DepartmentTHE DISTINGUISHING CHARACTERISTICS OF LACTOBACILLUS ACIDOPHILUS HAROLDR. CURRAN, L. A. ROGERS AND E. 0. WHITTIER Research Laboratories, Bureau of Dairy Industry, United States

THE DISTINGUISHING CHARACTERISTICS OFLACTOBACILLUS ACIDOPHILUS

HAROLD R. CURRAN, L. A. ROGERS AND E. 0. WHITTIERResearch Laboratories, Bureau of Dairy Industry, United States Department

of Agriculture

Received for publication September 28, 1932

It requires only a casual examination of the literature to showthat there is much confusion in regard to the differentiation ofthe members of the lactobacillus group. While bacteriologistswho have worked with this group may have quite definite con-cepts of the various species it is difficult to translate those con-cepts into concrete terms which define a biological species. Thisis especially true of the acidophilus cultures, since they have beenselected more on the basis of their ability to grow under certainprescribed conditions than because of compliance with taxonomiccriteria.Much of the confusion in bacteriological nomenclature is due to

the use of a limited number of cultures in determining a basis fordescribing species and to reliance on characters which are incon-stant and have little taxonomic significance. The natural group-ing of bacteria is based on fundamental physiological characterswhich the members of these groups possess in common. Corre-lated with these are more superficial, and sometimes more obvious,characters which may permit an easy identification. In establish-ing a classification it is clear that the emphasis should be placedmore on the stable basic characters and less on those which aresuperficial.The isolation of cultures for identification is more or less a

matter of chance and, if only a few are studied, there is always apossibility that these are not representative of the species as itoccurs in nature. We have used as many cultures as the laboriousnature of many of the determinations would permit.

595

on August 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

596 HAROLD R. CURRAN, L. A. ROGERS AND E. 0. WHITTIER

The 103 cultures on which complete, or nearly complete, deter-minations were made were obtained through the courtesy oflaboratories in different parts of the country. With the exceptionof a few cultures, presumably bulgaricus, and cultures of dentaland vaginal origin, nearly all were labelled acidophilus. Theorigins of these cultures in so far as they could be traced are givenin table 1.

iMlicroscopically all the cultures were Gram-positive, non-sporulating, non-motile rods, varying in size when grown in milkfrom short rods to long filaments. Acid was always produced inmilk with no evidence of gas formation. Colony form varied

TABLE 1

Source of cultures

NUMBER OFCULTURES

Human vagina................................................ 9Intestine, human or rat....................................... 23Dental................................................... 24Acidophilus preparations ...................................... 13Unknown ................................................. 34

Total ....................................................... 103

from the fuzzy and filamentous to the regular smooth sand-graintype with all intermediate gradations. Before use each culturewas plated on tomato peptone agar and incubated at 370C. in anatmosphere of 10 per cent carbon dioxide. After forty-eighthours well isolated colonies were examined with the 16 mm.objective. Film preparations from these colonies were stainedby the Gram method and examined. If no evidence of contamina-tion was then apparent the remainder of the colony was trans-ferred to sterile litmus milk, subcultures of which were used laterfor the experimental work. These cultures were replated atfrequent intervals in order to eliminate possible contamination.Throughout. every experiment each control and test culture wasexamined microscopically before and after incubation to verifypurity and positive growth.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

CHARACTERISTICS OF LACTOBACILLUS ACIDOPHILUS 597

In this study the following cultural and physiological charactershave received attention: total and volatile acid production,nature of the volatile acids, form of lactic acid, limiting undisso-ciated lactic acid, sugar fermentations, carbon dioxide production,limiting growth temperatures, limiting phenol and indol concen-trations, and surface tension relationships.

TOTAL AND VOLATILE ACID PRODUCTION

Except for carbohydrate fermentations most of the previousbiochemical studies upon the lactobacilli have been limited to theprincipal non-volatile product of fermentation, viz., lactic acid.Very few published data are available concerning the volatileacid production of these organisms. Heinemann and Hefferan(1909) found that B. bulgaricus produced about 6 per cent ofvolatile acid in milk. Barthel (1913) studied 10 aciduric bac-teria isolated from a variety of sources, and found that the amountof volatile acid formed, expressed in per cent of the total acids,varied in the different cultures from 1.1 to 7.1 per cent. Hartet al. (1914) studied the action of certain bacteria concerned inthe ripening of Cheddar cheese. When 2 strains of Bact. caseiwere incubated for seven months in milk most of the total acidthen present was volatile. Kopeloff (1926) refers to unpublishedanalyses by Zoller in which the total volatile acid approximated5 to 10 per cent of the combined volatile and non-volatile acids.Apparently this work was limited to a small number of authenticacidophilus cultures.

In our study of volatile acid production the following procedurewas adopted: 260 cc. of Swiss cheese whey was inoculated with 1cc. of a forty-eight-hour milk culture and incubated at 37°C.After six days, 10 cc. of the culture was withdrawn and titratedfor total acidity with 0.1 N sodium hydroxide. The remainderwas made acid to Congo red paper with phosphoric acid and dis-tilled with steam until 1 liter was obtained. An aliquot of thiswas titrated with 0.1 N NaOH and the remainder made alkaline tophenolphthalein with NaOH and evaporated to dryness over asteam bath. The residue was collected in 10 cc. of distilled waterand used for the identification of the volatile acids.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


The volatile acid production of our cultures is representedgraphically in figure 1. The strains are arranged according tothe proportion of total acid which is volatile. It is evident thatmost of the cultures are concentrated in a small volatile acid range.In 50 per cent of the cultures the percentage volatile of the total26_

22Z!#:~-Ele,1_.cE8

20 ..2.0 FIG. 2

Kcv~~~/8 /~~8 aep

6'~~~~~~~~~~~6

FIG.7/1 FIG.

