otion o f an involved character which may be

resolved into irreg ular and periodic . The irregular motions comprise

those sudden and rapid fluctuation s in the direction o f the needlewhich can n o t be predicted . The periodic motions are the solar vari

a t io n s which include the solar-diurnal variation depending upon thehour of the day

,the annual variation depending upon the day Of the

year and the solar-synodic variation depending upon th e‘

syn o d ic

revolution o f the sun , the lunar variations depending upon the

moon ’ s hour-angle and her other elem ents o f position,and partaking

o f the character of the tides , and the decennial variations which m ay

depend upon the frequency and magnitu de o f the solar spots . Boththe irregular and periodic m otions referred to are o f such small a m

pl it u d e in al l except the polar regions of the earth that they d o noteffect any of the practical uses Of the magnetic needle on the sea

,

but besides these there is another motion , having an amplitude reaching thirty or forty degrees in som e parts o f the worl d

,which is also

supposed t o be Of periodic character , and which , although not perhaps s o intimately connected with th e '

m e t e o r o l o g ic problems of the

day as the variations o f smal ler amplitude and period,i s doubtless

Of radical importance in meteorologic science .

At a particular instant of tim e the lines o f m agnetic force at any

place,to which a freely suspended m agnetic needle will set itself tan

gent,will have a certain direction and strength . The angle between

the plane o f the astronom ical m eridian and the vertical plane passingthrough the needle , o r the line o f force , i s the m agnetic declination ,or the variation of the com pass ; the angle between the horizon andthe direction of the needle

,measured in the vertical plane passing

through it,is the dip

,o r inclination ; and the force with which the

needle is held in the direction o f the lines of force i s called the mag

netic intensity . The declination and inclination,o r the directional

elem ents , which alone are concerned in a discussion of the motion of

the magnetic needle,have always been treated separately in investi

gating the secular change o f the m agnetic needle . From 1634,when

the fact o f the secular variation o f the decl ination was established ,and from 1676 , when the inclination o r dip was discovered

,reliable

observations o f these respective elem ents are recorded for the greatpopulous centers o f Europe

,and soon observations of the declination

o r variation Of the compass,a knowledge o f which is necessary to

m ariners in the navigation o f their ships,had been made by navi



that th e Inclination at that place will be found to pass through acycle o f changes and return to the same value at regular interval s oftime

.While the separate investigation o f series o f Observations of

declination and inclination i s o f.

g r e a t practical usefulness in gaining

a knowledge o f the rate o f secular change of these elements and predicting values beyond the range o f the observations

,in seeking to dis

cover the causes Of the secular change in the direction of the magneticneedle and to establ ish o r disprove its periodic character the declination and incl ination Should be viewed as component effects o f theforces that are acting . Such a view brings us to the investigation ofthe successive directions in space assumed at successive epoch s by a

freely suspended magnetic needle or the consideration o f the o h

served values Of the declination and inclination conj ointly , instead

o f the separate consideration of values o f the direction of the co mpass needle and o f the dipping needle . As a freely suspended mag

netic needle assumes its successive directions for different times , i tdescribes a conical surface whose vertex is the center Of gravity of

the needle .

If a 'sphere Of any convenient radius be described , with it s center

coinciding with the center o f gravity o f the needle,and the conical

surface be extended through the surface of the sphere,the line o f in

t e r se ct io n will be a serpentine curve whose geometrical nature shouldbe fully investigated

,since it represents the actual secular motion o f

the needle . Prelim inary analytical and graphical attempts have beenmade by Quetelet

,o f Brussel s

,Schaper

,o f Lubeck

,and the mathe

m a t icia n s of th e Coast and Geodetic Survey . The scantiness o f datah a s prevented any safe deductions as to the future course o f the needle .

At the present tim e we know,with moderate accuracy

,the values

o f the three m agnetic elements for the inhabited portions Of the world,

a n d also , with a lesser accuracy , the rates o f secular change in theelements

,but w e have no knowledge as to whether the needle , when

it points in a certain direction at a given place,will ever return t o

the sam e position again,or whether it will at the end o f a certain

period assum e the same direction again,and again sweep over the

same path in the same period . Nor do we know that the secularvariation period

,if there shal l hereafter be found to be one

,will be

the sam e in all parts o f the world .

To promote the study of the secular change it is proposed that thisCongress shall take steps t o secure the co -operation of observers atthe following-named place s to make yearly observations Of the dipand declination at selected stations and to arrange and transmitthem to the U . S . Hydrographic Office at Wash ington where theirdiscussion will be undertaken

BAROM ETRIC PRESSURE AND OCEAN CURRENTS. 177

Ch ris tia n sh a abSa in t Jo h n sAca pu lcoM a za tla nM e xico

Ve raSa n Ju a n de l Su r Nica r a gu a .

Ca lla o Pe r u .

Co nce pt io n Ch ile .

Va lpa ra iso do .

Be lize Br it ish Ho n du ra s .

S. o f Co lo m bia .

Co lo n do .

Pa n a m a do .

La Gu a yr a Ve n e zu e la .

Kin g st o n Ja m a ica .

Po r t Ca st r ie s We s t In die s .

Sa in t Th o m a s do .

Ba h ia Bra zilPa r a do .

Pe r n am bu co do .

Rio de Ja n e ir o do .

M o n t e vide o Ur u g u a y.

Bu e n o s Ayr e s Arg e n t in e Re public.

Sa n dy Po in t (Pu n t aAr e n a s ) Pa t a g o n ia .

Th e Azo r e sTh e Ca n a r ie s

Ca pe Ve rdeBe rm u da

6 .— RELATIONS BETWEEN THE BAROM ETRIC PRESSURE ANDTHE STRENGTH AND D IRECTION OF OCEAN CURRENTS .

Lie u t . W. H. BEEHLER, U. S. Na vy .

The student Of ocean meteorology can hardly fail to notice astriking similarity between the average annual curves o f isobars andthe general circulation o f the m ain currents in the five great oceans .The general circulation of the winds around the alm ost permanent

centers o f high pressure in the North and South Atlantic , the Northand South Pacific. and Indian oceans , deduced from observations Ofwind direction s extending over many years

,h a s been demonstrated

by their coincidence with the curves o f isobars to be in accordancewith the first principles o f meteorology .

There i s a most intimate relation between the barom etric pressureand the wind force and direction . The cha f a cte r o f the gradients o fbarometric pressure is the best evidence of the force o f the wind , andthe great practical value Of the barometer to mariners consists in thefeature that the changes in the barometer readings are the mostrel iable o f all the indications of change in the weather .

12

Ca p e o f Go o dCo n g o Rive r

De l a g o a BayLibr e villeLo a ndaPo rt Na ta lQ u il im a n e Rive rZa n ziba rPo r t Lo u isHe l l vil l e

Ade nSing a po r eSa ig o n

Pe kinHa ko da t eNa g a sa kiVl a divo st o kPe t r o pa u l o vskSitkaUn a la skaHo n o lu luTAh itiLe vu kaApiaM e lbo u r n ePo r t Da rwin

Sydn e yAu ckla n dWe l l in g t o

Afr ica .

do .

do .

0 0 0 0 do .

do .

do .

do .

do .

M a u r it iu s .

M a da g a sca r .

Ara bia .

M a la y Pe n in su la .

Sia m .

Ch in a .

Ja pa n .

do .

Sibe r ia .

Kam ch a tka .

Ala ska .


The North Atlantic Pilot Chart for June , 1893 , has"

three charts ofthe North Atlantic

,the main chart and two small subch a r ts, o n e of

which i s a chart o f the curves of i sobars and isotherms which obse r

va t io n s o f many years indicate to be the normal condition for themonth o f June , and the other i s a chart showing the average annualset o f the surface currents o f the North Atlantic .

