No. 4131.

NOVEMBER 1, 1902.

Presidential AddressON

THE PATHOLOGY OF INFECTION.Delivered before the Pathological Society of London on

Oct. 21st, 1902,

BY SIR J. BURDON-SANDERSON, BART.,M.D. EDIN., F.R.C.P. LOND., F.R.S.,

REGIUS PROFESSOR OF MEDICINE, OXFORD UNIVERSITY.

[AFTER a few introductory words Sir J. Burdon- (:Sanderson proceeded :-] ]In common with the older sciences pathology seeks to

investigate the "causes of things "-the relation to eachother of the phenomena which in endless sequence con-

stitute the life of the organism, no less in disease than inhealth. 50 years ago the only part of pathology which hadreality in it was pathological anatomy. On this foundation,and more particularly on the new anatomical facts whichtowards the middle of the last century the microscope hadrevealed, Virchow had founded the new science. In his

hands and in those of the men who worked with him, firstat Wurzburg and afterwards at Berlin, the science rapidlygrew into what it has been for the past 40 years, comprisingin its widened scope not merely the structural changeswhich result from disease but the processes which producethem.

Considering that this is our first meeting since the death ofthe great master it might seem appropriate that I shouldtake as my subject the inestimable services which he has inhis long life rendered to medicine ; that I should remind youof the "cellular pathology" and of the succession of dis-coveries relating to the minute anatomy of inflammation andtubercle, to the genesis of morbid growths, to the processes ofdegeneration-in short, to all the fundamental processes ofdisease which were at that time within reach of anatomical

investigation. For several reasons I feel reluctant to

attempt this task this evening. The time at our disposal’would be inadequate and nothing that I could say would benew to you. Those who are old enough to remember the dimtwilight of fifty years ago can perhaps better realise theclaims of Virchow to our grateful remembrance thantheir successors ; but all of us who are earnestly working’for the advancement of medicine, whether at the bedsideor in the laboratory, feel that thev belong to his schooland are directly or indirectly his disciples. The observa-tions for which I ask your kind indulgence will relateto a subject which some 30 years- ago occupied muchof my attention. It had no prominent place in the"cellular pathology" which Virchow, Recklinghausen, and’Cohnheim had taught us, but nevertheless has now so closea relation to it that I felt justified in designating it on aformer occasion the "cellular pathology of to-day. 1

" Forthe encounter of disease-producing agents with the livingorganism presents to us problems which are as truly problemsof cellular pathology as are those relating to the processesof inflammation, tubercle, or cancer. The subject on whichI am about to speak is the pathology of infection. Before Iproceed further I wish to observe that had I known that myfriend Dr. W. H. Welch would make infection the subject ofthe Huxley Lecture I would have selected some other topic forthis evening. As, however, the aim I have in view is muchhumbler and very different from his I do not apprehend thatI shall say anything discordant with that admirable ex-position of the experimental data relating to immunity.The purpose I have placed before myself is to translateinto language which would have been intelligible to thepathological student of 20 years ago the technical languagewhich is unavoidable in dealing with notions which have,so to speak, sprung fresh from the laboratory and have notyet had time to clothe themselves in plain English.May I for a moment ask yo ar a tention to the form in

which the question of the nature of infection pre,enteditself at the end of the sixties 7 Lister had taught us the

1 Address to the Thirteenth International Congress, THE LANCET,August 25th, 1900, p. 563.

etiology of traumatic inflammation and of the disease whichare associated with it. From Chauveau 2 we had learnedthat in the liquid contagia of many diseases communicableby inoculation-glanders, clavelee, small-pox-the morbificagent could be separated from the liquid in which it wassuspended by mechanical means such as filtration or sub-sidence ; and it had been inferred from this that all such

’ morbific agents were particulate. Villemin 3 had shown thattuberculosis could be communicated with certainty byinoculation. I myself4 did my best to complete the workwhich Villemin had begun by exploring the channelsby which the tuberculous process is disseminated withinthe organism and demonstrating its intimate relationwith the lymphatic system. Then followed the excellentexperimental and histological work done by Klein 5

’

which did the utmost that could be accomplished bythe technical resources which were then at the disposal of

I the pathologist. In looking back at these researches IL cannot refrain from noting how many of the anatomical data. which our present methods have brought into prominence

were accurately described half a century ago by Klein.But notwithstanding these discoveries the central questionremained unanswered. It was not known whether the tuber-culous process depended for its origin on a specific infec-