FIG. 1. DISTRIBUTION OF CULTURES ON THE BASIS OF VOLATILE ACID PRODUCTIONFIG. 2. DISTRIBUTION OF CULTURES ON THE BASIS OF LIMIITING UNDISSOCIATED

LACTIC ACID CONCENTRATION

acid was between 10 and 20; in 38 per cent between 4 and 10.In only 8 per cent of the total cultures was the volatile acid lessthan 4 or over 20 per cent of the total acid. The two strains inwhich over 30 per cent of the total acid was volatile were obvi-ously out of place in a collection of acidophilus cultures. The


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


compactness and single peak character of the distribution curve isevidence of a certain uniformity in the amount of volatile acidproduced. Upon the basis of volatile acid production thereforethese cultures are relatively homogenous and can not be dividedinto well defined groups. As will be shown later, however, thequantitative production of volatile acid correlates well with manyother characters and lends itself readily to a classification basedupon certain of these criteria.The total acidity produced in these cultures varied from 0.227 to

2.70 per cent, calculated as lactic acid. In over 75 per cent of thecultures the titratable acidity was between 0.5 and 1.5 per cent.Approximately 20 per cent produced less than 0.5 per cent acidwhile very few yielded more than 1.50 per cent total acid. Thelow acid group is somewhat larger than might be expected and tosome readers the absence of appreciable numbers of high acidformers in a collection of this size may occasion some surprise.The titratable acidity and total volatile acidity increased only

slightly after the first four days. That the ratio between the totalacidity and total volatile acidity is practically constant is indi-cated by the fact that determinations on new samples after thelapse of one year showed only small variations from those orig-inally found.

NATURE OF VOLATILE ACIDS

The positive identification and estimation of individual mem-bers of the volatile acid series is not readily accomplished. Thepartition method recently proposed by Werkman (1930) does notdifferentiate accurately between formic and acetic acids both ofwhich were formed in our cultures. The Duclaux method offractional distillation, though admittedly inexact and subject tolimitations, is capable nevertheless of yielding valuable compara-tive information and was therefore used in this study.The alkaline residue obtained in the quantitative estimation of

volatile acids was made acid to Congo red by the addition of con-centrated phosphoric acid; the solution was then diluted to 110 cc.with distilled water and fractionally distilled in 20 cc. incrementsaccording to the method of Duclaux (1A900). The separate por-tions were then titrated and the results calculated as percentages


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


of the distillate in 100 cc. The titration constants for the fivefractions showed a remarkable constancy for the entire group ofcultures. These figures are omitted to economize space. Over90 per cent of the cultures upon Duclaux distillation gave distilla-tion constants indicating a mixture of the same acids in the sameratio. This surprising similarity of the titration figures affordsunmistakable evidence that the cultures produced the samecombination of volatile acids. It is equallyT clear that this phiaseof their metabolism offers no satisfactory basis for differentiation.The distillation constants fail uniformly to coincide with any twoacid combinations in the Duclaux table. This lack of corre-spondence with known solutions suggested a complex mixture ofthree or more acids. The titration constants most nearly ap-proached those for formic, acetic and butyric acids and as theconstants for formic and acetic acid fall very close together itseemed likely that the volatile mixture was made up of these threeacids. As this evidence was only presumptive, positive methodsof analyses were then sought. The Jones method for the identi-fication of formic acid based upon the reducing action of thealdehyde radical was used for a time but later abandoned when itwas found that the sterile uninoculated whey medium containednon-acid volatile reducing substances. The mercuric oxide testfor formic acid which depends upon the solubility and stability toheat of mercuric formate was finally chosen to demonstrate thepresence of tlhis acid. This test was applied as follows: Thevolatile acid distillate was warmed to 40°C. with an excess ofHgO and filtered through paper. The filtrate was then boiled ina water bath which decomposed the mercuric formate yielding aheavy grey precipitate of metallic Hg. Every culture gave apositive reaction to this test which was accepted as positive proofof the presence of formic acid in these distillates. The identifica-tion of the other volatile acids was accomplished by repeatedfractional distillation of the volatile acid mixture and observationof the boiling points. In order to provide a sufficient quantity ofthe acids for this purpose all the individual volatile acid distillateswere combined. A composite 3-liter sample was extracted withether and the aqueous and ether phases were separated by means


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


of a separatory funnel. The ether was then distilled off and thewater-free volatile acid residue fractionally distilled until rela-tively pure samples were obtained. The odor, relative quantity,and boiling points of the samples indicated the presence of formic,acetic and butyric acids in the proportion of 6:3: 1. These valuesare necessarily approximate and probably vary at least 15 or 20per cent with the individual cultures. These results, confirmingas they do previous deductions, constitute, we believe, satisfactoryevidence of the identity of the volatile acids generated in thesecultures.

FORM OF LACTIC ACID

It seems probable that whenever a cell produces lactic acid, itproduces both the dextro and the levo form, at least ultimately.The ratios between the quantities of the two forms vary with aconsiderable number of factors, among which have been includedkind of sugar fermented, temperature of incubation, the species oflactic organism used, and the presence of organisms which do notproduce lactic acid (Pederson, Peterson, and Fred (1926) andreferences which they give to earlier work). Pure cultures grownunder the same cultural conditions produce consistently the sameratio of the two forms of acid and it is for this reason that thismethod was tried as a basis for differentiation of species in theacidophilus group. Kopeloff and Bass (1927) have reported thatthree cultures which they designated L. acidophilus gave dextro-rotatory acid in excess of levo.The method used in our work was that of Pederson, Peterson,

and Fred (1926), with modifications suggested by Dr. Fred.In order to obtain sufficient acid for analysis the following pro-

cedure was adopted: Swiss cheese whey was concentrated in avacuum pan until a large part of the lactose crystallized out.This sugar was removed by centrifugation and the liquor restoredto the original volume with water. The final product was conse-quently normal whey with a greatly reduced sugar content. Twoliters of this reconstituted whey were inoculated with 1 cc. of aforty-eight-hour milk culture and incubated at 37°C. After sixdays the culture was made acid to Congo red paper with phos-

NEW YORK HOMEOPATHICMEDICAI COLLEGE ANDFLOWER HOSPITALPRENTISS LIBRARY


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


phoric acid and distilled until all of the volatile acids wereliberated. This residue was further concentrated to a volume ofabout 100 cc., maintained acid to thymol blue with phosphoricacid and extracted continuously for seventy-two hours with eitherisopropyl ether or a mixture of isopropyl and ethyl ethers. Ap-proximately 25 cc. of water was added to the extract and the etherremoved by distillation. The water solution was then made up to100 cc. and a 5 cc. aliquot titrated to determine the quantity ofextracted acid. The remainder of the solution was boiled forten minutes or more with an excess of alkali-free zinc carbonate.The excess carbonate was removed by filtration and the filtrateevaporated on a steam bath until crystallization began. Ethanolwas then added to give a concentration of 50 per cent, the solutionwell stirred and kept over night at 10°C. The zinc lactate wasthen filtered off, washed with 95 per cent ethanol several times,then with ethyl ether, air-dried and bottled.

