THE SPECIFIC CAUSE OF YELLOW FEVER. A REPLY

TO DR. G. SANARELLI.

BY

WALTER REED, M.D,SURGEON, UNITED STATES ARMY

AND

JAMES CARROLL, M.D.,ACTING ASSISTANT-SURGEON, UNITED STATES ARMY.

FROM

THE MEDICAL NEWS,

September 9, 1899.

[Reprinted from The Medical News, September 9, 1899.]

THE SPECIFIC CAUSE OF YELLOW FEVER. A REPLY TO DR. G. SAHARELLI.WALTER REED, M.D.

SURGEON, UNITED STATES ARMY,AND

JAMES CARROLL, M.D.,ACTING ASSISTANT-SURGEON, UNITED STATES ARMY.

Were it not for the fact that the general practi-tioner of medicine in this country takes such a livelyinterest in all that pertains to the causation of yellowfever, and might construe our silence as an admis-sion that we had followed faulty methods in ourwork, we would not at this time reply to the paperof G. Sanarelli, entitled “Some Observations andControversial Remarks on the Specific Cause of Yel-low Fever,” which appeared in the Medical Newsunder date of August 12, 1899. To those who areacquainted with the methods followed in the labora-tories of the Army Medical Museum, we believe thatwe could well afford to leave this matter until suchtime as our completed observations might be givenpublication.

We pass by, therefore, as unworthy of comment,Sanarelli’s insinuation that we could lend ourselvesto the support of any personal controversy betweenhimself and Surgeon-General Sternberg; and further,his advice that we should have “looked up for amoment some good treatise on bacteriology beforeattempting to launch such a paradox” as that ba-cillus icteroides should be considered to be a varietyof the hog-cholera bacillus.

That our papershould partake somewhat of a con-troversial character, we very much regret, since wewould have much preferred to publish our observa-tions in full, so that our fellow-workers might beplaced in possession of the complete details of ourexperimental work and thus be ina position to judgewhether our methods had been good or faulty. Butfor the reason given at the commencement of thispaper and because of the remarkable tone of Sana-relli’s article, we must take up what we consider tobe a somewhat unpleasant task.

Because we have recorded focal necroses in theliver of guinea-pigs and rabbits inoculated with ba-cillus icteroides (we might have mentionednecrosesin the liver of dogs, which Sanarelli has quite over-looked); and since, in Sanarelli’s opinion, we havefailed to reproduce that degree of “acute steatosis”which he and others have recorded, we are desig-nated as “victims to some deplorable neglect oftechnical precautions in the laboratory,” and the

conclusion is drawn “that a gross and deplorableerror was made in their laboratory and that somehowthe cultures of the two micro-organisms being exper-imented with got mixed for want of sufficient atten-tion.”

We are willing to admit that such a condition ofaffairs, as is implied in the latter part of this quota-tion, is possible in any laboratory; but in this con-nection we cannot refrain from paying our admira-tion to Sanarelli’s mental acumen in arriving at thisconclusion, since he must have realized that, if wehad not “mixed the two micro-organisms, ” his claimfor the discovery of the specific causeof yellow fevercan no longer be sustained.

And here let us avow that we hailed with delightSanarelli’s announcement of the discovery of his ba-cillus icteroides as the specific agent of yellow feverand earnestly hoped that his work would be con-firmed by other observers. When later directed bySurgeon-General Sternberg to carefully comparebacillus x with bacillus icteroides and report to himthe results of our investigations we were only influ-enced by the desire to ascertain whether these weredistinct bacilli, and if so, to see whether we couldconfirm Sanarelli’s observations in the several ani-mals. That we could be influenced in the slightestdegree by Dr. Sternberg in arriving at our conclu-sions is a grave aspersion upon our reputation asscientific workers.

If it can be proved that we have “mixed” ourcultures and hence arrived at incorrect conclusions,we will hasten to admit our mistake. Let us seewhether we have been careless in our technical pre-cautions, and hence have, by mistake, inoculatedour animals with the bacillus of hog-cholera whenwe had intended to use bacillus icteroides.

In the first place, before proceeding to considerSanarelli’s criticisms of our work, we desire to statethat the culture of bacillus icteroides with which wehave made the majority of our observations, was ob-tained by Dr. Sternberg from Pasteur’s laboratory.We are informed by Dr. Sternberg that the packagecontaining this culture was opened by ProfessorRoux in his (Sternberg’s) presence and the culture

2

handed to the latter. It was delivered to one ofus by Dr. Sternberg on the 24th day of September,1897, and bore the label of the Laboratory of Hy-giene, University of Montevideo. We may remarkthat this culture was not in the shape of the ordinarystroke culture, but contained a number of well iso-lated colonies, such as would be obtained by inocu-lating an agar slant with a small drop of the bloodof an infected animal. Several of the colonies onthis agar slant showed the peculiar appearance ofgrowth which Sanarelli considers as characteristicof bacillus icteroides. It was from one of these so-called characteristic colonies that we made our firsttransfer. Subsequent transplantations have beenmade by us at intervals of about four weeks, and havealways been labeled “Bacillus Icteroides,Original.”We have thus transplanted this original culture onagar, without transmission through animals, for aperiod of nearly two years. Like our other stockcultures, it has been kept in agar tubes providedwith rubber caps and always in a dark closet.