L tion or owed its characteristic peculiarities to pre-existingl structural conditions, for experiment seemed to show that.

subacute inflammatory processes of non-infective origin’

affected the lymphatic system and followed the course of thelymphatic channels in the name way as tuberculous ones, so

that the latter could not be considered as more infective thanthe former. It was not until several years later that un-i equivocal evidence of the specificity of tubercle was obtained,,

not by the discovery of the eorp2cs delicti itself, for that.

did not happen until still later, but by the discovery ofSalomonsen that if a minimal quantity of tuberculousmaterial is introduced into the anterior chamber of the eyeof a rabbit no inflammatory reaction follows, but after anincubation period of two or three weeks tuberculous nodulesappear on the iris, the process eventuating in general tuber-culosis. It is not a little remarkable that at the very timethat Salomonsen was engaged with Cohnheim in these

experiments Koch was perfecting that great discoverywhich was the foundation of the bacteriological method-the discovery that the bacillus of anthrax could becultivated in successive generations outside of the bodyand that thes generations retained the power of com-

-

municating the- disease for an unlimited period. It thuscame to pass that the years 1877 and 1878 were amongthe most notable in the history of our science ; for whileSalomonsen’s discovery afforded to the many searchers afterthe specific micro-organism of tubercle-among whom oneof the foremost in this country was my predecessor in thischair, Mr. -W. Watson Cheyne-the assurance that theirsearch would eventually be successful, Koch’s discoveryfurnished the method by which that success was renderedpossible.

Let us hasten over the next eight years during which newspecific forms of micro-organisms were observed and dis-criminated from each other in rapid succession and passon to the next important step in the investigation ofthe process of infection-the discovery by Fliigge andNuttall’ 7 at Breslau and by Buchner B at Munich ofthe fact that in certain animals the liquor sanguinis is pos-sessed of alexeteric properties by virtue of which it maycontribute to the defence of the organism ; and that inanimals in which the power of resisting specific infectionshas been acquired by immunisation this power can be trans-mitted to other individuals by the serum. It does not liewithin my scope to speak of the value of this discovery asthe harbinger of a new therapeutic method. I wish only todwell on its influence on the progress of research. To the

2 Détermination Expérimentale des Éléments qui constituent lePrincipe Virulent dans le Pus Varioleux et le Pus Morveux, ComptesRendus, February 1868.

3 Etude sur la Tuberculose, Paris, 1868.4 On the Communicability of Tubercle by Inoculation; Tenth and

Eleventh Reports of the Medical Officer of the Privy Council, London,1868 and 1869.

5 Anatomy of the Lymphatic System, London, 1873-75.6 Sur l’Inoculation de la Tuberculose, spécialement à l’Iris du Lapin,

Nordiskt Medicinskt Arkiv, 1879, tome xi.7 Studien über die Abschwächung virulenter Bacterien, &c., in-

cluding Nuttall: Experimente über die bacterienfeindlichen Einflüssedes thierischen Körpers. Zeitschrift für Hygiene, 1888, Band iv.

8 Untersuchungen über die bacterienfeindlichen Wirkungen desBlutes und Blutserums, Archiv für Hygiene, 1890, Band x., p. 84.

S

1170

scientific investigator of the infective process it at once

opened new possibilities, the result of which has been thatduring the last ten years the inquiry into the nature of thealexeteric properties of the blood, whether natural or

acquired, has been carried out with extraordinary zeal andsuccess in Germany, France, Scandinavia. Russia, Italy,Japan-as well as in the laboratories of this country and ofAmerica. But the re-alts which have been gained present tothe student who desires to bring them into intelligiblerelation with each other difficulties which, if I am not mi.4-

taken, are greater than any which he has to encounter inother branches of pathology. In attempting to discuss them,however imperfectly, ( am not vain enough to imagine that Ishall be able to add to the knowledge of those members ofthe society who are conversant with the subject. Theutmost that I can hope for is to be of some use to those whoare perplexed by the multiplicity of the experimental datawhich have been presented to them and (to use a homelyexpression) who find it difficult to see the wood because ofthe trees. The main source of this difficulty is no doubt therapidity of progress-the number of new experimentalresults which present themselves to us year by year andmonth by month. A few years ago the contest seemed to bebetween infected organism and infecting microbe. Now wehave also to do with toxins and antitoxins. In the

language of a century ago, a transition has taken

place from solidism to humoralism ; and even the oldwords have been revived. Happily the change does not