In order to determine whether the lactate was optically ac-tive or inactive, it was necessary to determine the water of crys-tallization. The active salt has two molecules of water ofcrystallization or 12.89 per cent; the inactive, three molecules,or 18.18 per cent. The percentage of inactive salt may be cal-culated from the formula

100 X per cent H20 -12.8918.18 - 12.89

A 2-gram sample was weighed in a moisture dish, dried over CaCl2to constant weight and the water of crystallization driven off bydrying to constant weight in an oven at 110°C.The results obtained were very clear cut as a basis for sharply

grouping the 96 cultures examined. For 69 cultures, the deter-minations indicated practically pure inactive salt which denotesthe presence of the acids in equal quantities. Of the remaining27 cultures, 26 contained from 65 to 95 per cent of active salt inexcess and one contained 52 per cent. In all 27 cases the rota-tions of the active salts were levo; therefore, the acid in excess wasdextro-rotatory.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


LIMITING UNDISSOCIATED LACTIC ACID CONCENTRATION

In the earlier years of bacteriology, the cessation of fermenta-tion by acid-producing organisms was assumed to be determinedby a limiting gross concentration of acid. More recently, with abetter understanding of hydrogen-ion concentration and bufferaction, the actual limiting factor was believed to be the concen-tration of the hydrogen ions. Hydrogen ion concentration in amedium is a function of concentration of titratable acid and ofcharacteristic intensity and concentration of buffer substancespresent. Undoubtedly hydrogen ion concentration is of itself thelimiting factor under usual conditions. However, still morerecently (see review by Kolthoff, 1925) the undissociated formsof acids have been shown to be inhibitory. The concentrationof the undissociated form is a function of the hydrogen ion con-centration and of the total concentration of the acid present.Van Dam (1918) showed that in S. lactis cultures in whey to

which differing amounts of neutral sodium lactate have beenadded, fermentation ceased at different pH values. The calcu-lated concentration of undissociated lactic acid was shown to bethe constant limiting factor under these conditions. This limit-ing concentration varies with the medium used, but appears tobe a reproducibly constant characteristic of a species of organism,since Rogers and Whittier (1928) were able to check Van Dam'svalues for S. lactis in whey. It was hoped that limiting concen-tration of undissociated lactic acid might serve to differentiatethe lactobacilli from one another. Differences in limiting valuesfor our laboratory cultures were found of sufficient magnitude tojustify making the necessary determinations on our collection oflactobacilli.The equation used in the calculation is the following:

(H lactate) =(total lactate)

+ a (H+)

in which K is the dissociation constant of lactic acid, 14.7 X10-5, a is the degree of dissociation of sodium lactate, varyingfrom 0.76 to 0.80 for the concentrations used; and the brackets


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


signify concentrations of the substances bracketed in mols perliter. The determinations needed are of total lactate concentra-tion and of hydrogen-ion concentration.To three flasks containing 100 cc. of Swiss cheese whey were

added respectively, 2, 3 and 4 cc. of approximately 6 M solution ofneutral sodium lactate. The flasks were plugged with cottonand autoclaved for twenty minutes at 120°C. After being cooled,each flask was inoculated with 0.5 cc. of a vigorous culture of theorganism being tested. The flasks were then incubated at 27'C.After four days, the cotton plugs were removed and sterile rubberstoppers inserted in the necks of the flasks. Cotton plugs permitevaporation of the whey which affects the final results. After atleast two weeks' incubation, the hydrogen-ion concentration wasdetermined by means of the quinhydrone electrode and the totallactate concentration by the method of Friedemann, Cotonio,and Shaffer (1927). The sample was filtered and the filtrate useddirectly for the lactate determination without other treatment.The undissociated lactic acid concentrations calculated fromthese determinations usually agreed within five units in the thirddecimal place.

Figure 2 shows the distribution of these cultures, dependingupon their limiting concentration of undissociated lactic acid.As may be seen, three rather distinct groups are indicated, one fora very low concentration of acid varying from 0.020 to 0.040moles per liter, one between 0.060 and 0.120 and one above 0.140.Thus, these cultures are not checked uniformly by a given coii-centration of the undissociated acid but apparently representthree distinct groups in which the limiting concentration of undis-sociated lactic acid varies within a considerable range.

Limiting pH values for all the cultures were determined in skimmilk. These values correlated with limiting undissociated lac-tic acid as might be expected, but very grossly. There was Inonoticeable correlation with other determined characteristics andconsequently no report is made of the results obtained.

FERMENTATION OF CARBOHYDRATES

The value of carbohydrate reactions in the classification ofaciduric bacteria is a question upon which there is no general


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


agreement. Rahe (1918) proposed a classification of these organ-isms based upon the fermentation of glucose, maltose, lactose,sucrose and raffinose, Kulp and Rettger (1924) differentiatedL. acidophilus and L. bulgaricus by their ability to attack maltose,sucrose and unheated levulose. Albus and Holm (1926) declaredthat sugar reactions were unreliable for the separation of theseorganisms. Day and Gibbs (1928) also failed to confirm theresults of Kulp and Rettger and concluded that the fermentationof maltose, sucrose, and levulose does not offer a means of dis-tinguishing between L. acidophilus and L. bulgaricus.The fermentation reactions of Doderlein's vaginal bacilli were

compared with those of the intestinal or acidophilus types byJoetten (1922), Lash and Kaplan (1926), Thomas (1928), andothers. In general, no consistent differences were observed.Dental and intestinal strains have been compared directly byMlorishita (1928); Rosebury, Linton, and Buchbinder (1929);Hadley, Bunting and Delves (1930); Howitt (1930); and manyothers. Of these, only Morishita found significant differences innature of the carbohydrates utilized.The basic medium used in our work consisted of casein-digest

broth prepared from C. P. casein by the tryptic digestion methoddescribed by Kulp and Rettger (1924). Ten per cent solutionsof the test substance were made up in distilled water. Levulosesolution was sterilized by filtration through a Berkefeld filtercandle, while all the other sugar solutions were sterilized by steamunder pressure. These were added aseptically to the sterilebasic medium in sufficient quantity to yield a final concentrationof 1 per cent. In order to furnish more favorable growth condi-tions each 10 cc. of the medium also received 3 drops of sterilehorse serum. This serum was heated to 60°C. for one-half hourprior to its use in the maltose solutions (TenBroek 1920). Thefinal reactions of all the media were between pH 6.8 and 7.1.The inoculum was prepared by growing the culture at 37°C.