In the second place, at the time that we beganour investigations (September, 1897), no culture ofthe hog-cholera bacillus was in our laboratory, nordid we procure a culture of this bacillus from theBureau of Animal Industry until the following spring(May, 1898) about seven months after we had begunour work.

We may be excused for particularizing somewhatin this connection in order to set ourselves right be-fore those who are interested in this subject.

Let us then state that the hog-cholera bacillus wasone withwhich neitherof us was acquainted, except asto some ofits mostprominent morphological and bio-logical features, such as size, motility, growth inmilk, and in the three sugars; but we had not wit-nessed its pathological effects in the smaller ani-mals. We had more than once commented on thefact that its action on litmus milk, and in the threesugars was similar to that of bacillus icteroides; butas we were not prepared at that time to look uponbacillus cholerse suis as a secondary invader in hu-man beings, we did not follow up the subject further.

During the month of February, 1898, we had as-certained that the blood-serum of a dog partially im-munized against bacillus icteroides, although exer-cising decided agglutinative action upon the latter,was negative as regards the typhoid bacillus and ba-cillus coli communis, except in low dilutions. Atthis time our observations were almost completelyinterrupted by our duties with the class of the ArmyMedical School. We had scarcely resumed ourwork when we were honored with a visit from Dr.Theobald Smith, formerly of the Bureau of AnimalIndustry, and who didall of the bacteriological work

recorded in the “Report on Hog Cholera; ItsHistory, Nature, and Treatment,” to whose instruct-ive pages Dr. Sanarelli has so courteously invitedour attention. We communicated to Dr. Smith thefact that bacillus x (Sternberg) when injected intothe circulation of dogs would cause the same gen-eral symptoms, and the same extensive hemorrhagiclesions in the stomach and intestine as was producedby bacillus icteroides, and hence that these lesionswere not confined to the action of the latterbacillus.Dr. Smith then remarked that in reading Sanarelli’svarious articles, he was continually reminded of thehog-cholera bacillus.

When further told by us thatbacillus icteroides pro-ducedmultiple necroses in the liver of guinea-pigs,heremarked that he was of the opinion that bacillus ic-teroides would yet be placed in the hog-choleragroup. Fortified by this expression of opinion byone, than whom no one is more conversant with thehog-cholera bacillus or its pathological action, wefor the first time proceeded to procure a culture ofbacillus cholerse suis, and to test the action of ourdog’s serum, with the result already given in ourpreliminary note. As we had already, many times,obtained these focal necroses in guinea-pigs withlus icteroides, how could we have “mixed the twomicro-organisms?” We could scarcely have mixedwhat we did not have in our possession. If, asSanarelli has stated, these “focal necroses are en-tirely specific for hog-cholera,” we had alreadyproven by the production of these that bacillus ic-teroides was a variety of the hog-cholera bacillus.It is very important to note that cultures taken fromthese animals gave, in about thirty per cent., colo-nies similar in all respects to those described by San-arelli as characteristic of bacillus icteroides.

Turning now to Sanarelli’s criticisms ©four work,these may be stated as follows;

First. The marked difference in growth on potatoof his bacillus and bacillus choleras suis.

Second. That the toxin of his bacillus resists atemperature of ioo° C., while that ot the hog-chol-era bacillus undergoes modification as low as 6o° C.,and is entirely destroyed at ioo° C. Sanarelli herequotes from Selander’s observations.

Third. Dissimilarity of growth in gelatin plates ofthese two bacilli.

Fourth. That while the hog-cholera bacillus pro-duces a large number of foci of coagulation necro-sis (the organ is not stated), these do not occur inanimals inoculated with bacillus icteroides.

Fifth. That we have failed to produce acute steato-sis, or fatty degeneration of the liver in dogs in-jected with his bacillus.

As regards the difference in growth on potato

3

Sanarelli says, ‘ ‘it is sufficient to recall that the ba-cillus of Salmon forms on potato a luxurious brown-ish-colored culture somewhat like that of Mauve,while my bacillus icteroides develops on potato com-pletely without color and is scarcely visible. ’ ’

Salmon and Smith, 1 in their report, impliedneglectof which on our part to even casually glance at, hasoccasioned Sanarelli so much apparent regret, stateas follows: “growth on boiled potato when at950 F., appears as a faint straw-colored depositwithin twenty-four hours after inoculation. At20

0 C. or 25 0 C., it appears one or two days later.It slowly spreads in all directions as a layer of per-ceptible thickness. The color changes to a darkbrick-red, or may remain whitish (italics our own).In general the growth is darker the more rapidly thepotato dries up. ’ ’

Smith and Moore, 2 in describing the growth ofseveral varieties of the hog-cholera bacillus, state“that there may be slight variations in the depth ofcolor of the growth on potato due to variations inthe potatoes used. Now and then potatoes are en-countered on which no growth appears. The sur-face assumes a glistening appearance, but multipli-cation perceptible to the eye does not take place. ’ ’

Welch, 3 describing the growth of the hog-cholerabacillus, says; “the growth on potato assumes gen-erally a brownish or yellowish tint, but it may bewhite and sometimes is itidistinct (italics our own),although microscopically the growth is abundant. ’ ’

Our own observations show that the hog-cholerabacillus, when transplanted on potato grows, as arule, as a moist straw-colored or brownish layerspreading over the greater part of the surface of themedium; but that exceptionally it presents a moist,almost colorless layer. Cover-slips from this moist,almost colorless layer show that a marked multipli-cation of the bacilli has taken place.

