3"6015 OO2O7 564 University of Michigan - BU HR t | 2 | / º º º, º, | A. º, |, º AW/ AºA º ºſº/º º: º *. A. gº Tº - º º ſ § Z. d ſº - , º ." º AA º , ºr | º ſº, \ iſ. A.A. º ſ N | * * * &. º.º.º. A 2 - .” - * * * ſ ſ º 5. . . . . . . ; º tº w * . . . º º .* Y A ** | * ºf 2. ' ". 5 ſ º ſ/ * f º º * * t /. . Q º # ºl. , ſº “, | * * ; ; , º º W. ſº ſº. W.; H | 2 | " . . . º a . ; , ; º º g & Kº & º § * º sº jº Aºº | º º º º º º: tºº ſº /* ſº : |. f ſN* º A. ſº |; | N | -| | R || . .' ' ||". |-º: d * : * .|ſ sºdºd7.*||ſs | ºg.§ºſ . N N ºº | ſº ſº. 5. At º 's º º -ºš . , , , , "... * N -- N. - ~ * "C ºr N \, \! N W s ſ N N N \\ Nº ; , , , º * , *s N w M A Tº ' . º, N A *\\ º S. \ſ." N N Nº C C : A N R . . . . S. C a--- *x s N. " N & ſ ‘s W ſ s º N. : x ~, "Nº sº I \, \! § ſº º } ºf N c. - - W \ * , , , , ? º •---- · * • {• |- |-. w · *~~ ~ ~ §. ~~ +··--- |--·s]-· ··`ſ - ĖJĮĮĶĶĹĹĻĻĽĽJŪRĘ EĎ~！\§§§$ff###########№= <！-- - - - - - ~~~~ ~~~~ ！± EË\\$}}|}}}|[[[[[[[[# № E\\$sſſſſſſ！！:HOE. S.}}ſ'，}ſ= №ſ ſii № ~0.5- }}| F Sſ{{E ## { §||ſā B## §||;№ §§§3– §@₪$# }\，%\，}5ē}} （T_jēķ| Ķ7;§§Ź,■■~：----| №ſ •%ſ，%ſ，%ſ|\||\\]]{{IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIEſjä. İİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİİ.ÎÎÏÏĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪĪiſſiſſiſſä ·- <----- t .-- ·----w |×,，- --< - •.. ' "-|-~~ ~ ~ .|- . . -- -*•«… ---- Tį ·•---- - * … -| -7.- . '…^_|- -'.----، ... •- -|-·- |-* （ .---{ --*----- （~~~» - ... ~ ·º ~ •^-,“…” , \ .~~·- |-|-- ----}|--· •, ! , ----- -*- . ;- _r ，----... ···… ” - *-+ · · ?|-<！---|- / -~~ ~- “... ~~|-- v.·|- -– • "--~~ -}- |--*~ ~~ |× ----· --|-- - º !-- ... -- - •.-~...,|-¿• ~- • - |× ·+ ...’~）;•- i}«.--- - - ~... } ， ·"…~-|- -· · ·t -•·; · J -* :... ~ · · · -~- - -,^| ſ – ! ~~- |-•-- ſ º- -…º ~• - •-|- |-~ - }|-·• … …* ·→, '|- { •|- - -* ..»" |-'.•• .- -- u · *- -a.---- į, *»?* ►»）.*ae→ %，*----*„ ”*ș ſº*-+-+�ae” **** „+”■**+*«. }，**<.} ¿?*.* w **º ºr ----**ą. *„**-**** …{ ----****{ **¿●* / 33*~~→ ---- 4…，-· *----→|-*-- ... • •º,~,|-|’， • s}*！» Ķ· , ！*^•• •！”*- … �** * →-----*•-* \,;~*--***•••！\~ -º#«. *& �>& ： *<.*ær…º•* ~-）, º*3 » ' …* &vºj*ae *--**ş ~* ^ ） »�•}**!»„**~****----… =?.**} ×ș^{****rºſ,~,ŞJ #*（?：%*~~~， ,*` v.z.så}\~· §§） *（\&\-?) （; * * *.}^^}{ §.-+ *****4*, ，， ***Ļ** |×*.**…º.*** șºĢ. *§x**, •*}*-----}*--> �*«*, , , *{\T^º4 ſ«** ,- **º șºx → *,\,«ș && (* *.*‘ą-„***ą، º'，}• （）（* ;ºs;„ ' ^）-*…*~^ £**ą,*****\,~ ~ … 2±- * * *yº | }*, •}-} � *~ ș ~ **?**）, *）. •^*…，•? *， **• �* „. “，**→…šºr_;----{ ºg，�|-** •*sºſ*@;*-*• • •'; ſ.ſ.º ^ ^ ^ ^ _ ^ ^ „ -?^ „‘ą' -<. -«,! ***T！ ---- , º* „…#* ººJ ^_ --* *• „ “... , “ } ;º.º.）.§ ~; ~.*' &*\,!*w_&*ș„~ ~---- }~）;• ºº ºz. ）^ •}\, ，|×.*¿.---- •**|×**** *.#$wº» "&•ş~Zº;” **---- *---*）. **, *）（ * ? – ， ,}„» º（ , ） ----→→ •« …, *'，A..*，ę*ºs_,· **;*&*-*ae4„“.* .}} …?� å.* …*---- →→→•-*---} ¿.*¿¿.* , ° ′> !… ;--~~» \^ ~\,.*，? }&#*）^#*（?:.*）$* … • ~~«-ºr ķº “A №sº,&！” „łºș -**.**}*----*---- §§șğ&] ??? :-$§.*¿.*’ ,<ſ----Åy�*，**_■ *~ Ķſeº, º º}·&） ; * ~~~~ ° §„fººt rY* ,*~~ ----* ,\* * §§ § §.*,·,≤ ſ” ~ ~ ！,’ ’ ”~ Lº " […！º~º ~º} { , și º gºsº“, ſº*[× »>**----••→ • • ►► §§§$ $Ř„’’ * * · · ·.*&ș**•„**\,* §§§§ſ*Řºw's,3*****<<！ș •·**�.** å ···ae,|-ș. *** § →ģzřů º*… ► | ►►*，** **）* \，.**__-^*×»→ş {&&§§§*»|×*, ** ~*~） 3.35.& ar.<^* --- *...** &ºr; --- resºr: : **. ºf & v. -> 3. < * 7. *~3.4 r. .** 3’ # .* sº zºº. yº º sc. * * * , ‘. . . . **. - . " § ºº- *:::::::: *… . - • º º * *- 4. §. 3. º 2 : . . . . ~~ - - - ** * º * & % * FROM THE |EPARTMENT OF PATHOLOGY & -zº. OF THE & © () ollege of fluim and $num º COLUMBIA UNIVERSITY N. Y. VOL. IX. FOR THE COLLEGIATE YEAR 1903–1904 IETERIE’IETINTS NOTE. These reprints of the more important studies published by the workers in this department in various journals during the collegiate year 1903–1904 are brought together in this form for the convenience in reference of those interested in these themes. It has seemed to us that by this form of publication—first, the wider circulation which current medical journals afford, and the subsequent grouping together of reprints—the aims of such varied studies are more certainly accomplished than they would be by a special department or university publication. The cost of reprinting and issuing these studies is defrayed by the Alumni Association of the College of Physicians and Surgeons. T. MITCHELL PRUDDEN, Director. DEPARTMENT OF PATHOLOGY COLLEGE OF PHYSICIANS AND SURGEoNs CoLUMBIA UNIVERSITY, NEw York November, 1904 l50906 [Reprinted from THE MEDICAL NEws, Feb. 14, 1903.] & A STUDY OF A BACILLUS RESEMBLING THE BACILLUS OF SHIGA, FROM A CASE OF FATAL DIARRHEA IN A CHILD; WITH REMARKS ON THE RECOGNITION OF DYSENTERY, TYPHOID, AND ALLIED BACILLI.” RY PHILIP HANSON HISS, J.R., M.D., OF NEW YORK ; INSTRUCTOR IN BACTERIOLOGY AND HYGIENE, COLLEGE OF PHYSI- CIANS AND SURGEONS, COLUMBIA UNIVERSITY, NEW YORK ; AND F. F. RUSSELL, M.D., ASSISTANT SURGEON, U. S. ARMY. SEVERAL reasons have led us to report the re- sults of our study of the organism isolated from this case of fatal diarrhea. Chief among these is the vital importance of establishing beyond doubt the etiology of the various forms of diarrheas and dysenteries. Upon such establishment of etiology must ultimately depend the rationale of hygienic, diagnostic, and therapeutic measures— especially those of serumtherapy, and diagnosis by serum reactions—upon which the recognition and cure of the disease or protection from the infective agent must be based. Moreover, there is the greatest need, in the present state of our knowledge of the etiology of dysentery, of the most careful study, description, and especially comparison of the bacterial forms supposed, in different cases and various localities, to be of etiological significance in the same or various forms of this disease. As will be brought out more fully in detailing the peculiarities of the organism we are about to describe, the organisms isolated from cases of dysentery—apart from a determination of super- ficial resemblance in growth characters on the usual nutrient media and their reactions in the presence of the more commonly employed sugars —dextrose, lactose, and saccharose—have been largely identified by their supposed specific agglu- tination reactions in the presence of serum from dysentery patients and immunized animals. That there are serous objections to placing too com- plete confidence in such measures alone, we be- lieve will appear from our study. .* Read at the regular Meeting of the New York Pathological So- ciety, January 14, 1903. The bacillus of acute dysentery, or diphtheritic colitis, first described by Shiga" in 1898 as etio- logically connected with the disease in Japan, and the bacillus isolated by Kruse” in 1900, from the same disease in Germany, have finally, after much comparison and discussion of minute differences, been recognized as identical species. Under the stimulus of Flexner’s” work, the same species has undoubtedly been isolated from cases occur- ring in America, as is shown by the correspond- ing cultural characters and serum reactions of the organism isolated in New Haven", Conn., and known as culture “New Haven.” Flexner’s culture, isolated by him in 1899 while in the Philippines, is not admitted by all observers to be identical with the preceding, although Kruse— aside from the general weaker growth of the Flexner bacillus, and minor differences in its agglutinating reactions, which are, as a rule, less marked in serum from dysentery cases than the reactions of “Shiga,” “Kruse” and the “New Haven” cultures—could determine no specific differences, and concluded that the variation was not greater than that occurring among various true typhoid cultures. Shiga”, in his latest article, is practically of the same opinion, and claims to have gotten similar reactions to those given by Flexner’s bacillus with cultures of his own, when these were grown for several generations in milk. From this he believes the Flexner bacillus to be simply a slightly modified form of the true bacillus dysenteriae. Martini and Lentz", however, and particularly Lentz" in a separate article on the cultural differ- entiation of these organisms, claim to have found not only discrepancies in agglutination, but cul- tural differences of fixed value in the Flexner bacillus, as brought out in mannite-litrnus agar (Drigalski’s medium). To a discussion of this point we shall return later. The bacillus which we are about to describe was isolated from the contents of a section of the colon of a child dead of an acute diarrhea. The patient was a child about one year old, who suf- fered from an attack of enterocolitis. The tem- perature was not high, never over one hundred. There was persistent diarrhea, from four to six stools every day. For the most part they were green and contained mucus. In the history there was no mention of any blood in the stools. The child lost strength progressively and died in two 3 weeks. * [We are indebted to Dr. David Bovaird, for the material examined, and desire here to: express our appreciation of his interest in this. work.] - Method of Isolating the Organism.—The method employed by us for isolating the Organ- ism was as follows: The contents and Scrapings from the mucosa of the gut were diluted in broth. and platings were made from this emulsion by transferring one to several loopfuls to the medium. recommended by one of us (Hiss)* for isolating. the bacillus of typhoid fever. In this medium. the colonies of the bacillus of Shiga do not form threads, as do the typhoid bacilli, and may usually be distinguished from the colonies of colon bacilli. and allied forms by being smaller and lighter in. color. Transplants were made from the Small, light-colored, transparent colonies which were: present, into tubes of the semisolid medium de- scribed by Hiss” in 1897. By the use of this me-- dium, which appears to us much superior to the glucose agar recommended by other workers, Or- ganisms corresponding in character to the Shiga. bacillus may be readily recognized. In this semi- solid medium, when incubated at 37° C., the bacillus of Shiga, possessing no active motility, does not spread in the medium and cloud it, but growth is confined to the original line of inocu- lation. Neither does the dysentery bacillus form gas in the medium, from the one per cent. of glu- cose, which is present. All colon bacilli form gas. in the medium and are also distinguished by other characters, more or less marked, which depend. upon their degree of motility. Typhoid bacilli are readily differentiated by clouding the medium uniformly. - If some such semisolid medium as this had been employed, instead of dependence being placed on the microscopic test alone, much of the uncertainty and consequent discussion regarding the motility or non-motility of the bacilli described by Shiga and Flexner, it seems to us, might have been avoided. As pointed out by Hiss some time ago, the character of the growth of the bacillus of Shiga in this semisolid medium in tubes and its inability to form threads in the typhoid plating medium * Since making this communication another case of fatal entero- colitis, in a child of nine months, has been investigated by me, and a bacillus of the same species as the one here described recovered from the gut contents.— A'fss. f This medium is composed of agar, 15 gm.; gelatin 15 gm.; Liebig’s meat extract, 5 gm.; sodium chloride, 5 gm.; dextrose, Io. gm. No acid or alkali is added. See ref. 8, p. 156. 4. were among the exceedingly few cultural char- acters until recently known by which this organ- ism could be separated from the bacillus of typhoid fever.* Description of the Bacillus.--An organism an- swering to the characters shown in the semisolid medium by the Shiga bacillus was found to be present in the gut contents of our case, and was subjected to various tests to determine, if pos- sible, its further identity with various cultures of Shiga's bacillus. Three cultures of this organism sent to us by Professor Flexner have been studied. These, as has been stated, are the cul- tures known as “Kruse,” “Flexner,” and “New Haven.” “Kruse” is the culture isolated during an epidemic in Germany. “Flexner” is, we be. lieve, one of the , cultures obtained by Professor Flexner while in the Philippines, and “New Haven,” sometimes referred to as “Seward,” was from a case of dysentery occurring in New Haven, Conn. Morphology.—The organism is a short, fairly stout bacillus, with rounded ends, in some cul- tures almost coccoid in form. It occurs singly, but often in pairs, and rarely short threads may be seen. There is a tendency in some prepara- tions to occur in clumps. The bacillus stains with the anilin dyes, but as is the case with Shiga's bacillus, the staining of all the bacilli in a prep- aration is not equally intense. Near well-stained individuals it is the rule to find many which have taken the dye very poorly. The staining of the protoplasm of individual bacilli is also at times irregular. The organism is decolorized by Gram’s method. Spore formation has not been observed. The bacillus is not motile. Flagella stainings have not, so far, been made. No cap- Sules have been noted. In general morphology and staining reactions the Organism answers most closely to that of the dysentery bacilli as represented by Kruse's cul- ture. Flexner’s bacillus is not so stout and is longer, than ours, thus corresponding morpho- ºly more closely to the bacillus of typhoid ever. Cultural Characters.—The bacillus grows well on the usual culture media, and corresponds on these in all essential respects to the Shiga bacillus. Gelatin is not fluidified; nor is indol, according to Our tests, produced. On the usual nutrient * Hiss : Jour. Med. Research, 1902, v1.11, p. 165. 5 broth, which is clouded uniformly by the growth, a small pellicle may be formed after ten or twelve- days at 37° C. When a fermentable sugar is present in the broth no pellicle is formed. The same observation holds good for typhoid and colon bacilli. This absence of pellicle is probably accounted for by the ease of obtaining oxygen from the sugar and subsequent inhibition of growth by acid production. A failure to recog- nize this fact has undoubtedly led to discrepan- cies in descriptions of organisms. Some faculta- tive organisms, usually not forming pellicles, will do so in sugar free media, or at all events grow more abundantly in the upper portions where the necessary oxygen is present. None of our cul- tures of true Bacillus dysenteria, have, however, formed pellicles. Fermentation Tests.