in casein digest broth containing 0.1 per cent glucose. At theend of twenty-four hours the organisms were sedimented bycentrifugation, the supernatant fluid was poured off, and the cellswere resuspended in sterile distilled water. This process wasrepeated several times, after which the washed cells were diluted


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


with sterile water to an arbitrary density comparable with thatof a standard suspension of barium carbonate. One-tenth cubiccentimeter of this suspension was used to seed 10-cc. portions ofthe test medium. After ten days' incubation at 37°C. the hydro-gen-ion concentration was determined colorimetrically by themethod described by Brown (1924). A distinct change in pHfrom that of the inoculated sugar-free control was accepted asevidence of carbohydrate utilization.The study of the fermentative activity of these cultures was

made upon the following substances: Lactose, maltose, sucrose,unheated levulose, raffinose, salicin, mannitol. They are ar-ranged in the order of availability. With one or two exceptionsall the cultures fermented lactose; sucrose, maltose, and unheatedlevulose were utilized with very few exceptions. Raffinose wasattacked by about 60 per cent of the cultures and salicin by morethan half. Acid was produced from mannitol by approximately29 per cent. It would appear that these reactions remain fairlyconstant from the fact that 30 of the cultures retested after thelapse of one year showed only slight variations in the final pHvalue. In none of these cultures was the ability to ferment agiven substance either lost or acquired.

CARBON DIOXIDE PRODUCTION

Carbon dioxide appears as one of the end products in the de-composition of carbohydrates by most bacteria. It may befound in exceedingly small quantity or occur as one of the majorproducts of the fermentation. This particular phase in themetabolism of the lactobacilli has received very little attentionin the literature. Eldredge and Rogers (1914) isolated and stud-ied a large number of cultures from Emmental cheese. While noattempt was made to identify the organisms, the recorded cul-tural and physiological characters indicate that manv belongedto lactobacilli of the casei-bulgaricus group. Approximately 95per cent of the lactose-fermenting rods produced carboin dioxidein varying amounts in sugar-free whey. Sherman (1921) studiedthe CO2 production of 38 cultures of aciduric bacteria. Thiscollection included strains of L. acidophilus, L. bulgaricus, and


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


L. bifidus, as well as unidentified cultures from intestinal contents,silage, milk, and cheese. Many of these organisms were foundto produce significant amounts of CO2 and the bulgaricus typesproduced less than the intestinal forms. Fred, Peterson, andStiles (1925) found that the granulated lactic acid bacteria fromcereals formed little or none of this gas in yeast broth containingglucose, fructose, or lactose.Carbon dioxide was measured quantitatively by the following

method: Twenty cc. of unfiltered Swiss cheese whey was steril-

14

/2 /z A ~~~~67eGoalp -

/6'

K,

7/

Ng6WX X

FIG. 3. PRODUCTION OF CARBON DIOXIDE GAS

ized in Eldredge tubes. These were immediately cooled to 37°C.and inoculated with 0.1 cc. of a 48-hour milk culture. Ten cubiccentimeters of Ba(OH)2 were then measured into each tube toabsorb the C02, after which the cultures were incubated at 37°C.Absorption of atmospheric CO2 was prevented by sealing theupright tubes after inoculation. At the end of five days theexcess of Ba(OH)2 was titrated with 0.1 N oxalic acid and theCO2 calculated from the difference. The quantity of CO2 wasexpressed in cubic centimeters of N/10 Ba(OH)2 neutralized.

Figure 3 shows the range of CO2 produced in these cultures andtheir distribution in the different zones. As may be seen the


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


cultures fall largely into two general groups, those forming 4 cc.or less and those forming from 4 to 7 cc.

LIMITING GROWTH TEMPERATURE S

The temperature relations were established by heavily inocu-lating litmus milk tubes and holding them in incubators at 100,15°, 20° and 300C. and in water baths at 370, 400, 430, 460, 480500 and 520C. Growth was determined by observing changes inthe color of the litmus. At the low temperatures the cultureswere held for two months. The optimum growth temperaturewas between 370 and 400C. and was uniform for the entire collec-tion. None of the cultures grew at 52° and only a very few at50°C. The upper limit for a large number of the cultures was460, and many grew at 430 and not at 460C. In the lower rangethere was a sharp differentiationi, with 79 per cent failing to showevidence of growth in milk at 200C. The remainder grew at 20°and many of these continued to grow at 15° and even 100C.