depend as was wont to be the case at former periodsin the history of medicine, on a swing of the pendulumof medical opinion, but on the real progress of experimentalinvestigation. Every advance brings us nearer to the sceneof action and thus enables us to see things on a larger scaleand in more detail. In our attack on the citadel of know-

ledge the obstacles are the more formidable the nearer weapproach the breach. Far from finding the work cut out forus easier it is ten times more difficult. To the microscopicalmethods we used before we have now to add chemical ones.If we had to do with substances of known constitutionwhich could be recognised by their chemical reactions it

might be otherwise ; but as yet we are far removed fromthis knowledge. We gratefully learn from the physicist andchemist how to use exact methods and instruments. Evenas regards these we find that we must largely depend onourselves. For the methods which in the present investiga-tion have done most for us-such, e.g., as the use of the

centrifuge, filtration through porcelain, ana, to mention themost recent, the cytoclastic method contrived by Mr. SydneyRowtand—have been all devised by pathologist themselves.In the interpretation of the newly discovered properties ofthe circulating blood and of the tissue juices on whichimmunising processes depend I hope to be able to show youas I proceed that we have already received invaluable directassistance from the exact sciences, but the hi-tory of patho-logy leads us to believe that the most important work willstill be done by men who, like Virchow, Cohnheim, Lister,Koch, and Ehrlich, are themselves pathologists.The consideration of a complicated question like that of

infection can often be best entered upon from the point ofview of the prevalent notions which relate to it ; for in

general such notions contain a kernel of truth. Now thereare two points relating to infection about which all are

agreed. One of these is that when a contagium enters thehuman or animal body its encounter with the livingorganism is of the nature of a struggle between two oppos-ing tendencies. We may accept this notion of reciprocalcounteraction or antagonism as fundamental and allow itfree scope in our speculations as to the nature of infection,not merely because it is in harmony with observed facts withwhich we have been long familiar but also because it is anecessary corollary to the one biological law to which thereis no exception-the law that in the living organism everypart, every organ works together with the rest for themaintenance and efficiency of the whole and consequentlyfor the counteraction of whatever is hostile to that end.The existence in the environment of specific disease-pro-ducing poisons is a fact like that of the origin of evil in themoral world behind which we need not at present strive topenetrate. The liability to infection which this fact impliesmust. in accordance with the law of adaptation, be asso-ciated with the power of counteracting it. Our question,therefore, is not why we are provided with the means ofguarding against these risks, but how contagia act and howthe organism reacts against their attacks. The progress of

research affords ground for believing that the principle ofantagonism has an almost unlimited scope in its applica-tion to the process of infection. The old notion that

every bane has its antidote is so far true that everyinj1lriolls substance which is capable of being assimi-lated (in the physiological sense) by a living cell is also

capable of exciting in it an abnormal reaction antagonisticto the first. We have long recognised this power ofreaction in the cells of the animal body, but Dr. Walker

by his previous studies on the bacillus of typhoid feverand still more by the experiments published only theother day has given reason for believing that it isalso possessed by the bacillus itself. He has shownthat when the bacillus of typhoid fever is grown ina bacteriolytic medium, those bacilli which escape dissolu-tion acquire a higher degree of virulence. In other words,the anti-bacterial reaction of the medium excites in thebacilli which evoked it an anti-antibacterial-i.e., a pro-bacterial—reaction vihich is, of course, equivalent to anincrease of virulence.A second point about which there is also complete agree-

ment is that of speca,fieity. I have already referred to thefour or five years during which the discovery of the specificcause of tubercle was anticipated and the micro-organismitself sought for, just as the discovery of the planet Neptune-was anticipated by the astronomers Adams and Leverrier.Since then we have been furnished with equally unequivocalevidence of the specificity of many other diseases which werebefore of doubtful etiology. But the most striking result ofall has been the discovery of the close analogy between thespecificity of man and of the higher animals and that oftheir diseases. This analogy may be expressed by sayingthat species in animals and species in infective diseases havethis in common, that they can both be distinguished bycharacteristic peculiarities in the liquid part of the circu-lating blood-characteristics which might be called chemicalwere it not that, although they belong not to cells but to,their fluid environment, they are physiological or patho-logical and not such as the chemist could take cognisance of.On this interesting subject time will not allow me to enter.