--A study of the organism in dextrose, lactose, and saccharose bouillon in fermentation tubes shows a complete correspond- ence with the bacillus of Shiga, as represented by the cultures “Kruse,” “New Haven,” and “Flexner.” All of these organisms ferment dex- trose with the formation of acid and a clouding of the open and closed branch of the fermentation tube. Neither lactose nor saccharose are fer- mented by any of these organisms, hence are not available for anaerobic growth, and the closed branch of the tube remains clear. These bacilli are not able to grow anaerobically in broth not containing some available carbohydrate or allied Source of Oxygen supply. It has been the common practice in our labora- tory for several years past to test organisms, es- pecially those of the typhoid, Gärtner, and colon groups, in one per cent. mannite broth in fer- mentation tubes. A test of organisms in this me- dium has in our hands always served to differ- entiate cultures of Shiga's bacillus—“Kruse,” “New Haven,” and “Flexner”—from typhoid bacilli. This point was first noted about two years ago in our laboratory by Dr. Charles Nor- ris, who in observing the fermentative activities of certain members of the typhoid and Gärtner groups, tested the cultures “Kruse” and “Flex- ner” sent to us by Dr. Flexner early in 1901, shortly after his return from the Philippines. Typhoid bacilli, Norris found, utilized mannite with acid production and clouded the closed branch of the fermentation tube, while the dysen- tery cultures could not ferment mannite, and did not grow in and cloud the closed branch of the 6 tube, which in glucose free broth always remained clear. The organism isolated by us when grown thus in mannite does not correspond in its action to the cultures “Kruse,” “New Haven” and “Flexner,” but ferments mannite and clouds the closed branch of the tube, as do the typhoid bacilli. In bouillon made from maltose of high purity it can, however, be differentiated from typhoid cultures. Typhoid cultures ferment maltose with acid production and cloud the closed branch of the fermentation tube, while our organism does not ferment maltose nor cloud the closed branch. It is well to note that cultures of Petruchsky’s Bacillus fecalis alkaligenes can be distinguished from dysentery, typhoid, and allied bacilli by its inability to ferment any known carbohydrate (at least, anerobically) or to cloud the closed branch of the fermentation tube. Serum-Water Media.-A most striking differ- entiation of organisms belonging to the dysen- tery, typhoid, Gärtner, and colon groups is af- forded by a study of these organisms in serum- water media. This medium as a basis for study- ing fermentative activities was devised by one of us (Hiss)” for use in the differentiating of pneu- mococci from Streptococci. The medium is composed of clear beef serum, one part, and distilled water, two parts.” This mixture can be boiled without causing a coagula- tion of the albuminous bodies present. The me- dium after boiling is an opalescent fluid. As was pointed out in the original communication on this subject, this medium is practically, if not abso- lutely, sugar free. The original monosaccharid content of the blood is rapidly destroyed after drawing, by the action of the so-called glycolytic ferment of the blood. That this is still active in serum may be demonstrated by adding small amounts of dextrose to sterile serum, and then allowing this to digest at 37° C. for an hour or so. Subsequent fermentation tests with organ- isms will show that the sugar has entirely or partially disappeared, according to the amount added, as determined by acid or gas production. In using the medium as a basis for fermentation tests it is necessary to boil it before adding the sugars, as many of these are acted upon by the enzymes of the blood—diastase—maltase (glu- case), etc. This may also be easily demonstrated - * Beef serum, one part; distilled water, three parts; is better as it is less opalescent, 7 by digesting the sugar, especially dextrin and maltose, for a short while in the unheated serum- water mixture, maltose being rapidly destroyed and . split into dextrose, and dextrin quickly changed into maltose and then to dextrose. The medium is therefore made by first heating the serum water at IOO’ C. for a short time, and then adding usually one per cent. Of the Sugar desired. To this is added litmus solution (Merck’s highly pure litmus, five-per-cent. Solution in water) in the proportion of one per cent. The medium is then tubed and may be sterilized at IOo° C. for ten minutes on three consecutive days without fear of changes taking place in the sugar from the application of the heat. The sugars so far tested by us are dextrose, galactose, mannite, mal- tose, lactose, saccharose and dextrin. These sugars serve eminently well to differentiate our organism from the dysentery bacillus of Shiga, and also to distinguish between these and or- ganisms of closely allied characters, such as typhoid bacilli, Bacillus fecalis alkaligenes, colon and the organisms of the Gärtner group. This is shown by the following tests of organisms: Bacillus fecalis alkaligenes: None of the sugars are fermented; all of the media remain blue and un- changed. Bacillus Shiga: Deatrose and galactose are fermented with acid formation as shown by a reddening and Solid coagulation of the medium. Mannite, maltose, lactose, Saccharose and dextrin are not fermented, the media remaining blue and fluid. This is true of “Kruse,” “New Haven” and “Flexner.” , Bacillus “Y” (Hiss and Russell): Deatrose, galactose and mannite are fermented with acid production, as shown by reddening and coagulation of the media. The fermentation of mannite differentiates this organism from Bacillus Shigae, as represented by the cultures “Kruse,” “New Haven” and “Flex- ner.” The maltose, lactose, saccharose and dextrin tubes remain fluid and blue. Bacillus “48”: From stool of typhoid fever in 1897. Same reaction as above with the exception that the mannite medium is not coagulated throughout the entire depth, but shows a blue fluid layer at the top. This organism, however, is actively motile and thus further distinguished from “Y.” Bacillus typhosus: Dextrose, galactose, mannite and maltose and dextrin are fermented with acid pro- duction, as shown by a reddening and solidification of the media. Dextrin tubes first show a change to purple, and later coagulate in the lower portion leaving for some time a blue fluid layer above. Lactose and Saccharose media remain unchanged 8 and blue. The fermentation of maltose and of the dextrin medium serve to differentiate typhoid cul- tures from our organism “Y” and from the cul- ture “48.” Bacillus “48” deserves further mention. It was isolated by Hiss in 1897 from the feces of a New York hospital patient supposed to be suffer- ing from typhoid' fever, and was thought at first to be a true typhoid bacillus, but was subse- Quently looked upon as doubtful, as it did not form thready colonies in the typhoid plating me- dium and, although clouding the tube medium uniformly, without gas production, formed a thin pellicle at the surface of the medium. This is never the case with true typhoid cultures. It has been preserved in sealed tubes until a short time ago when it was re-examined and found to differ from typhoid bacilli in its action on maltose and dextrin. Although its sugar reactions correspond very closely to those of our bacillus “Y,” it is, as above noted to be, readily distinguished from it by its motility. Agglutinating Reactions.—Unfortunately we were not able to obtain a specimen of blood or serum from the patient from whose intestine our bacillus was isolated, so that a most important link in the chain of etiological evidence is lack- ing. A study of the agglutinating reactions of the bacillus have, however, been made with vari- ous normal and immune sera, and with the serum of an animal immunized with the bacillus itself. The following interesting, and it seems to us, very important results have been obtained. Normal Human Blood.—The results of tests against supposedly normal human Sera were as follows: Serum A.—Adult man, in good health; no his– tory of bad diarrheic or dysenteric attack. This serum did not agglutinate the bacillus in any dilution. 3, Serum B.-Placental blood. No history fur- nished. Positive reaction I-20, two hours. I-30, two hours was negative. Serum C.—Adult man, in good health. Two years ago while in Porto Rico suffered from a prolonged attack of diarrhea, which was not, however, considered dysenteric in character. Positive reaction, I-80, in two hours. Negative, I-IOO, in four hours. This serum, on account of the previous diarrhea, probably should be ex- cluded from the normal list. & Human Sera, Dysentery: Serum A.—From a 9 soldier with chronic dysentery. Acute attack in the Philippines two years ago. Another attack subsequently in Porto Rico. Positive reaction with our bacillus, I-2O, one hour. “Flexner” did not react at I-2O. “Kruse,” — I-IO, - I-40; “New Haven,” + I-IO, +I-20, - I-40. Serum B.-From a soldier with chronic dysen- tery. Sent to us from the military hospital in Washington. Acute attack of dysentery in the Philippines two years ago. Our organism re- acted I-40, one-half hour. “Flexner” gave no reaction. “Kruse” was positive I-20; minus, I-40, one and one-half hours. “New Haven” positive I-20; negative, I-40, one and one-half hours. - Normal Horse Serum— Bacillus “Y,” +1-40, one-half hour; + — I-80. “Kruse,” + I-20, + — I-40. Dysentery Horse Serum.—From a horse im- munized against culture “New Haven.” Sent to us by Professor Flexner: Bacillus “Y,” + I-400, H- – I-500, twenty hours. “New Haven,” -i-I-IOO, H– — I-200, twenty hours. “Flexner,” +I-40, −I-2OO, twenty hours. Bacillus “48,” —I-I, -I-50, twenty hours. Bacillus typhosus, + — I-20. Normal Rabbit Serum.—Bacillus “Y,” + — I-2; — I-IO, I-2O, I-IOO, two hours. Typhoid Rabbit Serum.—Highly immunized against typhoid. Exact titration against typhoid Originally I-2O,OOO. Bacillus “Y,” +I-5OO ; +1-4,000. “New Haven,” —I-20. Bacillus “48,” –I-25 and I-IOO, twenty hours. “Y” Rabbit Serum.—This animal received sev- eral inoculations of boiled cultures of “Y.” The serum agglutinated “Y” in dilutions as high as I-5OO. “Kruse,” “New Haven” and “Flexner” did not react in any dilution (–I-Io; –1-500). Bacillus “48” also showed no agglutination. Bacillus typhosus reacted at I-80; --- I-IOO, in twenty hours. - Normal Beef Serum: Bacillus “Y,” +I-2OO; +1-320, four hours. “Flexner,” +-I-Io; – 1–2O. “New Haven,” +1-Io; – I-20. “Kruse,” +I-Io; – 1-20. Bacillus typhosus, +1-40; +1-80, twenty hours. - +-indicates a partial reaction. IO The results of these serum tests are important and indicate to us, after a careful search of the literature of dysentery, that we are in all prob- ability dealing with an organism not hitherto de- scribed. It is readily seen from our tests that “Y” agglutinates with dysentery serum from human infections and from immunized animals in most instances quite as well as, and in some bet- ter, than do bacilli fully identified in all respects with the bacillus of Shiga. The various dysen- tery-like bacilli described by other observers, al- though showing some agglutination in the pres- ence of dysentery sera, have been agglutinated to a markedly less degree than true dysentery bacilli, and for this reason alone have in several instances been excluded from recognition as fully identical with Shiga's bacillus. This is true of the organism described by Kruse as the bacillus of pseudodysentery of the insane. The organ- isms placed in this class were, according to him, so far as he could determine, in all respects similar to his and Shiga's dysentery bacillus in cultural characters. They were differentiated by not agglutinating with true dysentery serum, but with the serum from the insane patients. The true bacillus of Shiga, on the other hand, did not agglutinate with these sera in the same dilutions. These pseudodysentery cultures were agglutin- ated in normal sera in I-50, which is higher than that usually shown by the true dysentery bacillus. Kruse found that his true dysentery organism agglutinated in normal Sera as high as I-2O, and exceptionally I-50. In his later paper, however, he thinks this may have been due to previous in- fection. Another point of interest is the high agglutina- tion of our bacillus with sera from animals im- munized against typhoid. Here “Y” shows a very close relationship to the bacillus of typhoid. This bond is strengthened by the fact that typhoid cultures also show fair agglutination in “Y” im- mune serum. Both Shiga's bacillus and typhoid call forth agglutinins which have an affinity for the receptors of our bacillus “Y.” “Y” immune serum, on the other hand, contains some agglu- tinins active for typhoid bacilli, but none, so far as our experiments go, which affect Shiga's bacillus. * - . . . “Y” and typhoid bacilli, therefore, call forth Some agglutinins in common or have receptors capable of anchoring certain agglutinins of each. “Y,” it may be, anchors all of the agglutinins * 4. I I called forth by Shiga's bacillus and probably other agglutinins normally present in the serum tested. This is indicated by its higher agglutina- tion in the dysentery horse serum than that shown by the dysentery cultures themselves. This is the case; or it has fewer receptors that have to be satisfied before agglutination takes place. The same phenomenon is probably illustrated by Our organism and typhoid bacilli in normal beef serum. There are agglutinins present in the serum active for both typhoid and “Y.” Our Or- ganism “Y” is, however, agglutinated in much higher dilutions than the typhoid bacillus, show- ing that there is much more “Y” agglutinating substance present, or that “Y” is agglutinated with fewer receptors satisfied than is the typhoid bacillus. The action of beef serum is important, as it may be found to afford a practical means of dif- ferentiating true Shiga bacilli