LI'MITING PHENOL AND INDOL CONCENTRATIONS

The work of Kulp (1929) introduced an entirely new meansof separating L. acidophlilus and L. bulgaricus, based upon theirrelative sensitiveness to indol and phenol. Concentrations ofthese substances completely inhibitory to L. bulgaricus exercisedno restraining action upon L. acidophilus. While his study waslimited to comparatively few cultures, the results seem to justifythe conclusion that typical L. bulgaricus and L. acidophilus maybe differentiated by the determination of their tolerance for indoland phenol. As both of these substances are present in appreci-able amounts in the intestinal canal of man and animals, it isnot illogical to suppose that they may play a role in determiningthe implantability of lactobacilli.The plate method and technique described by Kulp (1929) was

exactly duplicated in all details. Tomato peptone agar (Kulp1927) was the nutrient medium and the test reagents were Kahl-baum, C.P. quality.The relative tolerance of these cultures to phenol and indol is

shown in figures 4 and 5. The distribution curve for phenol


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


(fig. 4) reveals three rather distinct peaks indicating that thecultures were not alike in their response to different concentra-tions of this substance but may be divided on this basis into two

26

241 .... 201/P X

Lcneou1P C'26~~~~j ~ ../_20 |EG0/

/2~~~~~~~~~~4/< 6 . . . - zGO9

0/0

.oA 1/770S/

FIG. 4 FIG. 5FIG. 4. DISTRIBUJTION OF CUJLTURES ON THE BASIS OF THEIR ABILITY TO GROW

IN THE PRESENCE OF PHENOLFIG. 5. DISTRIBUTION OF CULTUJRES ON THE BASIS OF AB3ILITY TO GROW IN THE

PRESENCE OF INDOL

or more groups. About 75 per cent of the cultures were inhibitedby a phenol concentration of 1 :450. Sixteen per cent grewbetween 1:00 and 1:550, while the remainder constitutes a small


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


group which grew in concentrations as high as 1: 750. The limit-ing concentrations of indol (fig. 5) present a similar picture; thedistribution of cultures as in the preceding figure shows two orthree distinct divisions. The first group comprises over 60 percent of the cultures, all of which failed to grow above 1:2100.The other two groups are somewhat larger and more distinctthan was true in the limiting phenol concentrations.

SURFACE TENSION

Albus and Holm (1925, 1926) were the first to study the growthof L. acidophilus and L. bulgaricus in media of lowered surfacetension. Using the drop-weight method for measuring surfacetension they found that L. bulgaricus did not grow when the sur-face tension of the medium was depressed by sodium ricinoleateto less than 40 dynes. The authors in consequence concludedthat surface tension offers a valuable means of distinguishingbetween L. acidophilus and L. bulgaricus. Kopeloff and Beer-man (1927) published results which substantiated these findings.Hyde and Hammer (1927) studied 12 so-called L. acidophiluscultures and obtained variable results with respect to growth inmedia at a surface tension of 37.4 dynes. Day and Gibbs (1928),attempting to throw more light on surface tension as a factor inthe development of aciduric bacteria, studied 20 strains of L.acidophilus and L. bulgaricus from authentic sources. Sodiumricinoleate was found to exert a toxic action upon both bulgaricusand acidophilus strains but they concluded that this action couldnot be employed as a means of identifying the two organisms.As these authors have pointed out, the weak link in the evidencewhich regards surface tension as a criterion for differentiatingbetween L. acidophilus and L. bulgaricus is that different soapsused as depressants do not give uniform results, indicating thatthe chemical properties of the depressant rather than surfacetension were probably the significant factors in earlier work.

In view of the conflicting nature of the evidence upon thissubject it seemed desirable to study the growth of our culturesin the presence of commonly used surface tension depressants.The yeast lactose brom-cresol-purple medium employed by Albus


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


and Holm (1926, 1928) was chosen for this work. Neutral sodiumoleate (Merck) and sodium ricinoleate (Merck) were used as sur-face tension depressants. Surface tension was measured by theduNouly tensiometer at 25°C. before and after sterilization of themedia; 0.2 cc. of a forty-eight-hour milk culture was used toinoculate each 8 cc. of medium. Positive growth was recordedwith any change in the color of the indicator after incubation at37°C. for forty-eight hours.When sodium oleate was used to depress the surface tension of

the media, all of the cultures grew at 34.5 dynes per centimeter.With sodium ricinoleate as the depressant, practically all thecultures grew at 38 dynes. Occasionally a culture would fail togrow initially at this concentration but subsequent cultivationswere always positive. Unless we are to make the highly improb-able assumption that our collection contained no organisms of thebulgaricus type, it would appear that the action of these surfacetension depressants can not be used as a means of differentiatingL. acidophilus and L. bulgaricus.

CORRELATION OF CULTURE CHARACTERS

The hopelessness of classifying these cultures in any arrange-ment approaching the usual dichotomous scheme will be realizedon examination of table 2. In this table there are, if we includethe combinations made possible by incomplete determinations,47 groups. This number could be increased indefinitely by add-ing more characters. It is only by gathering the cultures intogroups having as many characteristics in common as possible thatwe can hope to arrive at any order.The most clear-cut division which we have is on the basis of the

nature of the lactic acid produced. The distinction between thecultures forming inactive acid and those forming both inactiveand dextro acid is sharp and is based on a fundamental physio-logical reaction. The test of the soundness of this division willbe found in the number of other characteristics which correlatepositively or negatively with the type of lactic acid.

In making up these groups we have included a few cultures onwhich the lactic acid was not determined but which we considered


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


TABLE 2

Characteristics of 103 Lactobacillus acidophilus culturesI~

r-

1234567891011

CULTURE NUMBER

1049128, 216, 12, 785, 9886062, 6523, 32, 33, 35, 10611, 15, 24, 44, 81, 82, 9376

LACTICACID

} a)

C) C)d

Cs +_.+

+

+

+

+

+

12 79 + +13 73, 77 + +14 10, 36, 42, 47, 55,85,87 + +15 27, 31, 34, 38, 39, 40, 43, + +

46, 48, 98162 + +127 16,13, 53 + +18 3,50 + + +19 101 + + +20 99 + + +21 74 + ++221 + + +23 45 + + +24 111 + ? +25 37 + + +26 7 + + +27 92 + +

18,26

29 75 + +

30 4 + + ?31 41, 25, 22 + +32 30 + ++33 19, 64 + + +34 20, 59, 66, 67, 68, 71, 89, + + +

103

35 58, 72, 95,107+

36 96, 97 +

LIMITINGUNDISSOCIATEDLACTIC ACID

c~2co o_C0Po Qo CD

+

+

+

+

C(

p: C

+

+

+C

+

+

202

0

.0'O C

It

+

+

+

0

F.