If I have not occupied too much of your time with thesegeneral coniderations, I will ask your permission to pass onto the discussion of the two forms in which the infective

process presents itself-infection by toxin and infectionby bacteria In doing so I will ask you to regard these pro-cesses exclu-ively from the pathological point of view. Iam anxious that we should as far as possible confine ourattention to what happens in the tissues and structures ofthe infected organism when attacked by the infecting toxinor bacterium as the case may be. Of the first kind ofinfection diphtheria is the typical instance. Its toxic actioncan be measured with assured accuracy by determining howmuch is just enough to kill a given test animal in a giventime. A certain quantity of horse serum prepared by thenow familiar method of immunisation is proved by experi-ment to counteract that action-i. e.. in technical language,to neutralise the minimal lethal dose. This experiment,which has been repeated thousands of times, suggests theexistence in toxin and serum of two bodies which enter intochemical combination, but when we seek for the evidencethat this is so we encounter insuperable difficulty. Ifthe combination of the two antagonistic substanceswere a chemical one it would be indicated by chemi-cal reactions capable of being expressed in chemicallanguage. In judging of chemical neutralisation we

have recourse to a chemical test or indicator. Inthe present instance our indicator is not chemicalbut pathological. The reagent used is not a chemicalreagent in a test bottle, but a living guinea-pig in a cage.The indicator even if a little less exact is quite as certain.We are able to observe the pathological effects, but of thesubstances which produce them we have as yet no exact

knowledge. Later we shall see that certain concomitantphenomena of infection are already accessible to chemico-physical investigation, but this cannot yet be said to be thecase as regards the essential process of infection and counter-infection. We can best judge of this by considering whatwe know as to the essential nature of the action of the toxinof diphtheria.The quality which stands first is lethality. This we

measure in terms of the M. L. D. unit. It is a processduring which the guinea pig passes more or less graduallyfrom potential to actual death. But lethality is not all. Alltoxins in moderate dose do something eke. Side by side

1171

with the lethal process there is another process the effects ofwhich tend in the opposite direction. Its nature can bebest understood by comparing it to what in physiology iscalled stimulation. For just as the introduction of a sapidsubstance into the mouth evokes a corresponding specificsensation so when a toxin enters the organism it callsforth an equally specific reaction or response. It is

scarcely needful to point out that this response must haveits seat in the living organism or in the cells of whichit is constituted, and that. like other responses to stimu-lation, there is no constant relation between the intensity of Ithe response and that of its exciting cause. It is still more

important to notice that, although its effects extend not onlyto the living cells but to the medium in which they live, theaction itself is physiological or vital.The two properties I have endeavoured to characterise,

the lethal effect and that of inducing an antidotal or pro,tective action, are so different that it may seem superfluousto contrast them, but their assoc’at’on in the process ofinfection suggests the question of the nature of the con-nexion between them. That these two actions are whollyindependent cannot be asserted, but there are many indica-tions that they are not necessarily dependent on each other.Let us refer to one or two experimental facts. If after alethal dose of toxin has been administered to a guinea-pig adose of antitoxin slightly larger than that necessary toantagonise the toxin in vitro is given immediately the lethaleffect is averted But if the antidote is delayed, as in theexperiments of D5nitz,l the quantity of antitoxin requiredto prevent death must be correspondingly increased until

eventually no amount of it is adequate. The lethal process,therefore, is one of which you can measure the duration inminutes, whereas the reaction is a process of protracteddevelopment. It would be difficult to regard them bothas direct effects of the same cause. The completeseparateness of the one process from the other is still morestrikingly proved by a remarkable series of experiments, al-orelating to tetanus toxin, made a year ago by Dr. Ritchie.JoHe found that by subjecting tetanus toxin to the actionof very dilute hydrochloric acid for a limited period itis possible so to modify active tetanus toxin as to de-

prive it of its lethality while retaining its power of

exciting antagonistic action. He was thus ab’e to use it atonce in much larger dose than would otherwise have beenpossible and consequently to bring up his animal" veryrapidly t’J such a degree of immumty that (in one series ofexperiments) they tolerated more than a hundred times theminimal lethal dose. This very importatt result ems tome to make it impossible to question the duality of the twoactions of toxin-the lethal and the reactional-but it maystill be asked what the nature of the reaction is. It wouldbe rash to attempt a complete answer to this question butwe have in experiments such as tho- e of Donitz, to which Ireferred just now, an indication which can scarcely be mi,-taken. If when an animal is moribund, when he deathprocess is going on and would certainly very shortly end, itcan be arrested by an aotidote, it is difficult to decribp theaction of the antidote otherwise than by saying that itis anti-lethal. As to the intimate nature of the antagonism weare precluded from discussion by the circumstance that theactions which antagonise each other are only known as actions.Lethality is a property which we cannot mve"tigate in vitro.The sum of what has been said as regards the infections

that owe their origin to soluble toxins is that in every suchprocess there are two actions-respectively lethal and reac-tional-which have their seat in living cells. The latter isnot in itself protective but is able to awaken an anti-lethalreaction in the cells which come under its influence. Youwill observe that this statement embodies no theory of the