Unfortunately there are no m onthly charts o f the currents,bu t the

compari son of these three charts suffices to invite scientific investi

g a t io n o f this coincidence to determine if there be any l a wg ove r n in gthe relation and the manner and eff ect of its Operations .1 submit the remarks on the Pilot Chart in relation to this analogy

between the movements o f the air and the curves o f equal barome tricpressure

Th e s t r e n g t h o f th e s u r fa ce cu r r e n t s is in dica t e d by t h e pr o po r tio n a l qu a n tity o f th ea rr o ws o n th e ch a r t . Th e g r e a t e st n u m be r o f a r r o w s a r e draw n wh e r e th e cu r r e n t s a r e

s t r o n g e st . Th e r e is do u bt a bo u t t h e dir e ctio n a n d st r e n g th o f th e se cu r r e n ts in ce r ta in

pa rt s o f th e No rt h Atla n t ic, a n d o u r vo lu n ta ry co ~o pe ra t in g o bse rve r s am o n g m a rin e r s

o f a l l n a t io n s a r e r e qu e s t e d t o co n t in u e th e ir o bse rva t io ns t o a sce rta in th e e xa ct se t a n d

st r e n gt h o f su r fa ce cu r r e n t s .

In th e Bay o f Bisca y r e ce n t in ve st ig a t io n s in dica t e th a t t h e Re n n e ll cu r r e n t , a s sh o wno n th e m a in ch a r t , se t t in g a lo ng th e n o r th co a s t o f Spa in e a s t t o th e co a st Of Fra nce

,

a n d th e nce n o r th a n d n o r th -n o r thw e st a thwa r t th e cu r r e n t se t t in g up th e En glish a n d

Irish ch a n n e l s , do e s n o t e xis t , a t le a st du rin g th e su m m e r m o n th s ; bu t , o n th e co n t r a ry ,it is cla im e d th a t cu r r e n t s s e t in a so u th -so u th e as t e rly dir e ct io n in t o th e Bay o f Bisca y ,a n d t h e nce w e stwa rd a lo n g th e n o r t h co a s t o f Spa in . No do u b t t h e r e is a la rg e vo lum e

o f wa t e r fr o m t h e Gu lf St r e am wh ich e n t e r s th e Bay o f Bisca y a n d m u s t e sca p e a n d

ca us e s u rfa ce cu r r e n ts t o se t o u t , s o m e a r o u n d Br e st in t o t h e En glish Ch a n n e l, a n d so m e

a r o u n d Ca pe Fin is t e r r e do wn a lo n g t h e co a s t o f Po r t u g a l, th e s e t de pe n din g la rg e lyu po n t h e dir e ct io n o f th e pr e va ilin g w in d .

By co m pa rin g th e blu e win d a r r o ws o n m a in ch a rt with t h e sm a ll ba r o m e t e r ch a rt

a n d th e sm a ll cu r r e n t ch a rt , a s t rikin g sim ila rity a pp e a r s be twe e n th e cu rve s , sh owin ge qu a l ba r o m e t e r pr e ssu r e , dir e ct io n s o f th e win ds , a n d th e g e n e r a l dir e ctio n s o f t h e

o ce a n cu r r e n ts . Am o ng th e ca u s e s wh ich Op e ra t e t o pr o du ce a n d in flu e n ce t h e win dsa n d cu rr e n t s , th is co m pa r iso n su g g e sts th a t th e va ryin g ba r o m e t e r pr e ssu r e m aybe o n e

o f t h e o rigin a l ca u se s a s w e ll a s a fin a l in flu e n ce o n t h e dir e ct io n o f th e cu r r e n t s ,

dir e ctly by its va ryin g pr e ssu r e . a s w e ll a s in dir e ctly th r o u gh it s r e la t io n s t o th e win dsTo wh a t e x t e n t th e ba r o m e t ric pr e ssu r e is a fa ct o r in in flu e n cin g o ce a n cu r r e n t s invit e s

ca r e fu l o bs e rva t io n s . Th e s t r e n g t h o f th e cu r r e n ts de p e n ds la rg e ly o n th e co n t o u r o f

th e co a s t , a s , in th e n o r thw e st pa r t o f th e Ca ribbe a n Se a , wh e r e t h e w a t e r is ra is e d byth e w e st e rly cu r r e n t , a n d flo ws th r o u gh th e St r a it Of Yu ca t a n in t o th e Gu lf o f M e xico ,

a r e se r vo ir wh ich disch a rg e s th r o u gh th e Str a it o f Flo r ida a n d g ive s abn o rm a l s t r e n g tht o t h a t pa rt o f t h e cu r r e n t sys t e m o f t h e No r th Atla n t ic kn o wn as th e Gu lf St r e a m .

In the practical presentation Of meteorological conditions the PilotCharts meet the purpose for which they are publ ished , and Lieut .Comm ander Richardson C lover

,ex-Hydrog rapher , merely intended

to invi te scientific investigation of this relationship Of currents t owind and barometric pressure with the hope that it m ay lead to

ascertain truths o f practical application .

The late Prof . Wm . Ferrel publ i shed a number of letters and



right and form the Gu inea Current , and , during the northern sum

mer,the Equatorial Counter Current . The author concludes that

the winds are first,then configuration of coasts , then the rotation of

the earth,and

,finally

,the force o f gravity in their relative influence

t o produce currents .The wind blows horizontally parallel with the surface Of the sea ,

or inclined at an angle either u pward or downward . In the firstcase the parallel motion would have some eff ect by its friction , andmu ch less if the wind be upward , bu t where inclined downward thedownward pressure causes a depression and forms a ridge o f water infront o f this depression which offers resistance and is carried alongwith maximum eff ect . Where waves are formed the crests are im

p e l l e d along by the wind and a considerable volume o f surface wateri s necessarily transported by the wind . To what extent the windcarries the surface water to leeward depends upon its force and cont in u a n ce . In cases where a storm wave meets the sudden resistanceOf coast l ine the shores have been in d u n d a t e d to a depth of from 2 0

t o 40 feet, as i s reported in the account o f the six typical Bay o f Be ngal cyclonic storms in the “Hand Book o f the Cyclonic Storms in theBay o f Bengal ,

” publ i shed by the Meteorological Department of theGovernment of India .

These facts appear to me to indicate that the direction o f ocean

currents i s most frequently to leeward . On Berghaus ’ Physical AtlasNo . 2 1, Se e s trb

‘

m u n g e n ,the ocean currents in connection with the areas

of the permanent high air pressure in the diff erent oceans are indica t e d . The wind circulation s and curves of i sobars are here al soshown to coincide .

The analogy between the curves o f i sobars and the directions ofw inds and currents i s therefore evident . It only remains to d emonstrate the nature o f this relation and

,if possible

,reconcile all the

theories of scienti sts with the experience of mariners .The most effective manner in which the wind can act upon the

surface waters to produce a current i s where it i s inclined downward,

and where the friction of the moving air i s enhanced by its pressuredown upon the water .

Those areas o f high pressure,more or less permanent in latitude

28° north and south , must necessarily exert by their weight o f air a

g r e a t e r p r e ssu r e upon the water upon which they rest than lesserweights Of a ir in areas of lower pressure exert upon the water inother parts o f the ocean .

The diff erences in temperatu re , diff eren ces o f l e ve l , a n d rotation o f

the earth must combine to give a complicated,unstable resultant

efie ct of this atmospheric pressure upon the sea . The first three co ndit io n s may be in Operation

,but the varying operation of the atmos

BAROM ETRIC PRESSURE AND OCEAN CURRENTS . 181

ph e r ic pressure must cause the final resulting effect upon the surfacewater .The curves o f isobars around the “ high o n opposite sides of the

equator would leave the equatorial regions with less weight o f atm os

ph e r e than where the areas o f high pressure exist .The World ’ s Chart o f Isobars shows that there is a norm al atm os

ph e r ic pressure o n both sides o f the equator from abo u t N . 10 ° toS . This pressure is 760 mm .

,o r inches .

The area o f normal high barometer , inches,southwest o f the

Azores , is about square m iles,and the weight of that m ountain

o f air is tons g r e a te r than the weight Of a ir over an

equatorial belt o f equal area south o f N . where the barom eter isnorm al , or inches , o n e -quarter o f an inch lower than the

“ high .