It will thus be seen that the growth of the hog-cholera bacillus on potato is variable, and that itdoes not always produce “a luxurious brownish col-ored growth, ’ ’ as claimed by Sanarelli.

Our experience with the growth of bacillus icte-roides on potato is at variance with that given bySanarelli. Both of us promptly transferred to vari-ous media the original bacillus icteroides turnedover to us by Dr, Sternberg on September 24, 1897.As regards the growth on potato, we quote from ournotes made on September 29th, 1897, as follows:Dr. Carroll, “a moist brownish growth. ” Dr.Reed, “grows less rapidly than bacillus x, giving amoist, thin, brownish layer, which spreads over thesurface of the medium.” These were duplicate ob-

1 “Hog Cholera,” Bureau of AnimalIndustry, Washington, 1889.8 “Infectious Swine Diseases,” Washington, 1894.* Johns Hopkins Bulletin, vol. i, No. i, iBBg.

servations made independently of each other.Our subsequent observations have shown that ba-

cillus icteroides on potato may grow as a colorlessmoist layer, or as a faint yellowish layer, or that itmay show a decided brownish color,

A. Agraraonte, 1 who spent some months in thislaboratory comparing a bacillus which he had ob-tained from yellow-fever cadavers at Santiago, Cuba,during the summer of 1898, with bacillus icteroides,could find no difference between these. He says asregards the growth of this bacillus; “Potato culturesfor twenty-four hours at 390 C. gave an abundantpale-cream, almost colorless growth; when comparedside by side with others made of bacillus icteroidesit was impossible to differentiate them. At the endof a week the cultures remained of the same color;while some of bacillus icteroides had taken on abrownish tinge (italics our own), others remainedthe same.

All of which goes to show that bacillus icteroidesdoes not always grow on potato as a colorless,scarcely visible layer, as claimed by Sanarelli, andhence that he cannot safely base a differentiation onthis account between his bacillus and bacilluscholerse suis. It would be well if Sanarelli wouldrecall our experience with the typhoid bacillus. Aswith that bacillus, so with bacillus icteroides, it mayor may not grow on potato as a colorless layer.

With reference to the claim of Sanarelli that thetoxin of his bacillus will resist a temperature of ioo°C., while that of the bacillus of hog-cholera under-goes modification as low as 6o° C., and is entirelydestroyed at ioo° C., we have no observations ofour own to record. We propose to show, however,that his statement concerning the supposed effect ofcertain temperatures on the toxin of the bacilluscholerse suis, which he quotes from Selander, is en-titled to no credence whatever. Indeed, we con-sider Sanarelli’s reference to the work of Selanderand Metchnikoff2 as particularly unfortunate, sincesubsequent observations have conclusively demon-strated that at that time, 1890-92, neither Selandernor Metchnikoff were working with the hog-cholerabacillus!

We almost feel like suggesting to our distin-guished critic that he should follow the advicewhich he seems so willing to give to others, andspend a brief time in consulting the literature onthis subject.

Referring to Selander’s investigations, Welch andClement® say: “We wish, however, in this connec-

1 The bacillus icteroides (Sanarelli) and bacillus x (Sternberg),Centralblattfurßacteriologie, etc.. Band xxv, No. 18-ig, 1899.

8 Annals de Vlnstitut Pasteur , 1890-92.3 “Hog Cholera and Swine Plague,” Proceedings First Veteri-

nary Congress of America, Chicago, 1893.

4

tion, to express our conviction that the organismdescribed by Selander in 1890, and probably alsothat by Metchnikoff later, as the hog-cholera (orswine-pest) bacillus, is not identical with the gen-uine hog-cholera bacillus as we know it in this coun-try. ’ ’ After recording their inability to confirm theexperiments of Selander in increasing the virulenceof the supposed hog-cholera bacillus, they say:“Whatever this organism of Selander may havebeen, we feel convinced that it was not our hog-cholera bacillus, and we base this conclusion uponthe results of his experiments in Pasteur’s Institute,as published, and upon our experience with the cul-tures with which he kindly supplied us.We believe Metchnikoff’s hog cholera microbe to beone of the hemorrhagic septicemia bacilli, possiblythe swine-plague or schweine-seuche bacillus.”

This opinion expressed by Welch and Clement wassubsequently proven to be correct by Smith andMoore, 1 to whom Metchnikoff, having confirmedthe observations of Selander in increasing the viru-lence of the supposed hog-cholera bacillus sent theblood of a guinea-pig inoculated with the bacteria in asealed glass tube. These writers state that “cultivatedon various substratathey presented all the charactersofswine-plague bacteria as found in this country.”Certainly we have not claimed any affinity betweenthe swine-plague bacillus and bacillus icteroides.

Not only does Sanarelli, in quoting from the workof Selander concerning the effect of varying tem-peratures upon the toxin of Selander’s bacillus, con-found the bacillus of swine-plague with the bacillusof hog-cholera, but he also displays the most as-tonishing want of acquaintance with the biologicalcharacters of the last-named bacillus. He says:“We may recall, besides, that colonies of the twoforms of bacilli (bacillus icteroides and bacilluscholerse suis) on gelatin plates are as readily dif-ferentiable, even at first sight, as it is possible forbacilli to be.”