from organisms of the “Y” type. Shiga's bacillus, as above noted, does not agglutinate in beef serum to an appreciable degree. These facts brought out in regard to the serum reactions of “Y,” dysentery, and typhoid bacilli, have important practical bearings. Unless care- ful studies are made of all organisms resembling Shiga's bacillus, gross errors may occur in serum diagnoses based upon the use of such organisms in testing reactions; and errors may also occur in identifying such organisms themselves by simple tests against dysentery serum. This is illustrated by the fact that “Y” agglutinates better even than the true Shiga bacillus in dysentery serum, and unless true Shiga bacilli are tested against im- mune “Y” serum they may not be differentiated from each other. It is more than probable that organisms corresponding to “Y” have already been isolated and not recognized as distinct from Shiga's bacillus. This may then lead to the con- verse mistake of establishing the identity of a dysenteric or diarrheic attack with true dysen- tery by agglutinations obtained by the use of a bacillus differing from the supposed etiological organisms. These remarks are not meant to convey the im- pression that “Y” itself is not or may not be found to be etiologically related to some form of dysenteric or diarrheic disease, but are intended to bring out the fallacy of relying solely upon , serum reactions in identifying either disease or I 2 organism. It is fortunate, therefore, that the study of cultural characters has aided us in the recognition of differences in these organisms and may be of further service in the recognition and grouping of new organisms as they are discov- ered. It seems to us in the face of our knowledge of the discrepancies in degree of agglutinability of such well-known organisms as typhoid bacilli, which not only show variations among the dif- ferent cultures, but in the same species under practically non-determinable culture conditions,” that we should hesitate before placing too much confidence in differences determined by serum re- actions alone, but should endeavor in all cases by careful and extended study to establish their identity or dissimilarity by every cultural and bio- chemical test available. | Arguing from the analogy of our work on im- mune animal sera, an infection due to a bacillus of the “Y” species might, if tested against typhoid bacilli, give a reaction in a dilution considered diagnostic, and Sera from both dysenteric and typhoid patients, when tested against “Y,” would certainly give reactions in high dilutions. Con- versely, “Y” might be identified either with typhoid or Shiga bacilli if dependence were placed on serum reactions alone in its recognition. Pathogenicity.—Our bacillus is pathogenic for both rabbits and guinea-pigs, but not markedly so. No characteristic lesions have so far been noted. Some hyperemia of the small intestine is apparent. Rabbits may be readily accustomed to large doses of living cultures without showing a marked loss in weight. As noted earlier in this paper, Lentz in a cul- tural study of a large number of true Shiga bacilli and lother more or less closely allied or- ganisms isolated from dysenteric patients, claims to have been able to separate by the use of a man- nite litmus agar the cultures, which he designates as “Flexner I,” “Flexner Manila” and “Strong,” from the true bacillus of Shiga and Kruse. The results of his tests indicated that, grown in this medium, Shiga's bacillus does not give rise to changes in the medium, with the exception of a decolorization of the medium in its lower layers, the medium remaining of the normal blue or deep purple in the upper section. In the media in- oculated with the Flexner and Strong bacilli, the color changes to a distinct red, due to acid pro- duction. I 3 We have endeavored to repeat Lentz's experi- ments. The medium used by him in his earlier experiments was made according to the formula given by Drigalski for his typhoid isolation me- dium, with the substitution of mannite for the lactose. In Lentz's article on Dysentery in Kolle and Wassermann’s Handbuch der Pathogenen Mikro-organismen (p. 32 I) the method of prep- aration is somewhat simplified. This last method is the one followed by us. At the present moment we are unable to say whether our culture “Flexner” is one of the or- ganisms designated by Lentz as “Flexner I” or “Flexner Manila.” The tests of Our culture “Flexner” on Lentz's medium do not, however, agree with his published results. Flexner’s bacillus gives identically the same reaction in Lentz's litmus mannite agar as the re- sults obtained when “Kruse” and “New Haven” are tested. The culture medium of Lentz, which is used in tubes and inoculated by central punc- ture, in all three instances became slightly red- dened below, but remained blue at the surface. The medium in Lentz’s experiments with Flex- ner’s bacillus reddened. Our bacillus “Y” and the organism “48,” as well as true typhoid bacilli when grown in the medium give rise to a reddening of the medium throughout its entire depth. These results were what we naturally expected from our knowledge of the metabolism of these organisms. “Kruse,” “New Haven” and “Flexner” do not utilize man- nite either in the presence or absence of oxygen, or give rise to acid from it. Bacillus “Y” and “48” and typhoid bacilli ferment mannite with rapid production of acid, and hence redden the entire medium. This was to be expected from their fermentation tube reactions and the results of our experiments with serum-water plus man- nite. An explanation of these discrepant results may, however, be found in the fact, which seems to have escaped the notice of both Drigalski and Lentz, that the medium to which they add sugars, such as mannite, lactose, etc., to be tested, is not glucose free. No mention is made of the removal of muscle sugar from the nutrient infusion used by them, nor is this recommended in the direc- tigns for preparing the medium. The presence of a fermentable substance, not mannite, in Lentz's medium, is clearly demonstrated by the redden- ing and possible subsequent clearing of the lower I 4 portion of the medium when tested with our cul- tures “Kruse,” “New Haven” and “Flexner,” which do not ferment mannite. That the medium remains blue at the top simply indicates that in the presence of oxygen the alkali production neu- tralized or exceeded the production of acid. We have made up samples of the medium from sugar-free meat infusion and find that in tests of cultures “Kruse,” “New Haven” and “Flexner” no reddening of any portion of the medium Oc- curs. Furthermore, when the medium is used Soon after heating, the usual absorbed oxygen being then absent, the growth is confined to the surface, or upper parts of the inoculation line, since these organisms do not grow anaerobically in the presence of mannite in otherwise sugar- free media. In tests of mannite fermenting or- ganisms such as “Y,” “48” and typhoid bacilli, the glucose-free mannite medium is found red- dened throughout, and growth has taken place along the entire length of the inoculation punc- ture. None of the results of Lentz's sugar tests, al- though they may be roughly indicative, can be scientifically reliable so long as the medium used by him contains this original sugar, and the jus- tice of our criticism of his method is evident in the discrepancies noted by Lentz” himself between the action of colon bacilli on media made by him and the results obtained by Drigalski with colon bacilli. Furthermore, slight differences in the initial reaction of the medium (alkali content), which are unavoidable with their method of neu- tralization and alkalinization, would also deter- Imine differences in their color reactions. [NOTE.-Since writing the above I have had the opportunity, through the kindness of Dr. Flexner, of examining the cultures “Flexner Manila (Harris)” and “Flexner Manila (Gray).” and Strong’s Manila culture. These are the cul- tures examined by Martini and Lentz, and the culture “Flexner” examined by us is, according to our tests, not one of them. Both of Flexner’s Manila cultures, as well as that of Strong, differ from the true Shiga bacillus in fermenting man- nite, as shown by acid production. Strong's bacillus may be identical with our organism “Y,” but has not been under observation long enough for us to speak positively. Flexner’s Manila cul- tures as well as one of his Baltimore cultures (sent to me by Dr. W. H. Park) is further differ- *See foot note by Lentz; Zeitsch. f. Hyg., 1902, XLI, pp. 560, 561. . I 5 entiated from Shiga's bacillus and from our bacillus “Y” and Strong's culture by fermenting maltose. Saccharose is also fermented with acid production by Flexner’s Manila cultures and by the culture “Baltimore.” My tests also indicate that these bacilli will produce acid in the pres– ence of dextrin. Their fermentation of saccha- rose and the absence of active motility separates them from Bacillus typhosus. Their fermenta- tion of maltose, saccharose and dextrin is slower than their action on dextrose and mannite, and may not be apparent for some days. Control saccharose tubes inoculated with Bacillus typho- sus show no change, so that this cannot be at- tributed to the presence of dextrose derived from the splitting of saccharose in the preparation of the medium. Control tubes of maltose and dex- trin inoculated with Bacillus “Y” and Shiga's bacillus also remain blue and unchanged. We have, then, apparently, three groups of organisms. One represented by the true Shiga bacillus, fermenting dextrose and some other monosaccharids; one by our organism “Y,” fer- menting dextrose (monosaccharids) and the al- cohol mannite; and finally a group represented by the more active fermenters—Flexner’s Ma- nila cultures and the cuſture “Baltimore”—or- ganisms which split dextrose, mannite, maltose, dextrin and Saccharose with the production of acid, as indicated by the reddening and coagula- tion of the serum-water sugar media. This study shows plainly the inadvisability of relying upon Superficial characters or agglutina- tion tests alone in the identification of an organ- ism, or on the serum tests in the determination of the Special character of a disease, and also lends support to our supposition that organisms differing from the true Shiga bacillus have been isolated and confounded with this organism.— Hiss.] Résumé.--To review briefly : We have in this paper described an organism from a case of fatal diarrhea in an infant. This organism agglu- tinates in high dilutions with the serum from dysentery patients and animals im- munized with Shiga bacilli. Formerly this bacillus “Y” could not have been differ- entiated by valid culture tests from the bacillus of Shiga, and, the mannite test being unknown, might easily have been confounded with Shiga's Organism unless its agglutinating reactions had been carefully studied in such a serum as that I6 from typhoid immune animals, or from the nor- mal beef, in neither of which the true Shiga bacil- lus, according to our tests, agglutinates in appre- ciable dilutions. - The bacillus described by us is also closely re- lated to Bacillus typhosus, both in agglutinating reactions and in the usually observed cultural characters, but can be separated from it by the absence of motility and by its reactions in the maltose and dextrin media employed by us. This bacillus apparently differs from all hitherto described dysentery-like or “pseudodys- entery” bacilli, in its agglutination in high dilu- tions of dysentery serum, an agglutination in Some instances more marked than that shown by the true Shiga bacillus. A description of this organism, apart from its connection, which may or may not be etiological, with the case of diar- rhea, has seemed to us important, since tests of its ability to agglutinate—too often alone relied upon—might lead to its being confounded with Shiga's bacillus. A subsequent use of such an organism for serum diagnosis might thus give rise not only to confusion in determining the “specific” character of a 'disease, but to untoward results in the application of serumtherapy. Its etiological significance we have not been able to determine, but its close physiological con- nection with the dysentery bacillus on one side and the bacillus of typhoid fever on the other make it not improbable that further research will show it to be associated etiologically with certain diarrheic or dysenteric attacks. REFERENCES. I. Shiga: Centralb. f. Bakt., 1898, Bol. XXIII, p. 599; Bd. XXIV, pp. 817, 870, 913. 2. Kruse: Deutsche med. Woch., 1900, XXVI, p. 637; 1901, XXVII, pp. 379, 386. . . 3. Flexner: Bulletin Johns Hopkins Hospital, 1909, XI, pp. 39, 23 I: £h. Med. Journal, 1900, VI, p. 4I4; Centralb. f. Bakt., 1901, Bol. , p. 449. 4. Vedder and Duval: Journal Experimental Medicine, 1902, VI, p. 181. - 5. Shiga: Zeitsch. f. Hyg., XLI. H. 2, p. 353. 6. Martini and Lentz: Zeitsch. f. Hyg., 1902, XLI, H. 3, p. 540. 7. Lentz: Zeitsch. f. Hyg., 1902, XLI, H. 3, p. 559. 8. Hiss: Journal Medical Research, 1902, VIII, p. 148. 