”

The researches of the Ch a l l e n g e r expedition claim to have establ ish e d that the general surface o f the North Atlantic

,in order to

produce an equilibrium must stand at a higher level than at theequator . I claim that a difference of level must be the difie r e n ce o f

eff ect o f atmospheric pressure . The pressures under the areas of highbarometer would make those areas o f lower level than the equatorialregions

,but the section o f the Atlantic examined by the Ch a l l e n g e r

o n W: and between N . 38° and S . puts the higher level near

the extremities . By computing the effect o f heat, Dr . Croll state sthat the surface level at N . 38° i s 3% feet higher than at the equator .

But th e Ch a l l e n g e r researches did not consider barometric pressureas a cause to lower the sea level . Only a small part o f the ocean was

examined,and it i s probable that further research will demonstrate

that the lowest level o f the ocean is under th e area o f the highest

barometric pressure .

The isobars at about N . 40 ° show that the pressure there is thesame as on the equatorial belt. The gradients north o f 40 ° are

ste eper , and the d ifie r e n ce o f level should be great er . In the NorthAtlantic the area of low pressure is near Iceland , and the effect ofthe barometric pressure should make that the highest level .

.

In the sketch the downward pressure o f the atmosphere in the highand its upward pressure in the “ low ” are illustrated . Manifestly theenormous -difference o f pressure must have the efie ct upon the in co m

pressible water to push it away from the region of the “ high to that o f

the . low .

”

This causes variation in the sea level and surface currents . Thetrue final direction o f the currents must be in the direction of theresultant o f the force o f atmospheric pressu re , the wind , and the rotation of the earth

.Generally

,as the angle which these first two forces

make with each other i s small,the resultant will be nearly t o leeward .

Its strength will depend upon these forces , together with the spe cific

gravity of ‘ the wate r , the time the forces were acting, the acquired

182 CH ICAGO M ETEOROLOGICAL ’ CONGRESS .

momentum,the pre-existing condition of surface whither the currents

flow,and the limiting slope o f the raised sea level . The gradients of

h igh barometric pressure are constantly varying and unequal ly distribute d

,the pressure acts on the water with a downward

,outward

,

spiral motion as on the a ir , and currents flow with the wind or at anangle with it

,depending upon conditions o f surface j ust m entioned .

The investigations o f the Prince of Monaco in the yacht L’H ir o n de l l e ,

during the summers o f 1885 t o 1888, o n ocean currents show a muchc l oser analogy between the curves o f isobars and surface-water circulation

,especia l ly in the ell iptical m ovement in and around the Sara

gossa Sea . The Azores are shown to be o n o n e of the curves,and the

drifts of the L’H ir o n d e l l e floats describe ell ipses varying in‘

dia m e te r '

from 2 0 0 m iles to the coast l ines o n both sides of t h e Atlantic . The

drifts o f derel icts and thousands o f ocean current reports by bottlepapers o f the U . S . Hydrographic Office indicate the same conformityof surface drift with the curves o f i sobars .

The configuration Of the coast line h a s an eff ect upon the circulation

,and the explanation quoted from the June Pilot Chart fully

explains the abnormal Gulf Stream Current .The effect o f the barometric pressure o n the Gulf Stream has been

well established,and in Lieut . Pill sbury ’ s report o n

“ Gulf Stream

Investigations and Results,

” there is o n e chapter devoted to the causeOf the Gul f Stream and o f Atlantic currents . After a very thoroughexamination of the gravity and wind theories , he advocates the windtheory as the principal cause

,but in the closing pages Of the chapter

he explain s abnorm al currents by the effect o f barometric pressure .He states that a difference o f o n e inch in the barometric column , or

about half a pound in atmospheric pressure,wil l give over o n e foot

diff erence in the elevation o f the surface o f the s e a .

The chart o f i sobars f o r the year shows that there is a region inthe North Atlantic between about N . 10 ° and N . 40 ° of about

square miles where the barometricpressure is above t h e norm al ,inches , 760 mm . North o f this zon e there is an area of ocean surfaceof about square m il e s where the pressure is below thenormal .The maximum high is about inches and the minimum low isinches

,o r a difference o f about of an inch . If one inch in

the height of the barometer represents about half a pound in the atmos

ph e r ic pressure per sqii a r e inch , the total difference of the weight ofatmosphere upon these regions reaches an enormous figure , sufficientto cause a very decided d ifference between the level s Of the sea a t theareas o f the m aximum high and minimum low . (Thi s diff erence , Ibel ieve

,amounts t o

’

abo u t 2 0 meters . ) The surface water will beforced up an incline in the region o f the “low .

” The lack of pressure ,or rather the diminished air pressure in the low region t aken in con


BAROM ETRIC PRESSURE AND OCEAN CURRENTS. 183

n e ct io n with the lesser area,will sti ll farther enhance the accum ula

tion of the water in the region o f the “ low .

”

The atm ospheric pressure in the Atlantic ca u ses the accumulationin the western part o f the Caribbean Sea

,and the sea level there

and in the Gulf of Mexico is one meter higher than that Off SandyHo ok , N . Y .

The Gulf Stream , thus formed , unites with the waters of the Atlantic circulating around the “ high and flowing up along theBahamas and following the United States coast l ine t o Hatteras . Thewaters continue o n

, but after passing the Grand Banks they m eetwith no further coast resistance and are pushed o u t by the barometricpressure

,which is constantly diminishing

,into the Arctic until the

upward slope is so great that the dim inishing pressure can no longerforce the water there . A large volume o f water flows down betweenthe Azores and the coasts o f Portugal and Africa

,where the pressure

is less than the maximum,and then continues circulating around as

before .

The water thus continually pressed away by the high pressure from

the m id North Atlantic must be replaced , and. consequently there areundercurrents o f cold water from the Arctic and northern part of theNorth Atl antic to restore the equilibrium . These cold currents wil l ,o n account o f their specific gravities , fal l below the warmer surfacecurrents

,and wh ile this barometric pressure is acting , these cold cur

rents flowing south cannot appear on the surface , for if they d idappear they would

,under n ormal conditions , be necessarily brushed

back again toward the Arctic .

Where the configuration o f the coasts has deflected this circulationof water away from the shores

,the cooler currents may there appear

on the surface,and

,consequently

,we find a cold current from the

Arctic along the coast o f Labrador , sneaking in around Newfoundland and close along the United States coast to Hatteras and Florida .

In the report for 1891, Appendix No . 10 , Of the U . S . Coast andGeodetic Survey

,Prof . E . E . Haskell publishes an account o f o bs e r

va t io n s of currents in the Straits o f Florida and Gulf Of Mexico , and

on page 347 he states

Ove r a.wa t e r s u rfa ce u n e qu a l a tm o sph e r ic pr e ss u r e a n d win d bo t h be co m e ca u s e s ,

a ctin g g e n e ra lly a t a n a ng l e with e a ch o t h e r t o pr o duce a cu r r e n t . Th e fo rm e r is th e

e qu iva le n t o f a h e a d t o be.

spe n t a s a g r a vity fo rce in th e dir e ct io n o f th e t r e n d o f th e

ba rOm e t ric g r a die n t , wh ile th e la t t e r a cts by fr ict io n o n t h e s u r fa ce t o pr o duce a cu r r e n tin it s dir e ct io n . Th e r e is lit tle o r n o in fo rm a t io n e xt a n t a s t o th e cu r r e n t th a t a ny

kn ow n ve lo city Of win d a n d ba r o m e t ric g ra die n t will pr o duce , n o r is th e r e a defin it e

e n o u gh r e la t io n be twe e n dir e ct io n o f win d a nd t r e n d o fba r o m e t r ic g ra die n t t o pe rm it o fm a kin g m o r e

‘

th a n th e g e n e r a l s t a t e m e n t th a t th e cu r r e n t sh o u ld be in th e dir e ct io n o f

t h e r e su lt a n t o f th e two fo rce s .

I quote this by permission,and this pamphlet contains tables con

n ect in g

.

the,

observations of currents with meteorologic data . I also


have in a letter from Prof . Haskell the further statement that,

“ If I

had'

a t my command daily observations of the direction and ' force ofthe wind and the reading of the barometer from stations so located asto surround the Gul f, I could predict the currents much as our weatheris predicted .