On the contrary, we wish to state that plated,side by side, in gelatin, we have been unable to de-tect any differences in the colonies of these two ba-cilli with the naked eye, and that it is only bymeans of the microscope that faint differences can,as a rule, be made out, such as a slightly increasedradiation in the colonies of bacilli icteroides, ascompared with those of bacillus cholerse suis. Weare, therefore, constrained to the opinion that Dr.Sanarelli has never plated on gelatin the genuinehog-cholera bacillus, such as we know it in thiscountry.

Referring to the focal necroses found by ourselvesin the liver of guinea-pigs and rabbits inoculated

» “Infectious Swine Diseases,” Washington, 1894.

with bacillus icteroides, Sanarelli says: “I am verymuch surprised that Reed and Carroll should havefound this lesion, and should have betrayed seem-ingly good faith in describing it so carefully, with-out realizing that the lesion is entirely specific tohog-cholera and limited to that affection, and iscompletely unknown in infections with bacillus ic-teroides.”

We can pass by the discourteous part of the fore-going sentence; but we must confess that even hadwe not succeeded many times in producing thesefocal necroses in the liver of pigs and rabbits withbacillus icteroides, we were not prepared for thestatement that this lesion was ‘‘entirely specific tohog-cholera and limited to that affection;” espe-cially since Blachstein 1 had succeeded in producingthis lesion in the liver of rabbits by the in-travenous injection of bacillus coli communis andbacillus typhi abdominalis; and since, moreover,one of us (Reed 8 ) working in Welch’s laboratory,had caused the same lesion by injecting cultures ofthe typhoid bacillus into the mesenteric vein of rab-bits.

Blachstein, speaking of the lesions due to bacilluscoli communis, says: “The more striking and con-stant lesions, those most characteristic of the affectionare in the bile and liver.” Further he says: “onlyone or two such areas may be noticed, or they mayoccur in large numbers. . . , These areas occurboth on the surface and in the depth of the liver. . . .On microscopical section these areas are found to beplaces where the liver cells have undergone necrosis. ’ ’He adds: “These necrotic and inflammatory changesappear to be of essentially the same nature as havebeen observed aft;r ligation of the common bile duct(Foa, Salvioli, Pick, and others), and occur in otherinfectious processes, such as after inoculation withtyphoid bacilli, and with hog-cholera and schweine-pest bacilli.”

Indeed, scattered areas of coagulation necrosis inthe livers of theseanimals have been produced by otherbacteria, and notably by such toxalbumens as abrinand ricin, in the hands of Flexner, 3 and we know,as stated by the latter, the close relationship exist-ing between these bodies and certain bacterial toxins.

Again we ask, can Dr. Sanarelli be ignorant ofthe rich literature relating to the production of mul-tiple focal necroses in the liver of guinea-pigs, rab-bits, kittens, etc., by the bacillus diphtherise and itstoxins? Indeed, in limiting focal necroses in the'livers of guinea-pigs and rabbits to the agency of

1 “Intravenous Inoculation of Rabbits with the Bacillus ColiCommunis and the Bacillus Typhi Abdominalis,” Johns HopkinsBulletin, vol. ii, No. 14, 1891.

2 Johns Hopkins HospitalReports , vol. v, 1895.3 “The Pathology of Tox-albumen Hopkins

HospitalReports, vol. vi, 1897.

one particular bacterium, Sanarelli displays profoundignorance of the pathology of coagulation necrosis.

Turning to our statement that these lesions arecaused by inoculation with bacillus icteroides, weinvite attention to the fact that Agramonte, 1 to whomwe communicated our observations, obtained thesame necroses in pigs injected with a bacillus ob-tained from yellow-fever cadavers (33 percent.) atSantiago, and which he identified as bacillus icte-roides (Sanarelli). We note also that De Lacerdaand Ramos 2 record the finding of “scattered yellowpoints’ ’ [avec des points jannes dissemines) in theliver of one rabbit out of four inoculated with ba-cillus icteroides and which died on the second day.They say nothing about the microscopic examina-tion of these scattered yellow points.

We have failed to find, as a rule, necroses in thelivers of animals dying earlier than the fourth day,although we have generally found them between thefourth and ninth day after inoculation, especiallyprominent in the livers of guinea-pigs. Our earliestrecorded observation was made on November 13,1897, upon autopsying guinea-pig No. . 456, deadon the fourth day after receiving 1 c.c. of a 20-hourbouillon-culture of bacillus icteroides. Concerningthe liver, we note: “Organ pale; some recent ad-hesions, numerous small punctiform areas, yellow-ish white in color, to be seen under capsule. ’ ’

With regard to the several observers cited by San-arelli as in opposition to our findings, we have beenunable to obtain the articles written by some ofthese and hence cannot state just what lesions theydid find in the livers of the inoculated animals. Ofthose accessible to us, we find that Domennica dellaRovera 3 states the following as the result of his in-fection of guinea-pigs with bacillus icteroides: “Myexperiments have demonstrated as a new finding inthe liver, foci of small cells (focolai parvi cellulari)situated in the midst of the lobules; the protoplasmof the liver-cells appear very granular, the nucleidiscolored and feebly stained, as well as the nucleiof the cells lining the biliary canals and capillaries.’’This description of Rovera agrees very well withwhat we have found microscopically, except that wealso find larger foci, whose capillaries are crowdedwith polymorphonuclear leucocytes, and whose cellsstain brightly with eosin, especially toward the morecentral parts of the area. Sometimes we find anentire lobule the seat of coagulation necrosis. Thisquotation further shows the accuracy with whichDella Rovera has made his observations, and com-