9. Hiss: Journal *peºple. Medicine, 1897, II, No. 6, p. 677. Io. Hiss: Centralb. f. Bakt., 1902, XXXI, p. 302; Science, 1902, March 7, p. 367. EXPERIMENTS RELATING TO THE QUESTION OF FIXATION OF STRYCHNINE IN ANIMAL TISSUES.” S. J. MELTZER AND G. LANGMANN. (From the Pathological Zaboratory of the College of Physicians and Surgeons, Columbia Unizersity, AVezw York.) When only a toxic dose of Strychnine is injected into an animal the subsequent characteristic tetanic convulsions cease after a shorter or longer time without leaving any sequelae. Why the spasms should not continue indefinitely has always been a subject of various speculations. One of the earliest theories was that strychnine enters into a fixed combination with the tissues or becomes oxidized within the body. Dragandorff," however, proved that strychnine is excreted in the urine without any alteration. It has been further shown by Adam * in a case of human poisoning, and was experimentally confirmed by Kratter * and Ipsen,” that even medicinal doses are traceable in the urine within a few minutes, and that its excretion is completed in less than forty-eight hours. The recent discoveries of the fixation of toxines by specific body tissues gave an impetus to the revival of the old theory that the animal tissues are capable of fixation also of alka- loids, especially of Strychnine. Attempts were made to sup- port this theory by two different experimental methods. One is the mixing of Strychnine with some tissues and then injecting it into an animal. Thus Widal and Nobécourt” mixed Strychnine with the pulp of brain, liver, kidneys, etc., and injected it into animals; they report that the toxic effect was indeed perceptibly impaired. It will, however, be suffi- cient to mention the experiments of Thoinot and Brouardel 6 who have mixed Strychnine with such inert substances as talcum, charcoal, starch, etc., and have observed a retarda- tion in the effect of the strychnine, to show that the factor in * Received for publication, Dec. 14, 1902. I9 2O MELTZER AND LANGMANN. these retarding influences is a mechanical One — the impair- ment of the process of absorption. Another method is that which was introduced by von Czyhlarz and Donath,' who tried to prove the theory of the fixation of Strychnine by animal tissues by an ingenious experiment. It is chiefly the interpretation of this experiment which engages our atten- tion in this paper. The experiment of von Czyhlarz and Donath, who have published it in 1900, is as follows: After tightly ligating one hind leg of a guinea-pig, they injected a lethal dose of Strychnine below the ligature and found that, when the ligature was removed a few hours later, the char- acteristic effect of strychnine was not forthcoming. They believe the only explanation possible of this observation is that “by a prolonged contact with the subcutaneous tissue, the muscles, and the blood contained therein, the poison be- comes in some way fixed or neutralized in vivo.” Such a conclusion would lead us again to the now familiar idea alluded to before. The idea of toxins being fixed by the cells or fluids of the animal body is now quite a familiar one. Its extension, however, to alkaloidal poisons should not be accepted without repeated experimental confirmation. Soon after the publication of von Czyhlarz and Donath we re- peated these experiments * on guinea-pigs, animals used ex- clusively by these authors, and extended them also to more Sensitive animals like frogs and rabbits. As to the latter, our results were entirely negative, the prolonged ligation hardly caused even a retardation of the onset of the spasms. In guinea-pigs, an animal which is known to be somewhat refractory to Strychnine, the experiment yielded positive results, but only then, when the dose administered was not more than the toxic minimum. For instance, eight milli- grams per kilo animal brought out invariably a fatal tetanus. Judging from these and other experiments, we Suggested that the efficacy of Strychnine in the above cases was princi- pally diminished, not because the strychnine was fixed by the tissues, but because the absorption was materially im- paired in the ligated extremity, so that the Strychnine could be carried slowly and in very small doses at a time into the FIXATION OF STRYCHNINE IN ANIMAL TISSUES. 2 I blood, whence it was soon excreted again before it had a chance to build up an efficacious minimum. Kleine,” who soon after became engaged in the same sub- ject, has proven that strychnine appears in the urine even while the constriction is still on. It is, therefore, possible that, owing to the excretion of a fraction of the dose of strychnine in the urine, the dose remaining in the leg when the ligature is removed, is reduced to a sub-minimum one and is consequently ineffective. Furthermore, there is another essential factor to be thought of. During the long period in which the leg is being con- stricted the Strychnine has an opportunity to thoroughly mix with the blood and lymph of the extremity. Now it is an established fact that strychnine is much less efficacious in a colloid fluid than in a crystalloid. We may add that even for this reason alone the control experiment of von Czyhlarz and Donath is inadequate. When these authors found that Strychnine, injected into a ligated extremity shortly before untying the ligature, brought on the characteristic symptoms, we have to remember that the two experiments cannot be compared for the reason that the injected drug in the latter experiment was in an aqueous crystalloid solution and not, as in their primary experiment, in an albuminous colloidal form. In our previous paper we have stated other facts which show the insufficiency of their control experiment. In the later and more extensive publication of their ex- periments" von Czyhlarz and Donath have not modified their original opinion, although they take notice of our objec- tions in a foot-note, and have even partially confirmed our experiments. - Two theories, then, are offered for the explanation of the Surprising result in the experiment of von Czyhlarz and Donath. On the one hand it is assumed that the effective minimum is never reached after the release from the ligature, because: first, a part of the strychnine has been already ab- Sorbed during the period of constriction and has been par- tially eliminated; secondly, the strychnine is absorbed slowly and with difficulty from the ligated extremity because 22 - MELTZER AND LANGMANN. it is suspended there in a colloidal solution, and because many of the blood capillaries and lymphatics have been de- stroyed and obliterated by thrombosis, etc., after a tight con- striction lasting for hours. On the other hand, a new and far-reaching principle is proclaimed, namely: that the Strychnine combines or is neutralized in vivo by the animal tissues. It seems to us that there should be no difficulty in deciding which interpretation is more simple or more plaus- ible. Ehrlich has repeatedly emphasized, and only recently so,” that the alkaloids, being foreign to the animal organism, are not capable of forming with the protoplasm such stable compounds as are known to be formed by toxins. We should also remember that there is no adaptation or immu- nization for strychnine. The view of von Czyhlarz and Donath, however, was re- cently supported by new and thoughtful experiments re- ported by M. Carrara.” - This author repeated successfully the experiment of von Czyhlarz and Donath in nephrectomized animals. In conse- quence of the double nephrectomy, which operation obliter- ated the principal path of elimination, Carrara assumes that the poison, even if absorbed very slowly and in very small doses, ought to be retained almost in its entirety within the animal's circulation, and if, nevertheless, the animal fails to manifest any symptoms of Strychnine poisoning the failure cannot be due to slowness in absorption, and therefore is due to fixation of the strychnine by the tissues. This argument appears at first sight quite convincing. It rests, however, on a premise that has hitherto not yet been tested by experi- ment, viz.: the assumption that nephrectomy necessarily leads to a cumulative effect of Strychnine. This assumption is not necessarily correct; we know, for instance, that in ad- vanced cases of nephritis urea is excreted in situations where normally not even traces of it are found. Possibly, there- fore, a collateral vicarious elimination for other substances may also be established in nephrectomized animals. Guided by this consideration, Meltzer and Salant” have studied in an extended series of experiments on nephrectomized FIXATION OF STRYCHNINE IN ANIMAL TISSUES. 23 rabbits whether subminimum doses of Strychnine have a cumulative effect. They came to the following conclusions: if the doses of strychnine are not too large nor injected in too short intervals, such nephrectomized animals will tolerate the gradual injection of the triple lethal dose without exhibiting the least effect of strychnine. The above-mentioned pre- mise, therefore, taken as a matter of course by Carrara, did not stand the test, for subminimum doses had no cumulative effect in nephrectomized rabbits. Both von Czyhlarz and Donath, as well as Carrara, experi- mented on guinea-pigs, which, as we have already noted, are much less sensitive to strychnine. We thought it, therefore, necessary to test the premise of Carrara directly on guinea- pigs & From a large number of experiments we quote here a few abridged protocols: (1.) Guinea-pig, 420 grims. ; double nephrectomy; during the subse- quent twenty-four hours injected 7 subminimum doses of nitrate of strych- nine, together 3 mgr., equals 7 mgr. per kilo; animal killed two hours after last injection; at no time a trace of hyperesthesia. (2.) Guinea-pig, 600 grims. ; nephrectomy; during 29 hours 8 doses of Strychnine, together 5.3 mgr., equals 8.8 mgr. per kilo; no symptoms of Strychnine. (3) Guinea-pig, 230 grms. ; nephrectomy; in 30 hours 8 doses of o.35 each, together 2.8 mgrs., equals 18 mgr. per kilo; animal died 6 hours after the last injection; never a trace of effect of strychnine. (4.) Guinea-pig, 300 grms. ; nephrectomy; in 48 hours 12 injections of o.45 mgr., together 5.4 mgr., equals 18 mgr. per kilo; animal had no con- vulsions, no hyperesthesia, only after the eighth injection there were some transient slight twitchings. (5.) Guinea-pig, 215 grims. ; nephrectomy; in 24 hours 12 doses of o.31 mgr., together 3.72 mgr., equals 17.3 mgr. per kilo; animal died 11 hours after last injection without any symptoms of strychnine. (6.) Guinea-pig, 395 grims. ; nephrectomy; in 72 hours 13 doses of o.6 mgr., together 7.8 mgr., equals 19.7 mgr. per kilo; animal died 4 hours after last injection without showing any effect of the poison. These protocols suffice to prove that we can speak of a cumulative effect of Strychnine on nephrectomized guinea- pigs only to a limited degree. We see that after double nephrectomy quantities of strychnine, far above the fatal dose, 24 MELTZER AND LANG MANN. can be injected without obtaining a single symptom of the characteristic poisoning, provided that we do not use too large single doses, nor that we let them follow each other too closely. We have injected into nephrectomized guinea-pigs as much as eighteen to twenty milligrams per kilo without causing even a characteristic hyperesthesia. Carrara's experiment, there- fore, in which nephrectomized and ligated guinea-pigs re- ceived only up to eight milligrams per kilo without a consequent tetanus, are by no means a proof that the Strych- nine in the ligated extremity was combined with or neutral- ized by the tissues. We may explain the results just as well by assuming that in consequence of the ligature the absorp- tion of Strychnine takes place very slowly and in minimum doses, and that, therefore, no toxic dose is accumulated in the blood. - Thus the explanation of the experiment of von Czyhlarz and Donath remains at the same point which it had occupied before Carrara's experiments made their appearance, viz., that the promulgation of the extraordinary theory of a com- bination of Strychnine with living tissues is not warranted by the results of this experiment. - Our experiments, however, have elicited some other obser- vations which deserve to be recorded, and might even assist in the solution of the problem in hand. We have established that the toxic minimum is at least not smaller for nephrectomized than for normal guinea-pigs.” On the contrary, in the general run nephrectomized animals seem to require larger doses. Meltzer and Salant " made similar observations in rabbits. This is unexpected, inas- much as with natural elimination checked, we should rather * Contrary to Carrara, we still believe that four milligrams of Strychnine per kilo are not in all cases sufficient to produce convulsions in guinea-pigs; we met with many exceptions, and it would be necessary to use somewhat larger doses if we wish to make controls for von Czyhlarz-Donath's experiments. Moreover, we found that von Czyhlarz himself” records experiments (experiment 12, experiment 16, experiment 18) in which he employed four milligrams per kilo without causing more than hyperes- thesia. We might also mention again that in repeating the experiment of von Czyhlarz and Donath we never obtained a positive result with eight milligrams per kilo. These authors themselves employed invariably only six milligrams per kilo; Carrara, however, men- tions that he succeeded even with ten milligrams FIxATION OF STRYCHNINE IN ANIMAL TISSUES. 25 expect to produce a toxic effect with smaller doses. This fact therefore demonstrates also that nephrectomy does not further the cumulative effect of Strychnine. This rule, however, does not prevail during the first few hours after nephrectomy. At this time even subminimal doses show a distinct effect in many cases. We have seen animals which succumbed to tetantic convulsions within a few minutes after doses two, two and four-tenths, and two and eight-tenths milligrams per kilo. Such injections should not be made later than three hours after nephrectomy, nor before the animals are entirely out of the narcosis. Besides marked tetanus some of the nephrectomized ani- mals showed a reaction to subminimum doses of Strychnine in a manner which we have never seen in normal animals. Without the development of evident tonic or clonic convul- sions there appear slight short jerks either of the whole body or in certain parts which continue almost without remis- Sion. On touching the animals one feels a continuous vibra- tion. Soon the animal will lie prostrated with legs extended, nose and belly touching the table, while the vibrations and twitchings persist. Ordinarily these animals succumb within an hour. Every case of tetanus not ending fatally in a short time gradually presents the above mode of reaction. These jerks gave an impression as though the system had not yet developed new ways of ridding itself of the poison, while the absorbed quantity of Strychnine was not sufficiently strong to produce regular convulsions or a continuous tetanus. Hence the continuous sub-tetanic reaction. After the last-mentioned observation the question arose whether a slight alteration of Carrara's experiment might not have quite a different result. From Carrara's description we learn that his modus operandi was first to perform nephrectomy, then to ligate a leg wherever strychnine was injected. When, under these circumstances, three hours later, the ligature is removed and Strychnine gradually begins to enter the circu- lation, the system may have in the meantime found a way to Prevent a cumulative action of strychnine. We have, there- fore, changed the order of the several acts of the experiment. 