In investigating the ocean currents it must be remembered that themountain o f air in the region o f the alm ost permanent “ high ” is n o tconstant in extent or in exact local ity . I have taken the averageannual location and direction o f the areas of the “high in the NorthAtlantic . This varies

,and the Pilot Chart f o r each month shows

these variations graphically . Again,near the belt of normal atmos

ph e r ic pressu re the air ci rcul ation around the high ” i s accompanied

by other circulations,both cyclonic and anticyclonic

,and these storms

will tem porarily disturb the normal condition and cause variationsin the current both in strength and direction .

To follow the movement of a cyclonic circulation across theAtlantic toward Europe and the Arctic the waters under the centermust be relieved o f pressure by the extent o f the abnormal diff erenceof air pressure . The winds also in the cyclonic circulation flow in

,

around,and upward

,and these causes m ust contribute

,not only to

raise the level .Of the surface water,but al so to make thi s increase of

level to take place in a com paratively small circular space ; hence ,the rem arkable high

,almost vertical

,seas which are raised and fall

with such destructive eff ect al l around in a confused mass in thecenter o f a cyclone .

The storm waves quoted from the Bay o f Bengal typical cyclonesare explained o n this theory

,and further exam ples might be quote d

to show that the currents of al l oceans and at al l times are chiefly du eto the atmospheric pressure . Ou r knowledge o f the ocean currents i sfar from exact

,and the Obj ect of this paper i s to invite investigation

of the subj ect o f ocean curren ts in connection with the barometricpressure .

It is extremely difficult t o ascertain the direction and strength ofocean currents . As a rule the reports of currents experienced arereally the difie r e n ce between an estimated run o f a ship in twentyfour hours by dead reckoning a n d the more exact r u n as determinedby astronomical Observations . Al l the e rrors o f the estimated courseand distance made good for twenty-four hours are added and ascribedto currents which

,f o r half the time , may have been in one direction at

o n e rate and at other times in other dire ctions at diff erent rates .With the sextant it i s rare that a captain can determine his posi

tion more than once in twenty-four hours,and until he has means of

finding hi s position accurately at sea more frequently,the reports of

currents experienced wil l be unreliable .I have recently invented a nau tical instrument, the solaromet er ,



storms prevailing at such times . A very notable instance of o n e

species o f such displacement occurred in 1877 when anticyclonicweather conditions were very persistent in low latitudes

,as evidenced

by the extraordinary extent and severity of the d roughts which bel te dthe entire earth in the equatorial regions . Coincidently

,the diver

sion of storm tracks into h igher latitudes was shown,especially by

the phenomenal mildness of the winter seasons . Ten years later,in

1888 there was a repetition of these same characteristic featuresdependent upon atm ospheric distribution . In thi s instance the

mildness Of the winter season , particularly that o f 1889 seems t ohave extended even into the Arctic regions

,caus ing floating ice t o

appear off the coast o f Labrador and Newfoundland in great quantities throughout m onths in which it i s rarely seen at all in that location . At the same time on the North American con tinent there wasa marked deficiency in the severity and extent o f cold waves

,and

storm tracks had their centers far north for extended periods . Thesame mildness appeared , l ikewise , in the northern parts o f the

eastern hemisphere,while in India

,on the other hand

,anticyclonic

conditions predominated , there being a general rise of abnormal

barom etric pressure for a considerable period and scantyrainfall throughout the year .

”(Na tu r e

,June 5

,1890 , p .

It is not within the province of a brief summary,such as the pres

ent,to do more than indicate leading features . Suffice it to say that

the extraordinary persistence and extent of the distribution o f thetypes o f weather described in the years named is not only in strongcontrast to average conditions

,but is in stil l greater contrast to those

prevail ing in years when the d ivergence from the norm al is in anopposite direction

,anticyclones with greater dryness and stronger

cold waves prevail ing in higher latitudes,with corresponding dis

placement o f cyclonic weather conditions into diff erent and,for the

most part,lower latitudes . The present year affords an example of

this variety of divergence from the normal,both in this country and

Europe,the winters in northern latitudes being distinguished by a

severity in strong contrast to their mildness during the years previo u s l y named , and the areas distingu i shed by phenomenal droughtsduring the summer likewise being transferred to higher latitudeswith coincident strengthening of the storms and hurricanes of lowlatitudes . In th is connection it is worthy o f note that what i sthought to have been the lowest reading o f the a r o m e t e r ever r e

corded o n the Atlantic was met with during a storm far south lastDecem ber . So

,too

,the West Indian hurricane season now approach

in g prom ises to be severe . Coincidently,complaints o f droughts are

heard from the interior o f th e North American continent and fromthe north o f Europe

,Great Britain

,especially

,suffering severely . In

order to bring out completely these contrasts in weather conditions

FLUCTUATIONS OF STORM TRACKS . 187

it would be necessary to consider in detail the efie ct Of rearrange

ments i n the distribution o f atmospheric pressure upon rainfall,

cloudiness , temperature , the direction and force of the winds , and

the l ike , n o t only at d ifie r e n t seasons o f the year , but also in p a r t icular localities , which , m anifestly , i s a task o f considerable magnitude .

But without entering thus into detail it i s sufficiently evident for thepurposes of the present discussion that the d istribution o f cyclonic

and anticyclonic weather conditions throughout the globe varies indifferent years in the manner which has been described . The broad

features o f these types of weather are readily distinguishable,both o n

land and sea and in summer and winter,under all the m odifications

which they thus undergo,SO that their transference from o n e latitude

to another can be traced with a good degree o f confidence . Indeed

such transference and local ization o f weather types through m ore or

less extended periods is one o f the most striking and fam iliar facts ofm eteorological observations .

In like m anner,for brief intervals

,there may intervene strongly

marked but relatively transient divergences from the m ore persistenttypes o f weather prevail ing in the m anner which has been described .

Thus,upon some particular date

,anticyclones m ay suddenly appear in

higher latitudes than that which they had been accustomed to f r equen t for weeks o r m onths preceding, and begin to m ove eastward

with more activity and becom e stronger , the storms prevailing alongtheir peripheries being modified likewise in respect to the courseswhich they pursue and the energy which they display . In such cases

an impulse of som e sort appears to have been im parted suddenly tothe atm osphere

,as a whole

,the increased rapidity of eastward m ove

ment and intensification o f storm action being apparent in the case

o f al l cyclones and anticyclones prevailing at the time . Thus , in

stances have been noted in which storms in Am erica and Europe , and

Off the coast of Japan , have acquired phenomenal intensity on thesame day

,const ituting a well-d efin e d period during which the energy

o f atmospheric movements was largely increased , anticyclones as wellas cyclones being everywhere strengthened . In such cases as these

there does not appear to be any delay , such a s would be required ifthe i ncreased activity displayed were dependent upon any slow process o f warming up continents o r seas . On the contrary the r e

arrangement in the distribution o f atm ospheric pressure (and incidental storm action ) begins promptly and in such manner as to in

d ica te that its origin i s not dependent upon local terrestrial condi

tions,the part perform ed by such conditions being, apparently , to

modify rather than originate the activities in question .

Thus,there are in general rearrangements in weather conditions so

well defined,and upon such a large scale , that it i s diffi cul t t o . r e s is t

the conclusion that the atmosphere , as a whole , i s un der the control


o f forces which have a com mon origin and which u ndergo variations

at their very seat of origin . It would seem that it ought to be possible to determ ine t o the best advantage the nature and m ode o f opera

tion of such forces by a careful study of the instances in which thedivergences from the normal are greatest . For several years the

wri ter has been collecting information bearing upon this point,the

outcome being the conclusion that the divergences in question mustdepend ultimately upon som e form o f solar variability

,it being co n

ceded on all hands that the s u n is the great source of atmosphericcontrol . As to the essential nature of this

“

variabil ity there is,how

ever,a question . From current views respecting atmospheric con

trol,we should naturally expect i t to consist in variation in the sun ’ s

power Of emitting heat . It is true that the weather conditions towhich attention has been cal led present anomalies in respect t o tem

p e r a tu r e , but these have reference t o distribution rather than totalamount . Thus , investigators using data from different parts of theearth have reached precisely Opposite conclusions as to whether thesun is hotter o r colder , and that , too , in the very years in which thedepartures from the normal of the kind indicated have been largest .