1 CentralblattfilrBacteriologie, etc., Band xxv, No. 18-19,1899.2 “Le bacille Icteroide et Sa Toxine,” Archives de Medicine Ex-

perimentale, No. 3, 1899.3 “Sul bacillo Icteroide” (Sanarelli), La Reforma Medica, vol.

iii, No. 9, July, 1898.

pletely refutes the position taken by Sanarelli, viz.:that focal necroses are never found in the livers ofanimals inoculated with bacillus icteroides. Rovera1also says concerning lesions found in the livers ofrabbits; “I also saw foci of small cells, situatedwithin the lobules, such as I have already recordedfor guinea pigs ’ ’

In view of the foregoing observations comfirmingour findings as regards the agency of bacillus icte-roides in producing focal necroses in the liver ofof guinea-pigs and rabbits, we are constained to saythat Dr. Sanarelli in emphasizing the specific char-acter of these necroses for the hog-cholera bacillus,again shows a surprisingly limited knowledge of thework which has been recorded in this line. Weare quite willing to admit, however, that these ne-croses are probably more abundantly produced bythe hog-cholera bacillus and its varieties, such asbacillus icteroides, than by other bacteria.

This brings us to Sanarelli’s fifth criticism of ourresults, viz.: that we “did not succeed in reprodu-cing the acute steatosis, fatty degeneration of theliver in dogs that was always so prominent a featurein my own observations as well as those ofFoa” andseveral other investigators whose names we hereomit.

We have endeavored to extract from the literature,as far as the Library of the Surgeon-General’s Officewould permit, the experiments made upon dogs byvarious observers, but have not been able to finddefinite data except as regards the experiments ofSanarelli, 2 De Lacerda and Ramos, 3 to which weadd our own results.

We submit these in the following table:It will be seen that there are eight (8) observa-

tions by Sanarelli; twelve (12) by De Lacerda andRamos, and seven (7) by Reed and Carroll, makinga total of twenty-seven (27). Of these, six (6) dogsrecovered, leaving twenty-one (21) that died as theresult of the intravenous inoculation of the culturesor toxin of bacillus icteroides, or both of these. Inone instance, the toxin was injected both intraven-ously and into the substance of the liver. Omittingone dog, which was not autopsied, we have twenty(20) fatal results, with fatty changes in liver re-corded in thirteen (13), or sixty-five per cent. Ofthese, one, reported by Sanarelli (No. 3), diedwithin three hours and forty minutes after receipt ofthe injection, and “showed that the hepatic cellswere already rich in fat drops. ’ ’ He does not statethat there was any fatty degeneration present, andwe can well believe that such was not the case.

1Archives de Medicine Experimental, No. 3, 1899.2 11 Policlinico, vol. iv, No. 16 and 18, 1897.3 Archives de Med. Exper ., vol. ii, No. 3, May, 1899.

6

Injection of Dogs with Bacillus Icteroides.

Observer. No. Quantity received. Result. Remarks.

Sanarelli I io c. c. of a 24-hour bouillon culture, intra-venously.

Recovered.

2 Agar culture diluted with 1 c.c. of a bouil-lon culture, intravenously.

Died during the samenight.

Liver congested, but no fattydegeneration.

3 Agar culture diluted with 1 c.c. of a bouil-lon culture, intravenously.

Died in 3 hours and 40minutes.

Liver shows numerous yellow-ish spots. Fresh sectionsin osmic show the hepaticcells already rich in fatdrops.

4 5 c.c. of a bouillon culture, intravenously. Died on 16th day. Intense fatty degeneration ofliver.

5 Agar culture diluted with bouillon—quan-tity not stated, intravenously.

Died during the samenight.

Liver pale with zones of fattydegeneration.

44 6 2 c.c. of a bouillon culture, intravenously. Died on 19th day. Marked fatty degeneration ofliver.

“ 7 2 c.c. of a bouillon culture, intravenously. Died on 8th day. Fatty degeneration of liver.8 5 c.c. of a bouillon culture, intravenously. Died on 10thday. Marked fatty degeneration of

liver.

De Lacerda and Ramos. I 1 c.c. of an agar culture diluted with 10c.c.of sterilized water, intravenously.Died during the same

day.No change in liver.

44 U “ 2 10 c.c. of an agar culture diluted with ster-ilized water, intravenously.

Recovered.

i ( U ( ( 3 20 c.c. of a bouillon culture into liver. Recovered.

1( ( I u 4 20 c.c. of a bouillon culture 2 days old, in-travenously.

Died on 19th day. No autopsy.(( (( it 5 35 c.c. of a bouillon culture 3 days old, in-

travenously.Recovered.

it it it 6 35 c.c. of a bouillon culture 1 day Old, in-travenously.

Died in 16 hours. No trace of fatty degeneration.

it 11 I ( 7 18 c.c. of a bouillon culture 2 days old, and38 days later 60 c.c. of toxin, both in-travenously.

Died in about 8 hoursafter the 2nd injec-tion.

Fatty degeneration of liver.

8 20 c.c. of a bouillon culture and. 10 c.c. ofthe toxin, intravenously.

Died in 24 hours. Liver and kidney of a blackishcolor. Nothing said offatty generation.

i i it Ii 9 60 c.c. of the toxin, intravenously. Died during the samenight.

Intense congestion of all organs.No fatty degeneration.

it it 11 io 20 c.c. of the toxin, intravenously, and 20c.c. into the liver.