Inature of the process. It aims at setting forth whathappens in the simplest and most general terms.Just as diphtheria and tetanus have served us for the ex-

emplification of the process of infection by soluble toxins, so Iwe may take cholera as an instance of infection by microbes. -IWe choose cholera for the reason that the fundamental experi- Iment of Pfeiffer," which has led to w many important dis- I

9 Ueber das Antitoxin des Tetanus, Deutsche Medicinische Wochen-schrift, 1897, Band xxiii., p. 428.

10 Artificial Modifications of Toxins, with special reference to Immu-nity, Journal of Hygiene, 1901, vol. i., p. 125.

11 It should be noticed here that infection without microbes is anartificial process. There is no case that I know of in which a specificinfective disease is communicated in its natural unmodified form other-wise than by microbes or by a materies morbi which may containthem. Pfeiffer : Ein neues Grundgesetz der Immunitat, DeutscheMedicinische Wochenschrift, 1896, Nos. 7 and 8.

coveries, related to the cholera vibrio. I must ask to occupya few moments in recalling the conditions of that experi-ment to your attention. In the earliest experiments on infec-tive products (I refer to Klein’s and my own experiments inthe seventies) the peritoneal cavity of the guinea-pigwas used as constituting an admirable cultivation chambercontaining endothelial elements prone to proliferate and

leucocytes prepared to incorporate whatever particles werepresented to them. The value of the method is now

much greater than it was then and has acquired greatimportance. The experiment I have just mentioned consistsin this, that you introduce into the peritoneum choleravibrios of mitigated virulence in less than lethal dose. Theseare followed in a day or two by others of virulent culturesand those by successors in increasing doses. The effect ofthe proceeding is that the guinea-pig becomes "immune,"the signs of which change are (1) that fresh vibrios in manytimes the lethal dose can be introduced without lethal result ;(2) that the vibrios undergo what for the last ten years hasbeen known as bacteriolysis ; and (3) that the serum of theanimal acquires bacteriolytic and protective properties.Now with reference to this bacteriolytic serum it had alreadybeen discovered that it could be deprived of its bacteriolyticpower by moderate warming. But it was found that if theliquid so modified was introduced into the peritoneal cavityits lytic power was promptly restored ; whence it was con-cluded that the constituent of the serum which was destroyedby warming was in some form or other produced by contactwith living cells.

I have ventured to repeat to you this story, with which nodoubt every one present is familiar, because it contains as ina nutshell what is essential in the process of bacterial infec-tion, and affords the data on which the distinction betweenthe toxical and the bacterial form of the infective processcan be founded. In the case we first considered, theencounter of the organism with a soluble toxin, duringan active immunisation nothing more is required than thatthe two actions we have designated as severally lethal andreaetional should be so opposed to each other that the effectof the former may be Luore or less balanced by the reactiondue to the latter. When, as in natural infection, bacteriaappear on the scene as carriers of infection, it is necessarythat (if I may be permitted to employ teleological language)the infected organism should conform to the natural con-

ditions of the infected organism. Pfeiffer’s experiment, con-firmed as it has since been by evidence derived from othersources, affords ground for concluding that the power whichthe serum of animals immunised by his method possesses ofdissolving the vibrio is due to a constituent similar in natureto that discovered by Nuttall and Buchner several yearsbefore and for the belief that in the bacteriolysis, which inPfeiffer’s experiment took place in the peritoneal cavity,two agents took part, the first of which is clearly a specificproduct of the colli ion between bacteria and living tissue-i.e., of the protective reaction of the latter-the other pre-existing and non-specific, a normal constituent of livingcells.