In like manner,if averages are taken from a sufficiently large part of

the earth,the variation of temperature from year to year is in sig

n ifica n t . So,too

,the exposure of properly gu arded thermometers t o

the direct rays o f the sun in local ities best adapted f o r such e xp e r i

ments,and under the m ost suitable conditions

,has t hus far given n o

evidence o f variations in the sun ’ s power of heat emission adequatet o explain the facts to which reference has been made . It i s a question

,indeed

,whether variation in the amount o f heat falling upon

the earth,as a whole

,could produce the diversified local effects appar

ent and maintain them in the manner which actually appears . As amatter o f fact

,however

,it i s n o t yet known whether the sun is hotter

o r colder when freest from spots . The observations thus far madeagree

,however

,in showing that the interposition o f the atmosphere

and its contents,and especially of the aqueous vapor which it con-J

tains,modifies the transmission and radiation of heat to such an ex

tent that this and not solar variability may be the source o f thechanges in temperature distribution to which reference has been .

made . This being the case,these temperatu re anomalies are to be

regarded as a mere incident in the rearrangement Of atmospheric dist r ibu t io n rather than its cause . Thus it becomes necessary to look

el sewhere than t o variations in the heat-giving power o f the sun for

the source of atmospheric control . At this point and in this connection a relation of the weather conditions described to the distributionin latitude o f the spots on the sun becomes of very great interest. Itis found that the fluctuation in latitude of the belts o f anticycloneson the earth and consequent diversion o f s t o rm tracks to wh ich r e f e r



chiefly,if not exclusively from disturbed portions o f the sun

'

when

at the eastern limb,and that such eff ect may at times originate thun

d e r s t o rm s instead of auroras , the substitution o f one o r the otherof these two classes Of phenomena depending apparently upon thelocation o f the originating solar disturbance relative to the plane o fthe earth ’ s orbit when at the eastern l imb . There is evidence al so of

a terrestrial localization of these phenomena , dependent in part, a p

p a r e n t l y, upon the physical conditions existing at the time in variousparts of the earth

,and in part upon a concentration o f eff ect at

certain hou r angles from the sun . Thus , there are diurnal maxima

and secondary maxima both o f thunderstorms and auroras,and the

regions frequented by them have a belt-like distribution in magneticlatitude . From th is it appears that the lines of force along which theinductive effect proceeds have a very definite arrangement

,and that

there are modifications o f effect in particular portions of the field

which these l ines occupy,giving rise to thunderstorms instead o f

auroras,o r vice ve r sa

,as the case may be . It would seem that the

origin Of the whole process i s through electrification of particular p o r

tions of the sun ’ s immediate surroundings through the agency o f theturmoil o f chemical and other activities due to the special violence o fthe eruptive forces there in operation

,as compared with the rest of

the sun,and that the inductive effect i s propagated outward into

sp a ce'

f r o m these sections of the sun dynamically , or , in other words ,in virtue of the motion o f rotation .

The dynamic origination of electrical currents has been greatlyfam il iarized Of l ate by the commercial applications Of the principlenow in ordinary u s e for m any purposes . Such origination depends

u pon a very different s e t o f conditions from those involved in thermoelectric action to which these forms o f solar activity have generallybeen referred heretofore . There i s no evidence whatever that a magnetic storm is al lied to o r dependent upon heat radiations in anysense o r to any extent whatever . Its method o f origination and

propagation is quite different in every respect,and any heating effects

that may attend are incidental and remote . Thus there is a form Of

solar activity having it s o wn distinguishing characteristics and exi stin g as an entity by itself which deserves most careful study . The

earth'

certainly i s comprehended within the range o f its Operation,

and it i s altogether l ikely that the inductive effects thus experiencedextend throughout the entire solar system , reaching every particle of

meteoric dust and debris , and the vapors , if such there be , in interplanetary space

,a s well a s the planets themselves . Al l these masses

o f matter charged up by induction exhibit permanent,subpermanent

,

and temporary effects,in accordance with which they act and react

u pon each other and l ikewise upon the sun itself,in conformity with

the laws governing induction .

FLUCTUATIONS OF STORM TRACK S. 191

In the j udgment o f the writer it i s the reactionary effect upon thesun itself Of these inductive forces

,causing rearrangem ents in the

distribution of the va pOr s in its vicinity , as seen during eclipses , thatdetermines the formation o f spots and their varying locat ion in latitude in a manner altogether Similar to that seen in connection withthe coincident changes in latitude o f the belts o f anticyclones o n theearth . In any event , “ the fact that these rearrangem ents Of thevaporous surroundings o f sun and earth undergo sim ilar variations

in respect to location in corresponding years points to com m unity of

origin o f these effects , and their evident relation to m agnetic phe

n o m e n a o f the sort which h a s been described i s such that i t wouldseem not unwise to shift, if necessary , the point o f view f o r the co nsideration o f the entire subj ect

,and to attack the problem s at issue

along the line s indicated in this discussion . This involves nothing

less than a reconsideration o f every fact and conclusion in respect tometeorological science from a standpoint altogether different froman assumed variability of solar heat .

It may be necessary t o go so far as even to discard provisionallyand tentatively the convection theory o f the origin Of storms , as

ordinarily held,in order to determine fairly and com pletely the part

which electro-ma gnetic induction o f solar origin plays independently

o f heating eff ects . The distribution o f temperatures , both in a hori

zo n t a l and in a vertical direction in cyclones and anticyclones , and

the velocity and extent of the associated wind m ovements , likewisein a horizontal and vertical direction , are n o t easy of explanation in

conformity with the convection theory o f the origin o f storm s .

Thus,in tropical hurricanes

,where the violence o f the wind m ove

ment is extreme,the t emperature gradient is small . Again , in the

case o f a severe s torm remaining stationary like the New York blizzard

,it would seem that the gravitational inflow o f such enorm ous

masses o f air ought t o involve a fil l ing up process . Certainly , currentviews respecting the forces concerned in storm action do not give thesl ightest intimation a s to where all the a ir goes in such a case as this .

In view Of such facts a s these , and without m ultiplying furtherillustrations

,i t would seem t o be n o t only reasonable but necessary

t o institute an inquiry as t o whether forces other than those heretofore taken into the accoun t are n o t concerned . To this end especiallyim portant is the identification Of the solar and other conditions o n

which auroras and magnetic storm s depend , growing o u t Of which

there comes an apparent relation to thunderstorm s . By the a id o f

the clue thus Obtained it becomes possible to study the behavior o f

the atmosphere o n critical dates , and during critical periods , whose

identification is secured through a knowledge Of these solar and asso

cia t e d conditions . As has been intimated throughout the course Of

the discussion the fluctuation in latitude o f anticyclonic belts and

CH ICAGO M ETEOROLOGICAL CONGRESS.

storm tracks both o n the grandest possible scale , as aff ecting climatethrough series o f years , and in individual instances on single dates ,i s most l ikely to afford an insigh t into the meteorological relationsof electro-magnetic forces o f Solar origin

’

. These forces certainlyplay a part in the economy o f the solar system

,and there are indica

tions that this part i s far more important than has heretofore beensupposed .

8 .—NORTH ATLANTIC CURRENTS AND SURFACE TENIPERA

TURES .

Lie u t . A. HAUTREUX , Fr e n ch Navy .

It is impossible to speak o f m eteorology or the physical geography

o f the sea without the spirit o f the im mortal nam e o f Lieut. Maury .

He it wa s who system atized the best manner of making Observationsand enunciated the principles and general laws o f the circulation of

the atmosphere and the oceans to the scientific world . In his school

this science has been studied,and especial ly in the United Sta tes

,

where it has been most developed o n land and sea .

It i s in that vast country,washed by two oceans

,possessing both

tropical and polar climates , the highest mountains and most extensive plains

,rainless deserts and the m ost ferti le regions

,with coasts

washed by the greatest oceanic river in the world and annually r e

ce ivin g glacial tributes from Greenland— it is there where t h e ele

ments o f heat and cold , dryness and moisture , rage and produce withgreatest force the phenom ena caused by the conflict . There thescience of meteorology

,based upon actual observations

,has made

the greatest progress . There,also

,the public i s m ost promptly noti

fie d and warned o f meteorological disturbances,and m easures taken

to prepare f o r them .