Died in 24 hours. Fatty degeneration of liver,kidneys, and heart.

it 11 11 II 70 c.c. of the toxin, intravenously. Died at 1 o’clock of thesame night.

Fatty degeneration of liver.

it 11 11 12 50 c.c. of the toxin, intravenously. Died on 7th day. Disseminated points of fattydegeneration but little ad-vanced.

Reed and Carroll I 5 c.c. of a 24-hour bouillon culture, intra-venously, Sept. 24, 1897.

Died on 9th day. Necroses in liver. Moderatefatty degeneration.

2 5 c.c. of a 24-hour bouillon culture, Nov.16, 1897, and a second injection of 5c.c. 14 clays later, both intravenously.

Died in 28 minutesaftersecond injection.

Slight fatty degeneration ofliver.

it t i t i 3 c.c. of a 24-hour bouillon culture, in-travenously, Dec. 13, 1897.

Died on 9th day. Marked fatty degeneration ofliver.

it Ii t i 4 5 c.c. of a 24-hour bouillon culture, intra-venously, Feb. 9, 1898.Recovered.

44 44 44 5 5. c.c. of a 24-hour bouillon culture, intra-venously, March 17, 1898.

Died in 52 hours. Necroses in liver. No fattydegeneration.

44 ‘ 4 44 6 5 c.c. of a 24-hour bouillon culture, intra-venously, July 1, 1898.

Died in 16 hours. Necroses in liver. No fattydegeneration.

it it it 7 5 c.c. of a 24-hour bouillon culture, intra-venously, July 1, 1898.

Recovered.

7

What he found in this liver may readily be found inthe normal hepatic cells of dogs in a good conditionof health and nutrition. We would, therefore, ex-clude this case as well as the case (No. 12) recordedby De Lacerda and Ramos, in which they found“disseminated points of fatty degeneration, but lit-tle advanced,” in the liver. No results of the micro-scopic examination is given in this case to showwhat the true character ofthe “disseminatedpoints”was, whether fatty or necrotic, and since it is notcharacteristic of fatty degeneration of the liver thatit should be found in “disseminated points, ” wefeel justified in also omitting this case. We will notdo so, however, in order that we may not be ac-cused of unfairness.

We thus find twelve (12) positive results againsteight (8) negative results in dogs injected with bacillusicteroides. Of the positive results five (5) are recordedby Sanarelli in seven (7) animals that died (71.3per cent.); four (4) by De Lacerda and Ramos in eight(8) dogs autopsied (50 per cent.); and three (3) byReed and Carroll in five (5) deaths (60 per cent.).It will be seen that the percentage of cases showingfatty degeneration is highest in Sanarelli’s experi-ments, due probable to the fact that his dogs livedlonger, but that the percentage reported by De La-cerda and Ramos is less than that reported by our-selves. Since De Lacerda and Ramos are authori-ties quoted as sustaining Sanarelli in his observationsconcerning the frequent occurrence of fatty degen-eration in dogs injected withbacillus icteroides, andsince we have obtained this change in even a largerpercentage of cases, it naturally follows that Sana-relli, in criticising us for our failure to produce fattydegeneration in the liver of dogs, was speaking asusual without any knowledge of the true facts, justas we have already pointed out that he has done inthe matter of Selander’s observations, and in theabsurd statement that necroses in the liver ofguinea-pigs and rabbits “are entirely specific for hog-chol-era!” True we did state, and we wish here to re-peat, that, with one exception hereafter to be given,the fatty change found in the liver of these dogs isnot comparable to that found in the liver of humanbeings who have died of yellow fever, and we mayadd that there is an almost entire absence of thatnecroses of individual liver cells which is so promi-nent a feature in the human liver.

Sanarelli speaks as if yellow fever were the onlycondition in which fat is found in the liver cells.It is hardly necessary to remark that it may be pres-ent in considerable quantity in the livers of compar-atively healthy individuals. Steatosis may meanlittle or much; it does not necessarily imply de-generation. Pathologists well know that in yellow

fever the change that takes place in the liver is es-sentially one of degeneratioti and necrosis , in whichthe hepatic cells appear as denucleated bodies dis-tended with minute fat-droplets and containing onlya small amount of protoplasmic debris, or as some-what refractive bodies, which stain brightly witheosin. Their outlines are very much distorted andthe normal beam-work arrangement is no longerseen. Sanarelli’s illustrations in Plate 10, Fig. 5,II Policlinico, Nos. x6and 17, 1897, prove that thisdegenerative and necrotic condition was not presentin the livers of human beings inoculated with thetoxin of his bacillus. In his section of dog’s liverthe arrangement of the liver cells is perfectly pre-served, their nuclei stain well, and there is only amoderate deposition of fat in the cell protoplasm.

In connection with the production of fatty de-generation of the liver ofdogs injected with bacillusicteroides, we desire to invite attention to the com-parative degree of this change found in the case ofdogs 531 and 928 of our series of experiments.

Dog No. 531, weight 17 pounds, was originallyinoculated on February 9, 1898with 5 c. cm. of abouillon-culture of bacillus icteroides, from whichhe recovered. He was then subjected to intraven-ous injections of increasing doses until he was ableto bear injections of 150 c. cm., of a bouillon-cul-ture of this bacillus. On July 20, 1898, the animalhad received 150 c.cm. of the culture, and died afew hours after the injection of the same quantity onJuly 29th. Thus this dog had received during aperiod of five (5) months and twenty (20) days,

c. cm. of a bouillon-culture of bacillusicteroides; yet at autopsy the liver was found tocontain only a moderate amount of fat, and uponmicroscopic examination but slight fatty degenera-tion of the hepatic cells was to be seen.