’ These conjectures would not have assumed the definite

form that we are now able to assign to them had it not beenthat about the same time another line of investigation wasopened which promises to be as fruitful as that of Pfeiffer.It was di-coverert that the toxic effects produced by alienblood disks (i.e., blood disks of an animal of a different

species) when introduced into the living body correspondin many remarkable particulars with those of morbificbacteria. In this connexion time will not permit me to domore than to mention the very beautiful series ot experimentsby which Bordet 1’ first demonstrated that the action of thehsemolytic serum thus obtained depends on two constituents,both of intracellular origin, one of which is specific and wasdesignated by him" substance se2sibzli.scctriee," the othernon-specific, the analogues of the two anti-bacterial productsmentioned just now. The former is also called MMMMMMMC, .

a word which expresses perhaps better than any other itsspecific property of preparing the blood-disk or the bacterium,as the case may be, for the lytic action of the second con-stituent, called by Bordet the alexine and by Ehrlich theco7rcplement. This we will call lysine. The immense valueof this experiment of Bordet conisted in this, that it addedclearness to the interpretation that had already been

given ’ f Pfeiffer’s reaction, showing that the power whichblood disks and bacteria have in common of resi.ting

12 Les Sérums Hémolytiques, Annales de l’Institut Pasteur, 1900,tome xiv., p. 257.

1172

the solvent (lytic) action of blood and tis!ues is annulled

by the specific protective action called into existence

by the penetration into the infected human or animal

organism of alien blood disks or of bacterid, as the case

may be. I will ask you to take for granted the experimentalevidence by which during the last two years these conclusionshave been established and their meaning extended, in orderthat I may claim your attention to the recently conductedinvestigations relating to the nature and mode of action ofheamolysine at Copenhagen. The results of these inquirieshave just appeared in the inaugural report of the SerumInstitution just opened at Copenhagen, which contains aseries of most important papers-written in English. It is

impossible to examine this volume without being impressedby the fact that the country that has given us our Queenoffers us in our own language contributions to pathologicalscience of an excellence which we shall find it difficult toequal. I can only refer to the two papers with which thevolume opens. Two eminent men-Arrhenius, the well-known professor of physics at Stockholm, and Madsenwhose work on immunity is familiar to us-have, if I may so

express myself, put their heads together for the solution offundamental questions relations to haemolysis. Before Iendeavour to explain their scope I must remind you thatthe property of dissolving the coloured blood corpuscles ispossessed by many bacterial products and particularly bythe toxin of tetanus and that it was discovered by Ehrlichsome years ago that the hsemolytic constituent of this toxinwas different from the " spastic " and that anti-lytic serumcould be obtained by the ordinary process of immunisation.One of the most important of the Danish investigationsrelates to the question whether the neutralisation of this

tetanolysine by anti-Iyine conforms as regards the physicalconditions under which it takes place to the process ofneutralisation of an acid by a base. The answer to the ques-tion is in the affirmative. It is demonstrated that the modeof action of anti-lysine on lysine can be deduced by a processof calculation and that the experimental results agree satis-factorily with the calculated ones. For us the essentialpoint is that the lytic process which takes so important apart in the defence of the organism against infection admitsof exact investigation. At first sight this result mightappear to be of little moment, for lytic action, although anessential concomitant of anti-lethal action, cannot be identi-fied with it. But in reality it is of the utmost significance,for the proof that the :nteraction of lysine and anti-lysine isof the same nature as chemical combination brings us onestep nearer to the understanding of the fundamental

antagonism between lethal and anti lethal action. Therelation between the question of haemolysis and that of thegeneral action of toxin may be stated as follows. We havetwo kinds of poisons concerned in the production of specificdiseases or of morbid states. Toxins which act on parti-cular cells, as, e.g., the coloured blood disks ; toxins whichact on the whole body. The former have been brought bythe chemico.physical researches of Madsen and Arrhenius,for which thoe of Ehrlich paved the way, within the scopeof exact scientific investigation. The latter are not as yetamenable to direct investigation of’ the same kind, but theclose relation of these toxins to the others justifies theinference that the antagonism between them and their

opponents may be dependent on chemical interaction.I must now ask your attention to one more experiment

relating to hoemolysis. The investigation in question is thatpublished a short time ago by Dr. W. Bulloch.]3 3 Dr.Bulloch’s method may be compared with Professor Pfeiffer’s,with the substitution of blood disks for cholera vibrios. Itconsists in the injection of repeated doses of the blood disksof the ox into the peritoneal cavity of the rabbit so as torender its blood more and more hmmolytic. Serum was col-lected at frequent intervals so as to observe the effect of theinjections (1) on its hasmolytic power; and (2) on the numberof leucocytes it contained. The hsemolytic power was esti-mated by measuring in vitro the quantity of ox blood diskswhich it was capable of disposing of. From this datumDr. Bulloch was able. by taking advantage of the fact thatthe specific immunibine was always in excess, to determinethe variaticns of that constituent by adding an excess oflysine and then making a second measurement of the

hsemolytic power of the mixture. In this way, although,of course, no measurement could be obtained of the quan-tities of the two constituents, the variations of hasmolytic