1-

+~~~~

+

+

FERMENTATION

Il+'- +I++j- ±+I

0

i

+

+

+

+

+

+

+

+

+

+

++

+

C)U)

_o+

+++

+

+

+

+

6L)14If


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


TABLE 2-Concluded

LACTIC LMTNUNDISSOCIATED CO2 | FERMENTATIONACD LACT'IC ACID C

CULTURE NUMBER I -

coul be sael plce on the basi of ote chrces nti

t a 5 5o wi l c

37 61,105 ++i? ? +-o+ +38 14,90,94, 100, 102 + + + + -+ +39 8,e70 + +d n g a40 108 + + ++41 29 ? ? + + + +42 56,57 ? ? + + - +--43 52 ? ? + ? ? + + +-44 54 ? ? + ? ? -+ -+45 17 ? ? + + -+--46 51 ? ? + ? ? -+--47 110 ? ? + ? ? +---

could be safely placed on the basis of other characters. On thisbasis we have group A forming inactive acid only, and group Bforming inactive plus dextro acid.

It was observed that there was a very obvious correlationbetween the type of lactic acid formed and the temperature limitswithin which the cultures grew. Of the 73 cultures forming onlyinactive acid, 58 (including 5 on which lactic acid was not deter-mined but which were considered to belong with this group) or79 per cent, did not grow at 20'C. On the other hand, of the 30cultures which formed a mixture of inactive and dextro lacticall grew at 200, and nearly all grew at 10'C. In other words, thecultures formuing only inactive lactic acid had, as a rule, a narrowtemperature range with few growing at 20' or lower while therewas a 100 per cent positive correlation between the production ofinactive plus dextro lactic acid and a wide temperature rangeextending always as low as 200 and frequently to 10'C. For thepresent we will include in the former group (A) only the 58 cul-tures forming inactive lactic and failing to grow at 20'C. The 16cultures producing only inactive acid and growing at 200 aretentatively placed in a group designated C.

While it is generally admitted that the type of colony formed


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


on suitable agar is subject to some variation, the acidophiluscolony is so characteristic that it is commonly used for identifica-tion. Our results as shown in table 3 and figure 6 indicate thatthis use is justified.Alnost without exception the cultures of the A group form X

type colonies or a mixture of X and Y types. Fully 60 per cent

TABLE 3

Distribution of colony types

TYPE OF COLONY GROUP A GROUP B GROUP C

per cent per cent per cent

X................................ 60.0 3.5 20.0Mlixture of X and Y..................... 33.3 34.5 26.6Y................................ 6.6 62.0 53.3

6RQOZ/P .4 6;eO66' 15

FIG. 6. CORRELATIONS WITH FERMENTATION REACTIONS AND COLONY TYPES

form X colonies. On the other hand, the colonies of the group Bcultures are of the Y type or a mixture of Y and X with only 3.5per cent forming X type alone.The use of fermentation reactions without regard to proper

correlation has.tended to produce confusion rather than clarifica-tion in the classification of zymogenic bacteria but when thesereactions are considered in their relation to other characteristicsthey become of real taxonomic significance. Tlle fermentation


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


of raffinose, mannitol, and salicin is shown in table 4. In thefermentation of these substances the cultures of groups A and Bexhibit distinct differences which are correlated with other charac-ters. The A cultures, with very few exceptions, ferment raffinoseand fail to ferment mannitol. About one-half ferment salicin.The B cultures quite uniformly fail to ferment raffinose but fer-ment mannitol and salicin. The correlation between type oflactic acid, temperature relations, colony type and these fermen-tations is too marked to be a coincidence.

It has been pointed out that the kinds of volatile acid areapparently identical for all of these cultures. The relation of the

TABLE 4Fermentation of rajJinose, mannitol, and salicin

GROUP A-INACTIVE GROUP B-INACTIVE GROUP C-INACTIVEACID; NO GROWTH AT AND DEXTRO ACID; ACID; GROWTH AT

200C. GROWTH AT 20C. 200C.

SUGAR Number Per cent of Number Per cent of Number Per cent ofof cultures of cultures of cultures

cultures cultures cultures

Raffinose ......... 53 4 92.99 7.01 5 25 16.67 83.33 5 10 33.33 66.67Mannitol ......... 4 54 6.89 93.11 23 7 76.67 23.33 4 11 26.67 73.33Salicin ........... 32 26 55.17 44.83 28 2 93.33 6.67 8 7 53.33 46.67

volatile to total acid is also fairly uniform for the entire collec-tion but figure 2 indicates that in general the cultures of group Aproduce a higher percentage of volatile acid than those of groupB. While the cultures of both groups are distributed over theentire range, 83 per cent of the group B cultures show below 12per cent of total acid volatile, while 80 per cent of the A culturesform more than 12 per cent volatile acids. It is not certain thatthe ratio of volatile to total acid is constant for any one culturebut the distinct tendency of the A cultures to produce a higherproportion of volatile acid is of some significance.The production of C02, shown in figure 2, differs from that of

volatile acid in that the distribution of cultures shows two dis-tinct modes. In the cultures forming less than 4 cc. of gas areincluded 80 per cent of the B cultures. The A cultures are more

JOURNAL OF BACTERIOLOGY, VOL. XXV, NO. 6


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

616 HAROLD R. CURRAN, L. A. ROGERS AND E. 0. W1rHITTIER

vrariable but show a tendency to produce a distinctly greaterquantity of CO2. This agrees with Sherman's (1921) observa-tioIn previously quoted. This property is very probably corre-lated with the formation of volatile acids and subject to the samevariations.The concentration of undissociated lactic acid at which growth

ceases has never been used as a differentiating character, but itis well known that various species of streptococci stop growtlh ata definite hydrogen-ion concentration and that under uniformconditionis this point is quite constant for each species. For eaclhof our cultures there is a definite concentration of undissociatedlactic acid at which groN-th stops and the cultures fall on thisbasis into three quite distinct groups.Except for a few of the 13 cultures which will be discussed later,

there is little evidence of a correlation of limiting undissociatedlactic acid with other distinguishing characters. In figure 3 theA cultures are shown distributed in 3 modes made up of culturesof which one group was inhibited by a very low concentration, oneby concentrations ranging from 0.060 to 0.120, and one by con-centrations above 0.120. The B cultures are distributed throughthese 3 modes, but with 62 per cent in the group varying from0.060 to 0.120. There is no apparent relationship between theability to grow after a high concentration of undissociated lacticacid is reached and the activity of the culture in souring milk.In other words, a culture which curdles milk in a short time doesnot necessarily belong among the cultures continuing growth untilthe undissociated lactic acid reaches 0.120 or higher.The concentration of phenol and indol in which these cultures

will grow, although possibly not a determining factor, may beconsidered as an indication of their ability to maintain themselvesin the intestinal tract. Figure 4 shows the distribution of thecultures in relation to their ability to grow in varying concentra-tioIls of phenol. Eighty-four per cent of the A cultures grew inconcentrations as high as 1:250 to 1:400. On the other hiand,none of the B cultures grew in conicentrations as high as 1: 350 andless than half were able to groxv at 1:350 to 1:400. The difference