The grand laws o f meteorology,which Maury so admirably reduced

t o harmony from phenomena Often o f the m ost fleeting character andcom plicated by local anomalies

,have by experience been demon

s t r a t e d in detail to be o f great service for the security Of navigation .

.Som e o f these points,especially those relating t o the currents and

surface temperatures Of the North Atlantic,we will proceed to in ve s

t ig a t e in th is paper .

The scientific expeditions so wisely,

directed by the governments o fthe United S t ates , England , Germany , and France f o r the physical

examination o f the ocean,in connection with the deep-sea sound

ings for the trans-Atlantic subma rine cables,have covered the sea so

thoroughly with a net of Observations that scarcely any importantfeature has escaped their investigations . They have established thatn o part o f the ocean is at rest

,but that the entire mass of the ocean

from the surface to the profoundest depths is constantly in motion



There are besides the Gulf Stream other currents which are recog

n ize d as perm anent and have certain general features , such as theLabrador Current

,caused by the mel ted snow and ice of polar regions

,

and the Counte r Equatorial Current, produced by the southwest winds

o f the coast of Africa .

These currents have been found by observations o f navigators andfrom the drift of floating Obj ects , such as ice , wood , bottles , and h ull so f vessel s . The Observation s conducted for the Pilot Charts , publishedby the Hydrog raphic Office of the U . S . Navy

,at Washington

,D . C . ,

have been of great importance . These Show the precise resultant ofthe complicated causes to which a floating vessel is subj ecte d . If

the hull of a derelict vessel,im in e r s e d 6 t o 8 meters in the water

without exposing t o the wind more than a portion Of its dismantledhul l

,should for several days in a month drift in a certain direction

it is evident that the mass of water in which it floats must havemoved in the same direction .

There are other surface movements of the sea which are designatedas permanent currents

,but which facts show are subj ect t o important

and unforeseen variations . An examination of the Pilot Charts willshow several examples .

We beg the reader carefully t o examine these charts,and especially

certain supplements which have been published by the WashingtonHydrographic Office

,viz : The Drift o f Bottle Papers

,

” July,1891 ;

The Derel ict Schooner White ,” February

,1889 .

We'

will proceed to investigate the following : The Norwegian Current

,the Rennell Current

,the currents o f the coast o f Portugal and

the west coast o f Africa , the currents o f the Sargasso Sea , and thetemperatures of the sea from Bordeaux to the La Plata River

,from

Bordeaux t o New York , and in the Bay o f Biscay .

THE NORWEGIAN CURRENT.

In summer the Atlantic,north Of the Azore s , does not appear to

be so much under the influence of the Gulf S tream,and yet in that

season the stream has i ts greatest extension toward the north,a fact

which is dem onstra ted by the tracks o f the derelicts Twe n ty-o n eFr ie n ds

,in July

,August

,and Septe mber , the Wh i t e , in June , July ,

August,and September

,the E. Da vis

,in Au g u s t

‘

a n d Septe mber,and

the Hu n t,in July .

In the season when t h e sou thwest and west winds prevail the watersare pushed northeast and east . The fact is shown by the drift o f

num erous derelicts publ ished on the Pilot Charts,and o f the drift o f

bottle papers in the special supplement of the Pilot Charts,1891

Neverth eless , even in this season , the condition o f the barometricpressure o n the Atlantic prevents the westerly winds from allowingthese waters to reach the shores of Europe . The surface wate rs do

Curre nts of the No rth At lantic in 1892 .Paflis

‘

of‘

drltt ing w’

recks.Pl a t e VIII Ha u tr eux,

Tempe rature of wate r in the BayofBiscay.



north and south . This inust appear logical o n account o f the uniformregularity Of the coast l ine there

,so that this important sheet o f water

i s not influenced by eddies such as would be found upon a n indentedshore .

There is,fortunately

,at Arcachon a fishing industry having five

steamers,under the direction of Mr . H . Johnson , wh o readily co n

s ented to assist o u r investigations and ordered bottle papers t o belaunched for this purpose . The captains o f these vessel s made thefol lowing reports

,which have been forwarded to the Hydrographic

Office at Washington , D . 0 .

Extract from the report o f Capt . Pateau

Th e cu r r e n ts a r e n o t r e g u la r . Th e y a r e ca u se d by th e wind ; with n o r th win ds th e

cu r r e n t s s e t so u th , a n d with so u th win ds th e y s e t n o r th . At t im e s th e r e is n o cu r re n t

w ith th e win d e a st Off sh o r e . In win t e r th e cu r r e n t s a r e s t r o n g e r th a n in su m m e r .

Extract from the report o f Capt . Durand

I h a ve a lwa ys n o t ice d th a t wh e n th e win d is so u th o r so u thwe st th e cu r r e n t s se t n o r th

a lo ng th e co a st o f Fr a n ce , bu t with th e w in d n o r th e a st o r n o rt hwe s t th e y se t so u th t o

wa rd th e bo t t o m o f th e bay, a n d th e n ce flo w we s t a lo n g t h e co a s t o f Spa in , with a

ve lo city pr o po r t io n a l t o t h e st r e n gt h o f th e win d a n d it s co n t in u a n ce .

Extract from the report o f M . Silhouette,Of Biarritz

Fo rm e rly m a n y sm a ll t r a din g ve ss e ls fr e qu e n t e d Ba yo n n e a n d we r e Oft e n lo st . Th e

ve ss e ls we r e ca r r ie d pe rp e n dicu la rly t o sh o r e,h e a d o n . Su bs e qu e n tly th e ir st e m s w e r e

ca r r ie d so u th by th e cu rre n t fr o m th e n o r th .

Al l these reports agree that the currents in the Bay of Biscay areabsolutely depen dent upon the prevai ling winds .In order to confirm these reports we have

,during June and July o f

this year,thrown overboard a number o f bottles

,three-fourths full of

water,attached to floats by a line two fathoms long . The leng th o f

the line being such that the bottles might be readily recovered on thecoast at low tide . These bottle s were thrown overboard 12 to 30 milesfrom the coast in depths o f 40 to 60 fathoms .

The results o f these experiments were collected and sent to theHydrographic Office at Washington

,D . C .

,which office is hereby r e

quested t o communicate them to the Congress . The experiments werecom menced o n May 25 and continued at the rate o f three or fourevery week . The l ast o n e

,recovered o n July 3 , had been

'

thrownoverboard on Ju ne 2 1, by the steamer Oce a n iqu e . Ou t o f the eighteen o r nineteen bottles thrown overboard

,thirteen were recovered

a large proportion . (See table o n page 198 and Plate Ix . )The currents Observed on board the vessel s where the bottles were

thrown overboard were weak and the general set was south-southwest.The bottles were adrift in the water for an average period Of fourteendays

,and the resultant direction o f their drift was to the south-south

east . Not one bottle was found north o f the place whence it w a s setadrift . This i s contrary to the theory o f the Rennel l Cur rent .



dangerous character which has always been accredited t o the Bay ofBi scay . No sailing vessel can beat o ff the coast while the winds andcurrents both s e t her o n . These facts Show that the wind is thepreponderating factor in causing the currents whose direction and setare Often modified by the configuration of the coast . They prove al so

that near the coast there is a surface movement which i s dangerous

t o navigation and should be studied carefully .

Ta bl e of dr if tin g bo tt l e s n e a r Ar ca ch o n .

'

l‘

h r o wn in t o t h e s e a . Re co ve r e d. Dr ift .

De s ig n a t in g La t il e t t e r . t u de .

C u r r e n t D is t a n ce . Dir e ct io n . Days .

M il e s M il e s

CURRENTS OF PORTUGAL AND WEST COAST OF AFRICA.