13 Ueber die Beziehung zwischen Hämolyse und BacteriolyseCentralblatt für Bacteriologie, Band xxix., p. 732.

activity during the period of infection could be estimated.Under different conditions the result of these experimentswas most remarkable. Each treatment of the rabbit withox blood disks was followed after a period of incubation ofseveral days by a rapid increase of the specific immunisine,this being ushered in by an increase in the number of leuco-cytes, particularly of the lymphocytes. In all the rabbits usedin these experiments the serum obtained contained more ofthe specific constituent than was required to utilise the com-plement, and it was found that this excess was increasedafter every introduction of a fresh dose of the obnoxiousblood disks without any increase of the lysine, so that inorder to obtain ’the full hsemolytic effect it was necessaryto supply the lack of complement by the addition of serumfrom an intact animal.The fact that the production and gradual augmentation of

the specific property which must unquestionably be regardedas the essence of immunisation were preceded and accom-panied by leucocytosis recalls our attention to those lines ofinquiry which relate to the part played by leucocytes in theprocesses we have been considering. I do not suppose thatanyone questions that the constituents of bacteriolytic andbasmolytic serum are substances of cellular origin. As

regards the lytic constituent, Metchnikoff and his schoolhave abundantly shown that although both bacteriolysis andhaemolysis can take place without the direct intervention ofleucocytes endowed with the faculty of phagocytosis, thesestructures take an important part in bacteriolysis. In con-nexion with this subject I may be permitted to remind youthat during this year a communication was made to thissociety by Professor Muir of Glasgow on the Reactions ofLeucocyte-forming Tissues in Infections, in which he con-clu,ively showed that the leucocytosis which is observed inacute inflammations was associated with changes of a veryremarkable kind in the marrow of bone. These changes,which he appropriately designates reactional, are charac-terised by the conversion of yellow into red marrow and theabundant proliferation of leucocyte-producing tissue. Pro-fessor Muir’s observations further tend to ,show that thisI -reaction " is for the most part limited to the medullarytissue, neither spleen nor lymphatic glands taking part init. I must not omit to mention that although no one has,so far as I know, investigated the subject so satisfactorilyas Professor Muir, several pathologists, particularly Dr.Roger of Paris, have described changes in the marrow,accompanying acute infective processes, of the same natureas those to which I have just referred.What I have said may suffice to show that we have in

these secondary infective processes in the tissues which pro-duce blood disks and leucocytes on the one hand and in thelymphatic system on the other a lroductive field of researchwhich can be worked both chemically and experimentally-a field which is accessible to all who are thoroughly versedin histological methods In this field what has already beendone will afford indications of other lines of investigation ofequal interest.

In conclusion, may I express the hope that the veryimperfect outline of my subject I have been able to give maybe found to be conformable, not only to the experimental’facts to which it has been possible to refer, but to the muchmore numerous and not less important ones which I havebeen compelled to omit ? 7 In dealing with scientific progress.there is some risk in looking forward ; but a certain amountof looking forward is not only permissible but demanded.I confess myself wholly unable to forecast the discoveriesof the next few years. There seem to me, however, to betwo lines of advance for which two kinds of investiga-tion are required. The one thing needful is to continuethe investigation of the origin and nature of infective

processes by experiments on animals and by the clinicalstudy of the anatomical and functional changes which thespecific infections produce in organs and tissues, and theprotective reactions which are associated with them. But inaddition to this it is of great importance that the methods ofphysical chemistry should be applied to the interpretation ofthe experimental data. Of investigations of the first kind Ihave given a few instances out of many this evening. Ofthe second kind I could offer you no better example than the-splendid research conducted by Madsen and Arrhenius.

DRAINAGE OF TRURO.-On Oct. 21st an inquirywas held at Truro into an application made by the Truro,corporation for permission to borrow .621,000 for works oisewerage and sewage disposal.