26 MELTZER AND LANGMANN. We ligated and injected a limb, and three hours later we performed a double nephrectomy. When, now, after the effect of the narcosis had passed, the ligature is loosened, it might be expected, if the Strychnine did not become fixed, that a strychnine effect will be met with, at least occasionally, as at this early stage the system had not yet elaborated its compensatory mechanism for elimination. The slow absorp- tion, therefore, might occasionally lead to a cumulative effect. This expectation was indeed realized. In eleven such ex- periments we saw, in four instances, a fatal tetanus soon after the removal of the ligature, in two cases sub-tetanic reaction appeared, in two others a distinct hyperesthesia was present. The three remaining animals did not differ from the other nephrectomized guinea-pigs (controls) which had received no strychnine, except in a certain vivacity and greed for food. It seems to us that the few positive results with fatal teta- nus offer an unmistakable proof that in these cases a lethal dose of strychnine was absorbed after the removal of the ligature. In some cases, however, the constriction alters the absorption to such an extent that no sufficient quantity of strychnine can enter the circulation before the final com- pensatory mechanism is developed; hence the negative results. Carrara's experiment, therefore, cannot affect our view. On the contrary, its modification proves conclusively that, after the ligature is removed, efficacious doses of strychnine may enter the circulation from the extremity. FIxATION OF STRYCHNINE IN ANIMAL TISSUES. 27 . : I2. I3. I4. I5. p. I58. BIBLIOG QAPHY. Beiträge zur gerichtlichen Chemie, iii, St. Petersburg, 1872. Husemann. Handbuch der Toxicologie, Berlin, 1862–67, p. 510. Wiener med. Wochenschr., 1882, p. 214. Vierteljahrsschr. f. gerichtl. Med., 1892, iv, p. 15. La semaine médicale, 1898, p. 93. La semaine médicale, 1898, p. 140. Centralbl. f. inn. Med., 1900, No. 13. Centralbl. f. inn. Med., 1900, No. 37. Zeitschr. f. Hygiene, Bd. 36, 1900. Ehrlich in v. Leyden’s Festschrift. Zeitschr. f. Heilkunde, Interne Med., 1901. Centralbl. f. inn. Med , 1901, No. 20. Journal of experim. medic., Vol. vi. l. c. von Czyhlarz. Zeitschrift für Heilkunde, Interne Medizin, 1901, THE Journal of MEDICAL RESEARCH, February, 1903, Vol. IX., No. 1. A Résumé of Some Recent Researches Relating to Cy- tolysis and Immunity. BY T. MITCHELL PRUDDEN, M.D., NEW YORK, Reprint from the MEDICAL RECORD, February 14, 1903. WILLIAM WOOD AND COMPANY., 51 FIFTH Avenue, NEw York. I993. A RESUME OF SOME RECENT RESEARCHES RELATING TO CYTOLYSIS AND IMMU- NITY.4. By T. MITCHELL PRUDDEN, M.D., NEW YORK. IN the early years of the new bacteriological era it was chiefly upon the invasive agents in infectious diseases that attention was centered. The charac- ters, sources, variations, portals of entry to the body, and fate of the bacteria—these were the dominant aspects of infection, full of fresh and dramatic interest. To-day the tide is setting toward the nature and variations of the responses of the cells of the host under such new and often baleful conditions as infection involves. It is only now that light from both sides is falling upon the questions which relate to infection and immunity that these are taking their places among related problems of general biology. I shall not burden you with classification, but it is convenient, for the sake of getting associated facts together, to consider immunity to the infectious diseases as either natural or acquired. The absolute or relative insusceptibility of turtles and fishes for tetanus, of rats for anthrax, and of the lower animals for syphilis and for scarlatina, are examples of natural immunity which we need not further consider here. Acquired immunity may be the result of a success- fully weathered infectious disease, as in the case of smallpox. Or it may be experimentally secured through the gradual adaptation of the body to infec- tious or to toxic substances without the natural manifestations of infection or intoxication. *Read before the Practitioners' Society, January 2, 1963. Copyright, WILLIAM WooD AND CoMPANY. 3 There are many ways in which immunity by gradual adaptation may be obtained. For example, by the introduction into the body in increasing quan- tities or potency of living microorganisms whose virulence has been at first diminished. This reduc- tion in virulence may be secured in various ways; for example, through the action of heat or chemicals; by drying in the air, as in the preventive inoculations in hydrophobia; by the passage through a naturally insusceptible animal, as in vaccinia. Or, on the other hand, immunity by gradual adaptation can be secured by the introduction of the dead bodies of infective microorganisms, or of various toxic sub- stances in increasing doses. But all these methods belong to current knowledge and need not be further dwelt upon. We can all readily recall the time, little more than a decade ago, when the conceptions of the nature of immunity from infectious disease, whether it were natural or acquired, were most vague and unsatis- factory. Clinical lore relating to the subject was positive; the nature of the infectious diseases in the light of the new knowledge of microorganisms was rapidly assuming definite proportions, and the varied lines of research upon animals were converging more and more upon the conditions under which acquired immunity occurs. But still the heart of the matter eluded our grasp. Bacteria and other microorganisms are in some way destroyed within the body; this was evident; the body cells do, in some fashion, acquire a capacity to resist the deleterious agencies which are evoked in infection. But the nature of these conservative processes was not clear. Various hypotheses were framed. Nutrient mate- rial in the body becomes exhausted, said one, so that the microorganisms presently die. Self-engendered poisons, said another, 1ead to their destruction. The results of the brilliant studies of Metschnikoff were formulated in the doctrine of phagocytosis. A long series of observations on the germicidal power of the 4 pody fluids led many to adopt the humoral concep- tion of the protective agencies. But now a new and, in spite of the varied 10re which had gathered about immunity, a quite unex- pected phase of the subject suddenly developed. It is not easy to realize with entire clearness to-day the feeling of incredulity with which we faced the announcement—now somewhat more than a decade ago—of the antitoxic capacity of the blood serum of animals which had been artificially immunized against diphtheria toxin. This capacity seemed all the more striking because the serum of the immune animals containing this new stuff, on being transferred bodily to fresh individuals, even though of different species or race, still carried the power to avert the specific toxic effects through which it had been itself evoked. The protection thus secured was not against typhoid fever, or tetanus, or smallpox, but only against diphtheria. It seemed fairly incredible that the body cells should be able to respond to so peculiar and specific a poison as that of the diphtheria bacillus by the elaboration of a new substance so definitely specific and so definitely conservative. How were we to conceive of what seemed to be new physiological capacities called into being within a few weeks under the influence of agencies brought to bear upon the animal body now, apparently, for the first time in its individual or even in its whole racial history. This seemed teleology in its most striking form, and the cells of the body colony seemed to be fairly raised to the plane of sentient beings, suddenly purposeful and potent for the welfare of the organism. It all seemed romance rather than science. But diphtheria antitoxin is a real thing. We make it, through the good offices of the horse, in adequate quantities; we test its strength; we bottle and label and inject it. While the chemist cannot isolate it nor write its formula, he tells us that it most resem- bles the globulins, and its practical value is estab- lished beyond the need of proof or rehearsal. 5 The effect of antitoxin in the blood is not to kill or destroy bacteria, but to render harmless the corres- ponding toxin. But how is this effect produced? Does the antitoxin destroy the toxin? Apparently not, for after the so-called neutralization of the toxin the latter can in some instances be recovered. It was assumed that the effect is secured by some rapid action of the antitoxin upon the body cells, by which these become speedily invulnerable or are stimulated to produce new substances. But anal- ogous effects can be secured in test-tubes, and here the action of living cells is excluded. Altogether, the facts point toward a chemical union of toxin with antitoxin, whereby the former is deprived of its capacity to injure cells. This form of antitoxic immunity, secured by the conveyance of the blood serum of an animal grad- ually adapted to a toxic or infective agent, and containing the protective substance ready formed, is called passive immunity, in distinction to the insusceptibility secured by an individual for himself through a direct adaptation to the deleterious substances, which is called active immunity. In active immunity the protective stuff is elaborated by the directly immunized individual; in passive immunity it is furnished ready made. The induction of active immunity takes time; passive immunity is secured at once. Active immunity is usually of considerable duration; passive immunity is evanescent. Results similar to those obtained in the study of diphtheria antitoxin were soon realized in the inves- tigations of tetanus. But the practical value of the tetanus, antitoxin as a therapeutic agent was 1ess obvious. Presently it was shown that antitoxic sera may be developed in the body during immuni- zation to snake venom, vegetable poisons, such as ricin, and to other albuminous animal and vegetable materials. Thus, also, through experimental adap- tation, the living body cells may elaborate and set free into the serum substances which suspend the action . . . 6 of ferments like rennet, pancreatin, fibrin ferment, etc. Now, for a time, the possibilities of serum therapy seemed boundless, and the preparation of Sera against all sorts of pathogenic bacteria went on at a breath- 1ess rate. Pneumonia and typhoid, Streptococcus septicaemia and plague, tuberculosis, and cholera– these and many other sera have been faithfully and persistently studied. But one after another they have been found to be, for the most part, of doubtful value and have been set aside. The analogy of other infections with diphtheria and tetanus seemed to be somehow faulty, and the key to the difficulty was wanting. The nature of the immunizing process was still so obscure and the failures, except with diphtheria and tetanus, were so conspicuous, that Virchow, a little jealous always of the bacteria which had for the moment crowded his beloved cells aside, was not without justification in his assertion that with all the talk of serum therapy we had obtained, indeed, in single instances, empirical practical results, com- parable with Jenner's vaccination, but offering no more than did his discovery any theoretical insight into the nature of the processes through which the effects were secured. Just at this period, then, the condition of affairs, as regards artificial immunization in man, was about as follows: There were two practical methods in vogue, vaccination for smallpox, and preventive inoculation in hydrophobia; both elaborated in en- tire ignorance of the inciting microorganism, and both highly successful. There was then the method by antitoxic serum of the greatest practical value in diphtheria, of much less certain usefulness in tetanus. This serum method of so-called passive immuniza- tion was developed as the result of the most pains- taking researches on the bacilli known to be the incitants of these diseases, and leading, step by step, to one of the most noteworthy triumphs of modern SC16ºrn Ce. But so far as man was concerned the protective Sera for a long list of other infectious diseases, which had been widely tested, had been so uniformly unsuccessful as to 1ead to the conjecture that, while there is a form of immunity which can be appro- priately designated as antitoxic, in the greater pro- portion of immunizations other protective agencies must be at work which were still but incompletely accounted for in the doctrine of phagocytosis or in the current conceptions of humoral germicides. A good deal of 10re had been gathered about the germicidal action of blood serum and certain of the phagocytic body cells under various conditions. But the nature and details of the process by which bacteria are destroyed in the body remained obscure. Several years ago, however, Pfeiffer showed that the peritoneal fluid of a guinea-pig, artificially immu- nized against the cholera vibrio, was capable, under certain conditions of not only immobilizing and l:illing in a short time cholera germs placed in the peritoneal cavity of the living animal, but also of causing their disintegration and destruction. This significant capacity of immune serum was, after a long series of experiments, finally found to be due to two distinct substances. One of these appeared to be formed in the body, as the result of the gradual adaptation of the animal to the cholera microbe, and was called the immune substance. The other seemed to be normally present in the serum of the warm- blooded animals, and to be identical with the sub- stance which had long been regarded as in itself germicidal, and which had been called by Bueehner alexin. It was presently found that lysis of the cholera microbe occurred only when these two sub- stances act together, neither of them when separate naving lytic power. If the two substances, the immune substance and the alexin, lytic when together, were heated to 7 o' C., the lytic capacity was lost. But if a small amount of fresh blood serum containing alexin were now added, the lytic power was at once restored. These curious facts, 8 set forth in part by Pfeiffer and further developed by Bordet and Metschnikoff, obviously have an important bearing upon our conception of the proc- esses by which some phases of immunity are secured. But the study of the effects of lytic Sera upon bacteria is one of great technical difficulty, owing to the minute size of the organisms and the conditions under which they can be observed. So the advance of knowledge in this direction halted for a time, and the observations of Pfeiffer did not at Once assume their full significance. - It was evident, however, that two phases of im- munity should be recognized, antitoxic immunity and bacteriolytic immunity; though it was not then, nor is it yet, clear whether these may not frequently or always be associated in the protective processes. The “Side-Chain.” Hypothesis of Ehrlich.