On the contrary, dog 928, weight 19 pounds, wasinjected with cultures of the bacillus cholene suis forthe purpose of immunization, and received a totalof 170 c.cm. during a period of four (4) months andfive (5) days. The largest dose was 50 c.cm. in-jected on July 19, 1899, from which theanimal diedthirty-six hours later. At autopsy the liver wasseen to be quite light in color, and upon micro-scopic examination the hepatic cells were found tobe not only the seat of advanced fatty degeneration,but there were, also, to be seen many necrotic livercells. Sections of this liver, indeed, show a stageof fatty degeneration and necrosis closely approach-ing that of yellow fever in human beings, and yetthis animal was injected with bacillus cholerse suis!Or did we here again commit the deplorable errorof “mixing the two micro-organisms”? Ratherwould we say that the important determining factor

8

in producing this fatty change was the difference inresistance of these dogs, and was not dependent onthe particular variety of the hog-cholera bacilluswith which we had injected them.

We observe, with some surprise, that Sanarellidoes not comment on the remarkable agglutinativereaction which icteroides serum (supplied to us fromBrazil, and presumably from animals inoculated withpure cultures of bacillus icteroides), exerts towardthe hog-cholera bacillus. The scientific fact alone,viz. , that this serum diluted 100,000 times shouldarrest the motility and agglutinate another bacillus,should, we think, have momentarily attracted hisattention; or is it possible that we here again gotour cultures “mixed” and were really subjecting ba-cillus icteroides to the action of its own serum? Nordoes he appear to have been impressed with thefact that the hog-cholera bacillus should so com-pletely reproduce in dogs, not only the clinicalpicture of vomiting, rectal tenesmus, and profoundprostration, but also death with extensive hemor-rhagic changes in the stomach and intestines. Wetrust, however, that he will later procure a cultureof the bacillus cholerse suis and give us the result ofhis observations with it on various animals. In themeanwhile, we beg to observe that Agramonte, 1working in this laboratory for a period of fourmonths, was unable to detect any differences be-tween the “mixed” culture with which we, unwit-tingly, supplied him, and a bacillus which he hadrecovered in Cuba from yellow-fever cadavers, andwhich he identified as bacillus icteroides (Sanarelli).

Concerning the statement made by us in our pre-liminary note that bacillus icteroides and bacilluscholerse suis both resisted extremely low tempera-tures, we had not intended to “set this down as oneof the characteristics of bacteria by which theymight be recognized,” since we did not know anybacterium that would not withstand low tempera-tures. We were only dealing with bacillus icteroidesas the svpposed specific cause of yellow fever, andsince by a natural law the specific agent of this dis-ease is arrested and destroyed by a temperature lessprotracted, and considerably less in degree, thanthe temperature required to destroy bacillus icte-roides, we were justified in emphasizing this fact.We leave, however, to F. G. Novy, the easy taskof answering this part of Sanarelli’s article.

At the time we began our experiments with feed-ing bacillus icteroides to young hogs we were wellaware that every precaution should be taken toavoid error in experiments on these animals, asemphasized by Welch and Clement.2

1 Centralblattfilr Bacteriologies etc., Band xxv, No. 18-19,1899.2 Proceedings First Veterinary Congress of America, Chicago,

1893.

Our first experiment was made in a room where adog had already died some weeks before from theintravenous injection of bacillus cholerae suis. Onthis account we did not attach too much importanceto this observation. We proceeded at once, how-ever, to procure a second-story room with cementedfloor which had never been used for animal experi-mentation. We had it thoroughly wiped out witha disinfectant solution; we also covered the windowswith mosquito-netting in order to prevent the in-gress of flies. A new broom and mop were pro-vided for this room. We placed therein four newwooden boxes, and in each of these we put a younghog on April 6, 1899. These animals had beenpurchased in open market, and were fat and healthyin appearance. April Bth, after withholding foodfor twenty-four hours, two of these pigs were fedwith a twenty-four-hour-bouillon culture of “ba-cillus icteroides, original,” passed through oneguinea-pig for the purpose of ascertaining whetherthe bacillus grown on agar for a period of seven-teen months was still virulent or not. On April10th, 1899, the remaining two pigs were fed, un-der like circumstances and with a culture of bacillusicteroides having the same source. The culture ofbacillus icteroides was in each case mixed with asmall quantity of milk brought directly from thedealer. The temperature of the hogs was takenboth before and after feeding, and we were carefulto have the thermometer placed in a disinfectant so-lution immediately after being used. The hands ofthe attendant were also carefully disinfected on eachoccasion. Under these conditions, and with theseprecautions, the four animals promptly sickenedand died, as shown in the following table:

Careful cultures were taken in each case from theseveral organs, the blood, bile, and mesentericglands. From the liver in one case, and from themesenteric glands in all other cases, colonies of asmall, quite motile bacillus were obtained. This ba-cillus was promptly agglutinated by high dilutionsof Sanarelli’s serum. From the other sources cul-tures were negative.