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


between the two groups is even more striking in their reaction toindol as shown in figure 5. Of the 41 A cultures on which thistest was completed, 34 or 83 per cent grew in concentrations offrom 1: 1100 to 1:1900. Only one of the B cultures was able togrow at concentrations as high as 1:1900.A group (C) of 15 cultures producing only inactive lactic acid

has been set apart from the A group on the basis of their abilityto grow at 20'C. or lower. The validity of a separation on thisbasis is doubtful. Some cultures grew at 20°, but not at 15°C.In others growth occurred at 20°C. in one test and failed on thesecond test. Some grew at temperatures as low as 10°C. Thereare too few cultures in this group to warrant a statistical treat-ment but some indication of their relationship to the cultures ofgroup A may be found in the five charts. These data offer nosuggestion of a basis for separationin to homogeneous groups. Inthe sugar fermentations the proportion of raffinose fermenters issmaller than would be expected. The colonies are largely of thesmooth type. It is very probable that some at least of the C cul-tures should be included in the A group. If all were included thedescription of the A group would become somewhat less exact.It would then be necessary to state that while the typical culturedoes not grow at 200 a culture could not be excluded if it grew at20°C. or lower. The percentage of raffinose fermenters wouldalso be reduced and the percentage of mannitol fermenters in-creased. Finally, the occurrence of the Y type of colony in groupA would be increased. It is probable that some of these culturescould be looked upon as atypical members of the A group whileothers could be completely separated by a more detailed study ofa larger collection.

In the B group are 8 cultures which may possibly, be separatedfrom the group on the basis of their ability to carry the lactic acidconcentration to a high point. While all other cultures arechecked by a concentration of 0.120 moles of undissociated lacticacid per liter or lower, these continue growth until the concentra-tion reaches 0.140 to 0.180. These cultures are almost identicalin every respect and it is quite probable that in a larger collection


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

618 HAROLD R. CURRAN, L. A. ROGERS ANID E. 0. WHITTIER

they could be differentiated as a separate species. We hesitateto do this on the basis of only 8 cultures.The arrangement of these cultures according to source as

shown in table 5 is enlightening. It will be noted that 73 percent of the cultures known to be of intestinal origin fall in groupA. However, a few cultures which were evidently considered bythe laboratory in which they were isolated as typical L. acidophilusclearly belong in group B. It should be stated that of the 5intestinal cultures in this group, 3 were not obtained from thelaboratory in which they were originally isolated. Two wereisolated from feces of rats in our own colony which had been fedfor an extended period on glucose and lactose. The 9 D6derlein

TABLE 5

Distribution of cultures by sources

VAGINA INTESTINE DENTAL ACIDOPHILUS UNKNOWNPREPARATIONS

GROUP _

Num- Per Num- Per Num- Per Num- Per Num- Perber cent ber cent ber cent ber cent ber cent

A 6 66.7 17 73.9 8 33.3 5 38.5 22 64.7B 5 21.7 14 58.3 2 15.4 9 26.5C 3 33.3 1 4.4 2 8.3 6 46.1 3 8.8

cultures were all placed in groups A or C and it may be assumedthat they were originally of intestinal origin.Over half of the dental cultures belong in the B group. The 8 A

group dental cultures which came from four laboratories con-formed in all respects with the characteristics of the group. Thisis also true of the 14 dental cultures in the B group. In so far asthe evidence from this collection indicates, the lactobacilli occur-ring in carious teeth are not of one species and are not usually ofthe acidophilus type.On the assumption that group A is identical with Lactobacillus

acidophilus, a description of this species may be formulated.For convenient comparison this is arranged in parallel columnswith that of group B.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

CHARACTERISTICS OF LACTOBACILLUS ACIDOPHILUS

Group A(Lactobacillus acidophilus.)

Gram-positive rods; dimensions vari-able; frequently in chains; non-sporeforming; non-motile.

Facultative anaerobe.Colonies on tomato-juice peptone agarhave delicate filamentous outgrowthsgiving the colony a rough or woollyappearance. These may be mixedwith smooth round colonies. Inexceptional cases the colonies may allbe smooth.

Optimum temperature is 370 to 400C.The maximum growth temperatureis usually 430, but a few cultures growat 460 and sometimes at 480C. Withpossibly a few exceptions growth doesnot take place at 200C.

Milk is curdled, in some cases veryslowly, with a firm curd. Litmus, ifpresent, is reduced.

In the fermentation of lactose a mixtureof acids is formed which includesvolatile acids usually in the propor-tion of between 12 and 20 per cent ofthe total acid. The volatile acids areformic, acetic, and butyric.

The lactic acid produced is entirely ofthe inactive type.

Small amounts of C02 are formed in thefermentation of sugars. The amountis usually greater than that evolvedby the cultures of the B group.

Maltose, sucrose, and raffinose are al-most always fermented. Mannitol israrely fermented, and about one-halfthe cultures ferment salicin.

Most of the cultures are resistant tophenol and in a suitable medium willgrow in concentrations of 1:250 to1:400.

Cultures are usually tolerant of indoland will grow in concentrations of1:1100 to 1:1900.

Group B(Identity uncertain. Probably in-cludes L. bulgaricus and possiblyL. casei.)

Gram-positive rods; dimensions vari-able; frequently in chains; non-sporeforming; non-motile.

Facultative anaerobe.Colonies on tomato-juice peptone agar

are round and smooth, but in somecultures these are mixed with colo-nies of the rough or woolly appear-ance. Only exceptional culturesgive all rough colonies.