For this the observations o f the steamers o f the Messageries Maritimes

,taken from tim e to tim e for Six years with about four each

m onth,are considered in detail . These steam ers ply along the coast

o f Portug al and Africa as far as the Cape Verde Islands . The currents have n o t the perm anence to wh ich they are credited

,bu t in

each season there are certain features .Win t e r .

-Along the coast o f Portugal the current sets north andnorth-northwest ; from Madeira to the Canary Islands the currentsets north-northeast

,and from the Canary Islands to Dakar the cur

rent sets west-southwest . The Counter Equatorial Current setssoutheast .

Su m m e r .— Along the coast o f Portugal the current sets south t o

south-southeast . From Madeira t o the Canaries the current setssouth-southwest . From the Canaries to Dakar the current sets southo u t hw e s t . The Counter Equatorial Current sets east . A velocity ofabout o n e m ile per hour has been found between the Canaries andDakar and between the equator and Pernambuco .

The diff erence in direction in summ er and winter corresponds with

the changes in the direction o f the prevailing winds in these regionsin these seasons . During the winter southwest winds are frequent between Madeira and Cape Finisterre

,and they force the wate r to lee

ward to the north . In summer northerly winds prevail and force

NORTH ATLANTIC CURRENTS . 199

the water south . Near Dakar,in the period o f the southwest m o n

soon e o f the coast o f Africa,the Counter Equatorial Current carries

warm water ashore near the Arqu in Bank , Cape Blanco .

It is to be noticed that in the trade-wind regions the westerly co mponent Of the current is always greate r than that Of the wind .

THE SARGASSO SEA.

Southwest of the Azores the waters o f the Atlantic form a large

whirlpool analogous and corresponding to the general circulation o f

the surface winds . The m ovement is demonstrated by the drifts o f th ederelicts Te l em a ch

,Dr u ry, Wye r G. Sa rg e n t , and Fa n n i e E. Wo ls t o n .

The diam eter o f the Curves o f these tracks i s from 350 to 40 0 m iles .These drifts are evidently the resultant eff ects o f the surface winds .

For in this sea the drifts to the westward took place from July toNovem ber during the period when the trades extended farthest north

,

and to the north and east during the winter m onths,when the pre

dominant winds in that region were south and west . The agreementin the directions of the winds and the oceanic surface drift i s therefore com plete .

CURRENTS BETWEEN BERM UDA AND THE WEST INDIES .

Al l charts show in this region a prolongation Of the equatoriald r if t vto wa rd the north , and a large mass o f water passing north o f

the West Ind ies and j oining the right side o f the Gulf Stream .

Notwithstanding thi s the Pilot Charts show that the derelicts Vince n zo Pe r r o ta ,

Ida Fr a n cis,M a ry Do u g l a ss , and Ri ta were f o r several

months in that vicinity without being drifte d by that current . Inthe chart

,Drift of Bottle Papers ” N 0 . 1x ) , bottles No s . 5, 88, 10 6 ,

a n d 129 , also Show that if such a current exists at tim es it has notthe permanence attributed to it by the charts . It can be said thatthe drifts o f the floats are directly contrary t o the current indicated

on the charts .

From all these Observations we may conclude that the wind is thegreat cause f o r most of the surface movement of the sea . Its action

causes the deviations from the displacement due to the tides andglacial discharges .

TEM PERATURE OF THE SEA BETWEEN BORDEAUX AND LA PLATA .

The study Of the subm arine t em peratures has revealed the laws o f

the vertical circulation o f waters . In examining the subm arine

isotherms along a meridian o n e is at once struck by the marked incl in a t io n o f these i sotherm s near shoals , the horizontality Of the lines

in tem perate zones,and the i r rise toward the surface in warm er

regions .It is difficult to explain why the isotherm of 7° to 10 ° C .

,after

descending to a depth of t o m eters near the Canaries ,

2 0 0 CH ICAGO M ETEOROLOGICAL CONGRESS.

should ri se to nearly 2 0 0 o r 30 0 m eters below the surface,contrary to

all the laws o f gravity and to the movement o f the waters by thesurface winds . It m ust surely be due to the tropical evaporation andthe necessity o f replacing the equatorial waters that the verticalmovement acts like a vast pum p . If there is o n e place where this

movement is special ly emphasized,we believe such a place exists in

the vicinity o f Cape Blanco , Africa, near Arqu in Bank .

The observations of tem perature o f the mail steamers o f the Mess a g e r ie s Maritim es on their voyages between Bordeaux and the LaPlata ( see table below ) show that the temperature between Lisbonand the Canaries increases regularly by 4° to 6° C .

, and that thisincrease augments most rapidly in November and least in March .

Tem p er a tu r e of th e s e a be twe en Bo r d e a ux a n d th e La Pl a ta .

La t it u d e .

Mar

ch

Apri

l

.

May

June

From the Canaries to the Arqu in Bank the temperature fall s from2 ° to 235

° C . from April to Novem ber . The fal l i s less marked but i sfound al so during the winter m onths . The center o f the thermal

depression oscillates be twe e n N. 2 0 ° and N . i t runs near the coasto n the level of the peninsul a o f Ca pe Verde , and it i s deflec ted at thetime the currents change direction .

From the Arqu in Bank t o Dakar there i s a sudden thermal ri se of8° to 9° C . in Septe m ber and October ; the rise i s less in March andApril . The sharp bend in the isotherms has also been established by


NORTH ATLANTIC CURRENTS . 2 0 1

the exploring expedition Of the Ta l ism a n . On this vessel it was Ohserved that the density Of the surface waters was only or

less than that Of neighboring waters,and also that this diminution

in the saltness exists in deep layers when the temperature i s as lowas 7° C .

This low density accompanied by low mm p e r a t u r e in this warmregion proves decidedly the polar source Of these waters and their riseto the surface . The color Of this water is also different

,being green

while the neighboring trade-wind water is blue . The depths reachm eters .

The Observations o n the mail steamers show another point Of

thermal depression and rise to the surface . This i s n e a rO

Ca p e Frio .

In the warm season when the rainfall is most m arked the thermaldepression i s about 3° o r 4° C .

FROM I BORDEAUXI TO INEW'

YORK .

The Bordelaise m ail steamers plying between Bordeaux and NewYork have willingly given me the results o f their Observations forseveral years

,1882 to 1887 . The route o f these ste amers crosses the

fortieth meridian in N . 47° 80’ and passes the southern extremity Of

the Banks Of Newfoundland .

In the curve Of these i sotherms o n e i s at once struck by their irregularities between the fortieth meridian and New York

,and by

their absolute uniformity between that and the mouth of the Gironde .

Temp e r a tu r e of th e s e a f r om Bo rde a ux t o New Yo rk.

5.Eze., O

9 $4 g

0 °3 Q °3

Cu 2 <1 2

19 o 19 o 17 o

Therefore,these thermal differences near the American coast prove

the sources o f the waters to be different , while the main currents thatreach the European shores are more stationary and are pushed and

mixed by the Winds .


BAY OF B ISCAY .

In a region SO contracted there can n o t be great thermal diff erences .Such sm al l differences as exist are evidences Of the m ovem ent Of the

waters in that great bay . We have the Observations Of the mail

steamers,from tim e to time

,for a distance o f about 50 miles from

the m outh Of the Gironde to Cape’

Ortegal . A study Of the lines Ofi sotherm s

,m onth by month , shows that in the months o f June , July ,

and August there is,along the fourth m eridian west Of Greenwich

,a

mass Of water about 10 0 miles wide , Whose temperature i s about 2 °

o r 3° C . higher than that Of the French coast waters , and that from

this point to Cape Ortegal the tem perature gradually falls in thesumme r months until i t i s 2 ° C . lower than that Of the coast waters .

(See tables and plate . )This state Of things indicates that the Rennell Current does n o t

exist during the summer m onths as stated on the chart . TO supply

this m ass Of warm water along the fourth meridian,one can only

find an oceanic sou rce"

,and t o furnish the cold waters o f Cape Orte

gal it can only be ascribed t o the melting ice o f the mountains on thenorth coast Of Spain . These flow along the coast from east t o west

until t h ey m e e t the oceanic waters at Cape Finisterre .