-Just here, before passing to some recent researches bear- ing upon bacteriolytic immunity, I should like to recall to your memory the so-called “side chain.” hypothesis of Ehrlich, by which he strove to account for the phenomena of antitoxic immunity as exem- plified with especial clearness in diphtheria and tetanus. Without this hypothesis we are to-day still practically at sea and adrift in our views of the nature of antitoxic immunity. While this remark- able capacity of the body to manufacture off-hand the greatest variety of potent specific antidotes to the most subtle and virulent poisons becomes, for the first time, comprehensible in the light of this ingenious conception. It is now several years since Ehrlich's hypothesis was enunciated, and whatever may be its merit as a direct contribution to science, there can be no doubt that it has been, in a remark- able degree, an inspiration to the most fruitful research. It is practically impossible to follow the recent work on immunity without a knowledge of this hypothesis and familiarity with the nomen- clature to which it has given rise, and I, therefore, 9 ask your attention to a brief review of its more important features. - - In order fully to comprehend the hypothesis of Ehrlich regarding the origin and nature of antitoxin, it is necessary for us to form with him a clear concep- tion of cell assimilation. We conceive of the cell as a mechanism for the storage of energy derived from without and for its release under definite conditions. This storage of energy is possible through the assimilation and tºuilding up by the cell of complex molecular com- binations. These, owing to their instability, are readily resolved into less complex and more stable combinations with the release of the stored-up energy. Thus is the life of the cell manifested. This broad conception of the cell is a purely intel- lectual one, however, for the details of cell metab- olism still elude the keenest scrutiny of the chemist. He cannot formulate protoplasm nor express its chemical changes proportionate symbols. Now, accepting this Gondition of affairs as for the moment inevitable, Ehrlich seeks to express his view of the character of the cell’s performances in general terms, disregarding its morphological pecu- liarities, somewhat as follows: We may conceive of the cell as consisting of a central group of very complex molecular combinations which maintains the char- acteristics and special capacities of the cell as an Organism under all the vicissitudes of its existence. Associated with this central organic group are many and varied subsidiary atom-complexes which loy means of their unsatisfied affinities bring the central group into relationship with food material through those chemical combinations which in living protoplasm characterize assimilation. These unsatisfied affinities by which assimilable material is fixed or united to the cell have been called “side chains,” a term adopted from the chemist. Not to press too closely, however, the analogy between the chemical processes in lifeless substances and assimilation in living matter, these IO affinities or “side chains” of protoplasm are now commonly called receptors. If we seek to illustrate Ehrlich's conception we shall be obliged to use graphic figures of extreme crudity. If the arc of a circle in Figure I rep- resent a portion of the periphery of a cell, we may indicate the side chains or affinities or receptors by projections whose special shape shall indicate their special capacity to combine with any substance coming in contact with them under favorable conditions. Suppose in this figure we let a rep- resent a nutrient molecule which is capable of combining with the receptor b. Through its union with b, and only through this, is it capable of entering into the metabolism of the cell. This mole- Ø CA Boa, FIG. I. cule a cannot unite with the receptor c or d, but only with such receptors—to use the crude expres- Sion which our illustration requires—as it fits. Through the receptors c and d other forms of food ºcules may, enter into the metabolism of the Céll. Thus it is in Ehrlich's conception, which after all is only a graphic way of illustrating the pre- liminary phases of assimilation by living proto- plasm, that the cell is capable of Selecting or “fix- ing” out of the host of various substances with which it comes in contact, just those and Only those to which its receptors bear a definite chemical relationship. II The same thing is true of toxic as of nutrient Substances. In order to be toxic to the cell they must enter into chemical combination with a suit- able receptor of the cell. Then only can they lead to the forms of damage which we are here consider- ing.” As the result of many studies on the nature and effects of toxins, Ehrlich is led to believe that the toxin molecule consists of two forms of affinities; One through which the chemical union with the cell is effected—called the haptophorus group; and the other—called the toxophorous group—by which the damage to the cell is brought about when once the toxin molecule is anchored to it. This conception may be illustrated as in Figure FIG. II : II, in which the toxophorous group a of the toxin molecule can be effective in damaging the cell only when united to the latter by the haptophorous group b. Ehrlich was led to this belief in the com- plex nature of the toxin molecule through the very curious fact that diphtheria toxin may under a variety of conditions lose its toxicity but still retain its capacity of neutralizing antitoxin and also of *It is evident from what has been said about the con- ditions under which substances can be toxic that the immunity of one animal to a given agent which is toxic in another may be simply due to the fact that the cells of the former have no receptors with which this agent can unite, or if this union does take place that the cell is not thereby damaged. This consideration has an important bearing upon our conception of natural im- munity. I 2 uniting with cells, and thus inducing the formation of antitoxin in the animal body. Such toxin molecules deprived of the toxophorous group are called toxoids. Having now with Ehrlich conceived of the living cell as consisting of a central essential complex molecular group, brought into relationship with its food materials by means of a great number of the most varied receptors, through which, under normal conditions, assimilation is secured, 1et us see what may happen if toxins come in contact with living cells which are furnished with receptors capable of uniting with them. - The union of the toxin molecule with the living cell being effected, the cell is more or less damaged. If the damage be sufficient, the cell dies, and that is the end of it so far as our present inquiry is con- cerned. But suppose the damage to be but slight, as may be the case in artificial immunization or in early stages of an infection. The cell is at least deprived of the useful offices of the receptor to which the toxin molecule is now united. This is in itself a loss to the cell, and through the regenerative impulses common to all living cells it proceeds to regenerate the 1ost parts, which in this case are the receptors. But as more receptors are thrown out of function by the continued action of the toxin, the necessity for compensation continues. Now it is a fact long known to pathologists, and especially emphasized by Weigert, that the regen- erative impulse is apt to be in excess of the obvious requirements and leads to overproduction of new cells, tissues, chemical substances, etc. Just this is what now happens to the cells, some of whose receptors have been rendered useless by combination with the toxin. New receptors are formed, more than the cell requires; so numerous may these become that many are at last cast off into the blood. Now here is the point at last. These *eceptors or substances, normal and tuseful to the body, but now formed in excess through Over-compensation, I 3 and set free from the cells into the body fluids, are the antitoxin. For, cast off into the body fluids, these superfluous receptors still retain their combining power for the free toxin molecules, which, also, are floating in the body fluids, and unite with them. This union having been effected, the toxin molecules are no longer a menace to the cell, because the affinities are now satisfied through which they joined the receptors while these were still a part of the cell, and in this way became harmful. This now inert com- bination of toxin molecules and detached receptors is physiologically indifferent stuff, and may be got Cell B ody out of the body through the usual processes of ex- cretion. But the antitoxin which has not united with toxin in the body of the animal which produced it is still available on the transferrence of the serum to another individual whose blood contains diphtheria toxin, and who may thus secure passive immunity. If we have recourse again to our graphic efforts, the hyperproduction of receptors by the damaged cell, their separation, and their action as antitoxin, may be indicated, as in Figure III. The toxic mole- I4 cules a, uniting with the receptors b, 1ead through the injury to the cell, as well as by its deprivation of the normal use of b, to the production, and at length to the overproduction of new receptors of the same kind. These superfluous receptors c are now cast off into the body fluids,” where, as at x, they may freely unite with the floating toxin molecules, forming harmless compounds and preventing further access to the cell, where alone the damage can be done. Or when free, as at y, the receptors may be transferred in the serum, becoming effective as anti- toxin in another individual. It is well to note that so long as the receptor maintains its connection with the cell it is not anti- toxin, but an element of vulnerability to the cell. It is only when the receptor has been set free from the cell that it is antitoxin. The antitoxic substance which neutralizes the action of the toxin molecule of diphtheria is not effective for the poison of tetanus, for example, simply because it does not combine with the mole- cule of tetanus toxin. It is specific for diphtheria, because it was the diphtheria toxin which excited its overproduction through a chemical union iden- tical in character, whether this union takes place while the receptor is a part of the cell, in which case the toxin becomes harmful, or when the receptor is detached from the cell, in which case the combina- tion is harmless. In the light of this hypothesis the incredible becomes the inevitable. The specific character of the antitoxic substances, so mysterious as to seem at first fairly to transcend physiology, is, in fact, but the result of adaptation to unusual conditions of cell capacities evolved and fostered for the every-day maintenance of life. The specific relationship between the toxin and the antitoxin is not developed *These receptors of various kinds, cast off into the body fluids under the most diverse conditions, Ehrlich has called “haptines.” T 5 during immunization, but existed beforehand as a necessary condition of toxic action. We can account by this hypothesis not only for the formation, protective action, and specificity of antitoxin, but we see, also, a special significance in the incubation period, during which the conserva- tive forces are mustering. We can realize, further- more, in the light of this hypothesis, how it is that the protection secured in active immunization is 1ess immediate and also why it is more prolonged than in the passive, since in the latter the available antitoxin is limited to the dosage and is not replen- ished as in active immunity by the continued cell activities of the affected individual himself. The conception of Ehrlich as to the nature of antitoxin is that of the chemist, and carries over to the performances of protoplasm the presumptions upon which chemical reactions in general are con- ceived and formulated. But this is not an easy matter, since our knowledge of the ultimate phases of protoplasmic metabolism is very incomplete. The physiological chemist presents to us as his final achievement in analysis an extremely elaborate complex which he calls the proteid molecule. This he does not yet venture to formulate. Its obvious importance and versatility he can indicate, but can- not yet follow with precision. Thus it is that when we attempt to illustrate ingraphic fashion our conception of the performances of proteid molecules, either in normal metabolism or in poisoning, we are forced to use the crudest of symbols and to indicate the affinities or combining predilections of our molecular groups by forms which are more suggestive of the carpenter or of the 10cksmith than of the subtle forces which dominate this under-world of 11fe. The use of such pictures is not without hazard, for these toxins and these receptors are in truth not histological structures, but molecular groups; they are not alone upon the surfaces but through all the mass of the protoplasm. And if one cannot at 1ast translate these uncouth symbols into the nice con- I 6 ceptions of the chemist they will prove but stum- bling blocks and may lead to his speedy undoing. There has been much discussion of Ehrlich's hypothesis, and it has withstood many assaults, mostly inspired by misconceptions of the funda- mental claims. The scope of this paper does not permit me to touch upon the many and most ingenious experiments by which this view of anti- toxic immunity has been sustained, nor is it practicable now to call attention to many of the phenomena not yet accounted for or seemingly inconsistent with the interpretations here set forth. Having thus glanced at antitoxic immunity, let us now turn to a new line of research bearing upon bacteriolytic immunity which has already led to most significant results and opened biological fields of great scope and complexity. Cytolysis.