Optimum temperature is 370 to 400C.Upper limits of growth, 430 to 480with occasional cultures growing at500C. All cultures grow at 15° andusually at 100C.

Milk is curdled, in some cases veryslowly, with a firm curd. Litmus, ifpresent, is reduced.

In the fermentation of lactose a mix-ture of acids is formed which in-cludes volatile acids usually in theproportion of from 4 to 12 per centof the total. The volatile acids areformic, acetic, and butyric.

The lactic acid produced is a mixtureof the inactive and active. Theactive acid is dextro-rotatory.

Small amounts of C02 are formed inthe fermentation of sugars. This isusually less than is produced by theacidophilus cultures under the sameconditions.

Maltose and sucrose are almost alwaysfermented. Only occasionally docultures ferment raffinose. Manni-tol and salicin are usually fermented.

Cultures are inhibited by phenol inconcentrations of 1:400.

Cultures are sensitive to indol andrarely grow in concentrationsgreater than 1:1900.

619


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


SUMMARY

It was found possible to divide a collection of 103 lactobacilliinto two distinct groups on the basis of kind of acid fornmed, tein-perature reactions, type of colony, fermentation of carbohydratesand inhibition by phenol and indol. Oine of these groups, in-cluding 58 cultures, agreed Nith the generally accepted conceptof Lactobacillu.s acidophilus and showed a high positive correla-tion in the production of inactive lactic acid, the formation ofrough or fuzzy colonies, failure to grow above 46° or as low as200C., fermentation of raffinose and failure to ferment mannitol.These cultures formed, as a rule, more CO2, and a larger propor-tion of volatile acids, and grew in higher concentrations of phenoland indol than the members of the second group.The cultures of the second group produced inactive and dextro

lactic acid and grew- at 20°C. or lower. The colony was usuallyof the smooth or Y type or a mixture of Y and X, rarely of thepure X type. Only occasionally did cultures ferment raffinosebut mannitol and salicin were nearly always fermented. Thesestrains were less active than the cultures of the first group in theformation of volatile acids and CO2 and were inhibited by rela-tively dilute solutions of phenol and indol.A third group, containing 15 cultures, differed from the first

group in growing at 200C. There was also sufficient variation inother characters to warrant setting them aside at least as atypicalcultures.

All of the cultures produced some volatile acid consisting of amixture of formic, acetic, and butyric.A large proportion of the cultures of known intestinal origin

were in the acidophilus group. A few of the dental cultures wereof the acidophilus type, but the greater part belonged in the groupproducing inactive plus dextro lactic acid.

REFERENCESALBUS, W. R., AND HOLM, G. E. 1926. Jour. Bacteriol., 12, 13.ALBUS, W. R., AND HOLM, G. E. 1928. Jour. Bacteriol., 16, 197.BROWN, J. HOWARD 1924. Jour. Lab. and Clin. Med., 9, 239.DAY, A. A., AND GIBBS, WV. M. 1928. Jour. Infec. Dis., 43, 97.DUCLAUX, P. E. 1900. Traite de Microbiologie, 3, 385.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


ELDREDGE, E. F., AND ROGERS, L. A. 1914. Centralbl. f. Bakt., II, 40, 5.

FRED, E. B., PETERSON, W. H., AND STILES, H. R. 1925. Jour. Bacteriol., 10, 63.

FRIEDEMANN, T. E., COTONIO, M., AND SHAFFER, P. A. 1927. Jour. Biol. Chem.,73, 335.

HADLEY, F. P., BUNTING, R. W., AND DELVES, M. S. 1930. Jour. Amer. Dent.Assoc., 17, 2041.

HART, E. B., HASTINGS, E. G., FLINT, E. M., AND EVANS, A. C. 1914. Jour.Agric. Res., 2, 193.

HEINEMAN, P. G., AND HEFFERAN, M. 1909. Jour. Infec. Dis., 6, 304.HYDE, L. S., AND HAMMER, B. W. 1927. Ia. Sta. Col. Jour. Sci., 1, 419.HOWITT, B. 1930. Jour. Infec. Dis., 46, 351.JOETTEN, K. W. 1922. Archiv. f. Hyg., 91, 143.KOLTHOFF, I. M. 1925. Tijdschr. Vergelijk Geneeskunde, 11, 268.KOPELOFF, N. 1926. Lactobacillus acidophilus. Williams and Wilkins, p. 23.

KOPELOFF, N., AND BASS, L. W. 1927. Proc. Soc. Exp. Biol. and Med., 24, 773.KOPELOFF, N., AND BEERMAN, P. 1927. Jour. Bacteriol., 13,7.KULP, W. L., AND RETTGER, L. F. 1924. Jour. Bacteriol., 9, 357.KULP, W. L. 1927. Science, 66, 512.KULP, W. L. 1929. Jour. Bacteriol., 17, 355.LASH, A. F., AND KAPLAN, B. 1926. Jour. Infec. Dis., 38, 333.MORISHITA, T. 1929. Jour. Bacteriol., 18, 181.PEDERSON, C. S., PETERSON, W. H., AND FRED, E. B. 1926. Jour. Biol. Chem.,

68,151.RAHE, A. H. 1914. Jour. Infee. Dis., 15, 141.ROGERS, L. A., AND WHITTIER, E. 0. 1928. Jour. Bacteriol., 16,211.ROSEBIURY, T., LINTON, R. W., AND BUCHBINDER, L. 1929. Jour. Bacteriol.,

18,395.SHERMAN, J. M. 1921. Abst. Bacteriol., 5,6.TENBROEK, C. 1920. Jour. Exper. Med., 32, 245.THOMAS, S. 1928. Jour. Infee. Dis., 43, 218.VAN DAM, W. 1918. Biochem. Zeits., 87, 107.WERKMAN, C. H. 1930. Indus. and Eng. Chem., Anal. Ed., 2,302.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

Documents

DepartmentTHE DISTINGUISHING CHARACTERISTICS OF LACTOBACILLUS ACIDOPHILUS HAROLDR. CURRAN, L. A. ROGERS AND E. 0. WHITTIER Research Laboratories, Bureau of Dairy Industry, United States