Now we think that looking at these experimentscritically, there were two chances for error to havecrept in; but not through lack of technical precau-tions or neglect on our part. In the first place rhaving injected the guinea-pig with “bacillus icte-roides, original,” we might have recovered the hog-cholera bacillus from his blood and organs, but thishardly seems probable. Our cultures were purefrom this guinea-pig, and we have every reason tobelieve that we recovered bacillus icteroides. Inthe second place, as we only kept our young hogsfor a few days before feeding them the culture, we

9

Young Hogs Fed with Bacillus Icteroides.

cannot say that some one of these hogs may notalready have been infected with the hog cholera ba-cillus. They were bought in an open market, andwe kept no controls for these first experiments, notbeing able to obtain pigs from the same litter. Butadmitting this chance for error, we can hardly be-lieve that all of these hogs were already so badly in-fected with the hog-cholera bacillus that they shoulddie within twelve days from the feeding, and thatthe symptoms of illness, such as diarrhea and loss ofappetite, should have so promptly appeared afterthe inoculation withbacillus icteroides. If the hog-cholera bacillus were present in one, it was presentin all, and it is remarkable that the disease shouldhave run such an acute course after natural infection.We are informed by the dealer from whom thesepigs were purchased that they had been in his pos-session for about a week when sold to us. He alsohas stated to us that during the present year,although retaining his hogs for periods varying froma few days to several weeks, he has not had any pigsicken or die on his hands.

We think, therefore, that we were quite justified,as the result of these experiments, in expressing theopinion that bacillus icteroides was a variety of thehog-cholera bacillus. We may state that nothinghas occurred since the publication of our prelimi-nary note to cause us to change this opinion; butthat our subsequent results, such as the immunization

of guinea-pigs with gradually increasing doses of thedead culture of bacillus icteroides against a fatal doseof the hog-cholera bacillus, and securing the sameresult by injections of the dead cultures of the latterbacillus against a fatal dose of bacillus icteroides,have rather confirmed us in that opinion.

We have also been permitted to study sections ofthe liver from the case of a soldier who died in Ha-vana during the present year, and from the micro-scopical findings we are strengthened in the beliefthat bacillus icteroides should be considered as asecondary invader. From the blood of this caseduring life, bacillus icteroides was recovered bothby the members of the Havana Commission and byA. Agramonte. 1 This bacillus was also obtained oncultures after death by the same observers. It isimportant to note, however, that from the spleen,after death, the typhoid bacillus was also obtainedby Agramonte. Careful study of sections of theliver in this case shows an entire absence of fattydegeneration, while many areas of necrosis, such ascharacterize the liver of typhoid fever, are to beplainly seen. In other words, bacillus icteroides,though present during life in this patient’s circula-tion, failed to bring about that “acute steatosis”upon which Sanarelli lays so much stress.

We hope, at an early date, to present our com-plete report to Surgeon-General Sternberg.

1 Centralblattfur Bakteriologie, etc., Band xxv, No. 18-19,1899.

No. Date. Quantity. Result. Lesions. Remarks.

976 Apr. 8, ’99. 25 c.c. 24-hour culture. Died on 9th day. Diphtheritisof stomach and largeintestine.

Diarrhea and fever April10th, with thin, pea-soupstools. Weakness, ofhind extremities.

977 Apr. 8, ’99. 15 c.c. 24-hour culture. Died on 6th day. Ulcers on lips; hyperemia offundus of stomach; ulcera-ted Peyers patches; follicularulcers in large intestine withcork-lining exudate.

Diarrhea and fever, begin-ning April 10th. Ema-ciation; loss of strengthin hind legs.

978 Apr. 10, ’99 20 c.c. 24-hour culture. Died in 7 days. Congestion and ulceration ofmucosa of stomach. Ulcer-ation and diphtheritis ofcecum and colon.

Diarrhea April nth andsucceeding days. Ema-ciation. Loss ofstrength in hind ex-tremities.

979 Apr. 10, ’99. 15 c.c. 24-hour culture. Died in 12 days. Erosions on tongue. Diphther-itis of stomach and esoph-agus. Ulcers in stomach,small intestine, and cecum.Numerous small circum-scribed thickenings in sub-mucous layer of large intes-tine. The centers of some ofthese already undergoingnecrosis.

Same clinical symptoms, be-ginning on April 12th.

The Medical NewsESTABLISHED IN 1843.

A WEEKLY MEDICAL NEWSPAPER.Subscription, $4.00 per Annum.

The American JournalOF THE

Medical SciencesESTABLISHED IN 1820.

A MONTHLY MEDICAL MAGAZINE.Subscription, $4.00 per Annum.

Progressive MedicineQUARTERLY, 400-500 PAGES, ILLUSTRATED.

Subscription, $lO.OO per Annum.

COMBINATION RATESALONE IN COMBINATION

American Journal $ 4.00 % 11lMedical News 4.00 j $7-5° I $15.00 iProgressive Medicine io.co J 15-75 j.^6Medical News Visiting List.. 1,25 jMedical News Formulary ... 1 50, net J

• In all $20.75 for $l6“Progressive Medicine,” and “Journal” or “News,” $l3.“Journal,” “News,” “Visiting List” and “Formulary,” $8.50.

LEA BROTHERS & CO., PUBLISHERS,706-8-10 Sansom St., Philadelphia. m Fifth Ave , New York.

The specific cause of yellow fever :FRONTCover page

MAINChapterUntitledUntitled

BACKCover page

TablesUntitledUntitled