--It has been known for some time that the blood serum of one animal species, when injected into the vessels of another may do serious damage, and even kill the latter through a rapid separation of the haemoglobin from the red blood-corpuscles. This dangerous effect brought to a speedy end attempts which were at one time made to sustain the ebbing forces of life by the transfusion of alien blood. But the significance of this so-called “laking” of the blood by mixture with alien sera Was Overlooked and passed into the textbooks on physiology as another of the many curious and unexplained phenomena relating to red blood cells. Bordet, however, was led to inquire whether if the animal body be capable of adapting itself to toxic substances and to bacteria in such a way as to neutralize the effects of toxins and to destroy bacteria, as had been shown by earlier experiments, it may not respond similarly to the introduction of other foreign substances, such as alien red blood cells, for example. The blood serum of the guinea-pig is not normally lytic for the red blood cells of the rabbit; that is, it does not cause the separation of the haemoglobin I 7 from the stromata, with a partial destruction of the latter. Now Bordet injected a few cubic centi- meters of the whipped blood of the rabbit, contain- ing the serum and red blood cells, into the subcu- taneous tissue or peritoneal cavity of normal guinea- pigs. This operation, which does not markedly interfere with the well-being of guinea-pigs, was repeated five or six times with intervals of a few days. When now blood was drawn from the treated pig, allowed to clot, and the clear serum secured, it was found to have become markedly lytic for rabbit corpuscles. A very small proportion, mixed with rabbit's blood diluted with physiological salt solu- tion, in a short time brought the haemoglobin into a clear ruby solution in which the stromata or “ghosts” of the corpuscles floated as a pale and scarcely visible cloud. This process is called hamolysis: Serum possessing this capacity is called hamolytic or haemotoxic serum. This adaptation of one animal to the red blood cells of another species may be accomplished without difficulty with a great variety of animals. But a most remarkable thing about this newly acquired lytic capacity of the serum is that it is limited to the red corpuscles of the species of animal whose blood was used for the injection—in Bordet's experiment to the corpuscles of the rabbit. Red blood cells of the dog, cat, sheep, bovines, fowl, etc., are no more affected by this serum of a guinea- pig which has been adapted to the blood of the rabbit than they were before. In other words, the adaptation to foreign corpuscles is specific. I have said that this adaptation to alien blood is specific; that is to say, the serum becomes active only for the corpuscles of the species injected. This statement should, however, be so qualified as to recognize the curious fact that a slight degree of lysis may often be induced in corpuscles of species of animals very closely related to those from which the injected blood is derived. For example, if a rabbit be adapted to human blood by intraperi- toneal injections, the serum of this rabbit, now 18 strongly lytic for the corpuscles of man, may be slightly lytic for the corpuscles of monkeys. Simi- larly, serum artificially lytic for the red cells of goats may be slightly lytic for those of sheep, but not for the corpuscles of cats, dogs, man, etc. This form of test, delicate beyond anything hitherto known in physiological chemistry, may thus prove of value in defining the relationships and limitations of animal species. I show to you here, in illustration of this remark- able experiment, a series of slender test tubes placed against a white background. In this instance the material which I show was secured by adapting the rabbit through successive intraperitoneal injections to beef blood. In the first tube is a little simple salt solution of a strength of o.85 per cent., which is approximately isotonic for red blood cells. To this salt solution I have added a small quantity of beef red blood cells. You see that the corpuscles have settled to the bottom leaving the perfectly colorless salt solution above. In the second tube is the same quantity of beef corpuscles which have been mixed with normal rabbit serum. Here also the corpuscles have settled to the bottom, leaving the clear, color less serum above. You see from these two control experiments that neither physiological salt solution nor normal rabbit serum are lytic for beef red blood cells. The third tube is prepared in the same way as number two, except that here the fluid is serum from a rabbit which has been adapted by successive intraperitoneal injections to the beef corpuscles. Here the corpuscles have disappeared save for a faint cloud at the bottom composed of the colorless “ghosts” of the cells, while the fluid has the fine ruby color of dissolved haemoglobin. This adapted rabbit serum is therefore haemolytic for the beef corpuscles. In order to show to you the specific character of this adaptation I have here two tubes containing the serum of the beef-rabbit, to one of which I have I 9 added corpuscles of the dog, to the other corpuscles of man. In these there is no solution of the haemo- globin; the corpuscles lie in a red mass at the bottom; the serum above is colorless. This preliminary observation of Bordet was immediately followed by a series of studies upon artificial haemolysis, the results of which I shall summarize in the simplest fashion. In the first place, to what is this remarkable acquired lytic capacity of the serum due? Bordet heated for half an hour to 56°C. some of the lytic serum secured by adapting the guinea-pig through subcutaneous injections to the red blood cells of the rabbit. He found that it had completely lost its new lytic power. But when he now added to this inert serum a little fresh blood serum from a normal guinea-pig, which is not in itself lytic, the original dissolving power of the heated serum for rabbit corpuscles was at once restored. The inference from this experiment is obvious. The dissolving capacity of this artificially lytic serum is due to two distinct substances. One of these, that one which results from the adaptation of the animal to the alien blood, is stable at 56° C.; the other, which is present in normal serum, is rendered inert at 56° C.; that is, it is very 1abile. These two substances were named early and have been often renamed. For the present I shall speak of the stable substance resulting from the adaptation to the alien blood as the “immune substance” or “immune body,” and of the other, more sensitive to heat and present in normal serum, as “alexin,” a name which was long ago applied by Buchner to a substance or substances in normal serum, to which its germicidal capacity, first clearly demonstrated by Nuttall, was attributed. Here is a series of tubes illustrating these points. In this first tube is a little normal rabbit serum which has been mixed with beef red blood cells. The cells have settled to the bottom, 1eaving the colorless serum above. In the second tube is the scrum of a rabbit which has received a series of 2O intraperitoneal injections of beef blood to which I have added the same quantity of beef red blood cells as in tube 1. But these cells have disappeared and the serum is red with dissolved haemoglobin. This adapted serum is thus lytic. In tube 3 are the same substances as in 2, except that the beef- rabbit serum, seen to be strongly lytic in No. 2, has been heated for one-half hour to 56° C. Here you see there is no lysis of the corpuscles. They lie unchanged at the bottom of the tube. The fluid above is colorless. In tube 4 I have the same substances as in tube 3; namely, beef corpuscles and heated beef-rabbit serum (not in itself lytic) but I have added a very little fresh normal rabbit serum which, as you may see from the control tube I, is not in itself lytic. It does, however, contain alexin, and you see that it has “reactivated” the heated serum of tube 4, so that complete haemolysis has occurred. There now followed a series of important studies by Ehrlich and his associates which throw still further light upon these curious lytic agents. We have seen that in order to secure the immune substance free from the alexin one has only to heat the lytic serum to 56°C., for half an hour, when the alexin is destroyed. If one wishes to secure the alexin apart from the immune substance he makes use of a very curious property of the latter; namely, its capacity of uniting with the cellular element under whose influence it was elaborated. For example, if one places a small portion of the serum of a rabbit which has been adapted to beef blood in contact with beef corpuscles at a low tempera- ture” for a few hours, he will find that the immune substance has formed so staple a combination with the corpuscles that on their separation from the fluid by centrifugation, none of the immune sub- *It is necessary to reduce the temperature in this ex- periment, because at the ordinary temperature the in- tºy of the corpuscles would be destroyed through yS1S. 2 I stance, but all of the alexin, will be left in the fluid. That the corpuscles under these conditions actually contain the immune substance is readily shown by adding to them a little normal serum— containing alexin but no immune substance— whereupon the lysis will at once take place, as shown by the red color of the fluid. This union of the immune substance with corpuscles is strictly specific, occurring only with the corpuscles of the animal species used in the adaptation. Many other points of extreme interest and sig- nificance have been revealed in these studies on artificial haemolysis which I have not time to touch upon. The question of a multiplicity of immune sub- stances and of alexins has been brought forward, and it seems probable, especially from the researches of Ehrlich, that in adaptation of each animal species to a single form of cell several immune bodies may be developed. It is not unlikely, furthermore, that more than one alexin may be normally present in the blood serum of each animal species, and that a single immune body may be capable of uniting with several forms of alexin. It has been found that the haemolytic capacity of the normal blood serum, which in many animals, as we have seen, is very marked for the corpuscles of alien blood, is also due to two substances which in character and action are similar to those which have been so carefully studied in the lytic sera of artifi- cially adapted animals. It is evident that this artificial haemolysis, secured by the adaptation of one species of animal to the red blood cells of another, is quite analogous to the proc- ess by which immunity is secured against patho- genic bacteria—cholera, for example—and which is called bacteriolysis. Both are specific examples of the general process called cytolysis, meaning cell destruction. But this reaction of haemolysis is not only one of extraordinary delicacy, but is easily 22 observed under conditions quite within our control, and permitting such elaborations and variations as involve great technical difficulties when we are directly engaged with the phenomena of bacteriolysis. Thus these studies of haemolysis have a practical significance in their bearing upon our conceptions of bacteriolytic immunity quite apart from the inter- esting general biological field into which they have 1ed the way. The development of cytolytic capacities in the blood serum of the living animal as the result of adaptation to bacteria and to alien red blood cells being known, it was natural to extend the method to other cells. Thus it has been found that in the adaptation of one animal to the spermatozoa of another species by intraperitoneal injections, a serum is obtained which quickly brings to an end the movements of fresh spermatozoa of the species used—spermotoxic Se111111. - - Similarly, specific leucolytic sera have been pro- cured by intraperitoneal injections of emulsions of 1ymph nodes, spleen and bone marrow. Such leucolytic sera may not only dissolve the leucocytes outside of the body, but they are extremely toxic when introduced into the living species from which the tissues originated. The effects of these leuco- lytic—leucotoxic—sera in the body are most pro- nounced in the blood-forming organs. Here, as the studies of Flexner show, a very significant impulse to new cell formation may be associated with the action of the leucotoxic sera. Emulsions of liver cells of one species injected into the peritoneum of another have led to the develop- ment of hepatolytic serum; that is, serum which on injection into the body of an animal of the same species as that furnishing the liver cells induces note- worthy degenerative changes in the liver. Thus, also, nephrolytic, pancreolytic, thyreolytic, neurolytic, and other analogous cytolytic substances have been developed. 