During the months o f November and December the temperature Ofthe water i s higher at Cape Ortegal than near the mouth of theGironde . Th ese are the months Of the west winds o n the coast OfPortugal . These winds at the sam e time blow from the northwest onthe French coast and suffice to explain the thermal d ifle r e n ce .

These Observations of the temperature Of the sea clearly Show thatthe waters o f the Bay Of Biscay are diff erently affecte d in summer

and winter . During the fine season,in a period Of comparative calms

,

in the Bay the surface waters are rather stationary,and are heated

m ore than the coast wate rs by the influence Of the winds and tides .The experim ents with drifting bottles Show that in the summermonths that the coast waters are subj ect to a slight movement . TheRennell Current has n o t the permanence nor the dim ensions with

which it is credited . In the Bay Of Biscay there are variable coast

currents depending upon the force and direction of the winds .The conclusion that m ay be drawn from this study is that al l the

ocean currents marked on the charts are mor e o r less deflected by the

winds,and that even the m ost constant currents such as the Equa

t o r ia l Currents,the Gulf Stream

,and the Labrador Current are sub

ject to t heir influence . The branch currents which wash the shores

o f western Europe and Africa are even more sensitive to the winds .The currents o n charts Should be accepted as the general condition

,

but it must always be borne in mind that the wind dominates and isthe grand factor in surface movements .



p e r a t u r e s Of the waters in the neighborhood Of the Grand Banks asfar as the northern limit of the Gulf S tream .

Finally,Since this paper is addressed to the Meteorological Con

gress,they Should formulate this resolution

Th a t th e m e th o ds o f givin g in fo rm a t io n so a ccu ra t e ly e dit e d in th e Pilo t Ch a r tsh o u ld be a do pt e d by a l l th e m a rit im e n a t io n s o f Eu r o pe ; t h a t th e ir m e th o ds o f co l

le e t in g in fo rm a t io n be a do pt e d a n d t h e ch a rt s dis tr ibu t e d with th e sam e g e n e r o sity ; th a tth e de po sit o f r e po r t s fr o m l o g bo o ks be m a de Oblig a t o ry , a n d th a t th e ca pta in s be ihde m n ifie d by fr e qu e n t publica t io n s .

We think that by adopting this resolution the Congress wil l do auseful work

,and in a great degree increase the security Of l ife and

property on th e sea .

9 .— STORM S IN THE SOUTH ATLANTIC .

Ca pt . A . P. PINHEIRO.

Honored with the confidence placed in me by the General Comm it t e e Of the Congress o f Meteorology to submit a r e po r t o n thestorms Of the South Atlantic

,I have sought t o comply with this

courteous invitation with the means which I had at my disposal,by

touching upon the few data as yet known in the southern seas andtaking a general view of the observations to the present time .I,therefore

,divide this work into two parts : (1) My o wn Observa

tions from 1874 to 1893 , as to the present st ate Of thi s informationand as to what should be done for its progress and good practicalresult . (2 ) Historical Observations as to the South Atlantic stormsfrom 1789 to 1865 , covering a period o f about seventy-five years .l

1) Pe r so n a l obse rva ti o n s — Since 1874 I have always followed withinterest the South Atlantic storms

,and I Observe that the storms

follow a general direction from west to east and seek to throw themselves into the South Atlantic Ocean from the Pacific , or over theBraz il ian territory .

I,therefore

,bring before the Congress the fact that I am endeavor

ing once more,by util iz ing the recent reorganization o f the Meteoro

logical Service Of Brazil,brought about by Admiral Custodio de Mello

,

l ate Minister Of the Navy, tO realize my Old scheme Of connecting

by telegraph the service Of exchange of Observations between Chile ,Argentine Republic

,and Ur a g u av, in order t o fol low with more pre

cis io n the route o f the storms in the South Atlantic,and later o n I

shall Seek to connect the whole service Of South America and to placeit in daily correspondence with the United States Of North America .

Th e h is t o r ica l pa r t Of Ca pt . Pin h e ir o ’ s pa pe r co m pr is e s th e in t e r e st ing de scr ipt io no f s t o rm s in t h e So u t h Atla n t ic

,r e a d be fo r e th e Soci éte M étér o lo g iqu e d e Fr a n ce

'

in

1866,by M . M a r t in de M o u ss i

, a n d pr in t e d in th e An n u a ir e Of th e s o cie ty , Vo l . xrv.

1866 . pp . 15-2 2 ; o win g t o lim it e d spa ce it is n e ce ssa ry t o o m it it h e r e .— EDITOR.

STORM S IN THE SOUTH ATLANTIC . 2 0 5

When news reaches me by telegraph at Rio de Janeiro of the fallof the barometer at Valparaiso or Cordova

,I know from past e xp e

r ie n ce that after an interval o f from three and a half to six days,

according to the velocity o f the wind,bad weather may be expected

on the eastern coast Of South America , as storms follow the generaldirection from west to east, spread themselves over the interior OfBrazil

,a n d become Spent in the central states of Sao Paulo and

Minas Geraes,o r extend along the seacoast as far as the north of the

State Of E spirito Santo .

The winds o f the South Atlantic,between S . 2 3° and S . pro

ce e d generally from west, southwest, and southeast ; but n orth Ofabout S . they blow more constantly from northeast to east

,

following the configuration Of the coast to the mouth o f the Amazonas .Unfortunately the Observations made in the basin of the river arefew . On the principal river east winds predominate during thegreater part Of th e y e a r , notably in the dry season from Novem ber

until May,when they are rather strong . But in connection with the

pamperos o f the months Of June and July, that i s in the winter , theyare stil l stronger .

Upon a superficial examination Of the world we have in thenorthern hemisphere

,as well as in the southern

,movem ents o f the

air from the west toward the east . In storms these proceed withSpiral circulations ; in the former they are against the hands Of thewatch , and in the latter contrariwise . The tracks o f storms then goeastward in both hemispheres .The writer has establi shed meteorological stations Of the second

order at Belem,the capital of the State o f Grao Para, and at Manaos ,

the capital o f the State o f Amazonas,in Brazil . After a series of

Observations Shall have been made,it may be possible to connect the

general atmospheric movements o f both hemispheres . More stationswill soon be in Operation .

When I was at the International Conference in Munich two yearsago

,Mr . Wragge , Of Brisbane , Australia , had already Spoken to me

about my giving him,by telegraph

,information as to the southwest

storms which pass over Brazil and which afterward cross over tothose regions . In my opinion the meteorological service Of the world ,and more particularly the route Of the storms , will have taken agreat Step t o become be tter known , when , at a given hour , the directors o f the various meteorological services Of the world may beable to communicate with one another at a certain time (noon Greenwich ) and study the various evolutions of capricious storms over the

'

face of the terrestrial globe . A strict method in the Observations ,the mode of efie ct in g the same , the development o f the network o f

telegraphic stations,and a discussion in each of the countries where

Observations are taken In order that they may be subsequently sub


m it t e d to and appreciated by the permanent committee of mete orlogiste

,all thiswill be a great advance toward determining the general

laws o f storms in the South Atlantic , and also for those o f the wholeworld .

The small number o f South American stations , the l ittle relianceo n the Old instruments , and the manner o f setting them u p and Ob

serving them are sufficient reasons for declaring this knowledge stil lin a state Of em bryo in the South Atlantic

,and

,unfortunately

,very

little can a s yet be said upon the subj ect even after consulting thef e w reports here referred to : C . F . Martins (Amazonas ) , 1831 ; Martin de Moussi (Montevideo ) , 1843 Doazan (Buenos Ayres ) , 1866 ;Albert de Lisle (Montevideo ) , 1866 ; Manuel Eguia (Buenos Ayres ) ,1875 ; M o r sko i Sbo r n ik (Valparaiso ) , 1874 .

Let,therefore

,the above proposition go forth from the midst Of this

important conference,and f o r the realization o f my scheme I call t o

'

my assistance the Congress Auxiliary of the World ’ s Columbian Exposition .

IND 0 ? PART I .

Documents

€¦ · INTROD UCTIO N. The Congress Auxiliary of the World’s Colum bian Exposition was organized by authority and with the support of the Exposition Co r po r a tio n for the