23 So far as they have been studied, the nature of the active agents in these various cytolytic sera and their mode of action is analogous with those in haemolytic sera. Here, as there, the action is due to two groups of substances: one, the “immune body,” stable and increased by the adaptive process; the other, the alexin, occurring normally in the body, not increased in adaptation, and readily destroyed or rendered inactive by heat. All of these last-mentioned forms of cytolytic sera require more extended study before far-reaching conclusions should be drawn from them. But it is now evident that the different functional types of cells in one animal are capable in the adaptation to the economy of another of inciting more or less definitely specific responses, as shown by the various types of cytolytic sera. When one musters all the possible combinations in this form of adaptation and considers the prob- ability that multiple immune bodies may develop in each instance, and that these furthermore may correspond to multiple alexins, the complexity of artificial cytolysis becomes sufficiently evident. But now still another phase of this matter demands a word. These cytolytic or, as some prefer to call them, cytotoxic sera, when introduced into the living bodies of the species from which the cells inciting their formation are derived, act as toxins to which the organism responds, each after its kind, by the development of antitoxic substances. These are called anticytolysins or anticytotoxins. Let me illustrate this interesting point. The blood serum of the normal guinea-pig has, as we have seen, no lytic action on the red blood cells of the rabbit, but after the adaptation of the guinea- pig to the blood of the rabbit by repeated intra- peritoneal injections, the guinea-pig serum is strongly lytic for the rabbit corpuscles in test tubes outside the body. But this lytic serum is not less toxic when introduced into the body of the rabbit. Under these corditions the rabbit produces an anti- 24 toxin, an antihaemolytic substance which is in solution in his serum. If now a little of this anti- haemolytic serum be mixed with some of the lytic serum from the adapted guinea-pig, it will be found, on the addition of rabbit corpuscles, that the lytic serum has lost its power, just as diphtheria toxin loses its harmful properties on mixture with diphtheria antitoxin. Thus may be formed a great variety of specific “anti-bodies”—anticytolysins—from sera which are normally lytic or have become so through experi- mental adaptations. In view of the remarkable results of the adaptation of the body to alien cells from different animal species which we have reviewed, it was natural to ask how an animal would respond to the introduc- tion into the recesses of his body of cells—red blood- corpuscles, for example—from another individual of the same species. It was found, in fact, that under these circumstances lytic substances are sometimes, though not uniformly, developed. The possibility of the formation of isolytic substances was thus established. But if this be possible, why may not autolytic substances be formed by the adaptation of an animal to his own cells experimentally displaced. Such substances have, however, not been found under the experimental conditions thus far observed.* Still it is well known that cells and tissues worn out from use, or dead as the result of injury, inflam- matory exudates, etc., are constantly removed from the living body by processes apparently analogous, if not identical with those which can be experimentally evoked. So that autolysis in some form seems to be an important factor in the main- tenance of the integrity of the body. Just what the agencies are under which normal living tissues cells *Some observations are recorded in which, after profuse internal, hemorrhage, haematuria has developed, indi- cating, the possibility of autohaemolysis under special conditions. 25 are protected from the action of autocytolytic sub- stances is not yet clear. But the multiplicity of known “anti-bodies” justifies the conjecture that such substances—anticytolytic—may be constantly formed and act as safeguards to living and useful cells. If now we turn to the various hypotheses which have been advanced to account for the formation and action of these cytolytic substances, we find that an elaboration of Ehrlich's views as applied to antitoxin is here a source of great illumination. It is evident at once, however, that the matter is not as simple as in the case of antitoxin because we have here two substances at work, the immune body and the alexin. Neither the immune body nor the alexin alone can induce cytolysis. They must act together. The phenomena are, in the main, accounted for if we assume that it is the alexin which when the necessary conditions are fulfilled exerts the destruc- tive action upon the bacterial or animal cell. But the alexin cannot enter under ordinary conditions into direct chemical combination with the cell receptors. The union is effected only by the intervention of the substance which is increased in amount in the process of adaptation; namely, the immune body. A long series of experiments has led to the belief that the immune body has two free atom complexes which enable it to form chemical unions. Through one of these atom complexes it unites with the cell to be destroyed; through the other it is joined to the alexin. Then and not until then is the alexin so linked to the cell that its toxic or destructive action upon the cell occurs. This conception may be illustrated as in the case of antitoxin by crude figures. Here it should be remembered we are illustrating not the production of the cytolytic substances, which we shall speak of later, but the action of them upon the cells to be destroyed. 26 Let a-—Fig. IV.-A—be the cell which is to be destroyed with one of its receptors indicated at d. Let b represent the immune body with one atom complex e capable of uniting with the cell receptor d, and with another f capable of uniting with the alexin c through g. Now the aléxin which appears to be the effective agent in the destruction, cannot unite directly with the cell receptor. When, how- ever, it becomes linked to the cell by means of the immune body b, its destructive capacity can come into play. * -. º: - **:A *:: y - * _*** º: * sº § &: & §§ w §4. In similar fashion one may indicate the action of anticytolytic substances which may be effective through union either with the alexin or with the ºnmune, body, as shown in Fig. IV, B and C. In B the antibody” h prevents the linking of the alexin c to the immune body b by itself uniting with the former. It then acts as an antialexin. In C the “antibody” prevents the inking of the *ine body b to the cell receptor d, and hence acts as an anti-immune body. We shall see in a 27 moment why at present the substance here spoken of as antialexin is usually called the anticomplement. The experimental evidence that the anticytolytic substances may be thus due to the formation of adaptive substances of two classes, anti-immune Substances and anticomplements, cannot be entered upon here. In view of the rationale of cytolysis, as just set forth, we may consider the immune substance to be an intermediary between the alexin and the cell to be destroyed; or, on the other hand, we may consider the alexin as the complement to the immune substance, since only through their union is the toxic action possible. In fact, following Ehrlich, one sometimes speaks of the immune body as the “intermediary body,” or “infermediary substance,” but since it is furnished with two combining affinities, it is now usually called the “amboceptor.” Furthermore, since the experimental analysis of the lytic process by the new technique has shown that the germicidal and destructive action of blood serum, formerly sup- posed to be due to a single substance called alexin, is really due to the combined action of two substances, the use of the word alexin for one of them is misleading and has now been largely given up. The substance present in the serum of both normal and adapted animals through which lysis is effected when it is linked to the cell by the inter- mediary body is called the complement. Other names have been applied to these hypothetical complexes or substances which we need not now stop to consider.” It was natural in the early studies on bacterio *These strange and multitudinous names are stumbling- blocks to those who have not followed step by step the development of these new lines of research. But most Of them mark temporary phases in the development of the new conceptions and are no longer current. One should remember that immune bodies or immune sub- stances, amboceptors, sensitizing substances (substances sensibilisatrices), or fixatives (fixateurs), and intermediary 28 1ysis which were incidental to researches on im- munity, that the new substance which was found in the serum as the result of the immunizing process should be called the immune substance or immune body. It was natural also, although less appro- priate, to apply the same term, immune substances, to the analogous substances which appeared in the serum as the result of the injection in the same fashion of cells and other materials which were not infectious, not disease producing, and against which, therefore, the body is not, in the old sense, immunized. But these new uses of the word are, I think, un- fortunate because the word immunity has come to have a special and useful significance in relation to infection, intoxication, and other conditions of natural or acquired tolerance to obviously and seriously harmful agents. The process in both instances is, indeed, one of adaptation, and the newly- acquired capacities of the serum are due to sub- stances resulting from this adaptation. They arise from a functional modification of parts of the body, and hence may be appropriately called adaptive substances. It seems to the writer that it would be better to consider immunization as a special phase of adaptation, and so limit the application of the word that it shall still connote infection and intoxication in the traditional sense. If we continue to speak of immunization to the countless possible adaptations of the body to such innocuous substances as milk, pepsin, white of egg, etc., we may be forced at 1ast to regard intestinal digestion as immunization, and thus miss a distinction which is not without value in the expression of our Conceptions of immunity to the infectious diseases. bodies Or intermediary Substances, all mean the same thing; and that alexin in its modern sense, cystase, addi. inent, and complement are synonymous; and, further, that for the expression of Éhrlich's views, the words tºrtºnume bodies or amboceptors and complements suffice for practical purposes. (See Table showing synonyms page 39.) g synonyms, 29 This, then, is the rationale in accordance with Ehrlich's hypothesis of the action of these cytolytic or cytotoxic substances, either existing naturally, as they do in some animals, or being called forth in 1arger quantities in the process of adaptation to the cells which they destroy. This view has been most fully tested upon haemolytic sera, since here the reaction is most easily studied. But so far as One can see it applies as well to the phenomena of bacteriolysis, whose direct study is much more difficult. It should, however, be borne in mind that the erythrocytes are very delicate and very pecu- liarly constituted cells, and it is possible that inferences drawn from haemolysis are not applicable without qualification to other and less vulnerable cell types. The origin of the amboceptors of these cytotoxic substances is accounted for in the same way as in the case of antitoxin. The cells or substances which are introduced into the animal, and to which it proceeds to adapt itself, 1ead, through union with such body cell receptors as may be fitted to them, to the overproduction of these special complex receptors. These are presently cast off as super- fluous to the body cell producing them, and are then free as amboceptors in the body fluids. As in the case of antitoxin formation, it is prob- able that the cell receptors which are thus increased are normally concerned in cell assimilation, and it is not unlikely that their complex character may have some relationship with the complexities of the great albuminous molecules, which must suffer initial changes before becoming fit for assimilation. t any rate this hypothesis assumes that in the process of adaptation either to toxic substances or to foreign cells or other albuminous material, the body develops no new capacities, but only an exag- geration of those already existing. As to the exact source of the intermediary sub- stances we cannot yet speak with certainty. Phagocytosis.--It was inevitable that these re- 3O markable studies on cytolytic Sera should lead to considerable modification of the doctrine of phagocy- tosis; just as inevitable as that the recognition of the important offices of the cells in the formation of these lytic sera should lead to noteworthy conces- sions from the camps of the humoralists. I regret that the scope of this paper does not permit me adequately to review the later conceptions of Metschnikoff and his associates of the action of phagocytes in the adaptive processes by which immunity is secured. Metschnikoff fully recognizes the importance of the adaptive substances, some of which may be largely increased in amount in the processes of immunization. More strenuously than other Ob- servers, however, he insists upon the phagocytic cells, especially the leucocytes, as the originators of the substances concerned in cytolysis, and holds that under ordinary conditions it is only within these cells that these substances are effective. In artificially immunized animals, however, the interme- diary substances, it is conceded by Metschnikoff, may be set free from the cells which produce them and mingle with the body fluids. The complement, On the other hand, which he, in common with others of the French school, calls cytase, in recognition of its ferment-like characters, Metschnikoff does not believe to be set free in the body fluids except through some damage to the leucocytes in which it is formed. Such a damage, for example, as befalls the leucocytes in the clotting of the blood: for in this process it is assumed that the setting free of the fibrin ferment involves the destruction—phago- lysis—of the leucocytes. The views advanced by Bordet and others of the French school regarding the union of the ambo- ceptors with the cells to be destroyed are less pre- cise than those of Ehrlich. Both, however, recog- nize the importance of an association of the ambo- ceptor as a condition for the effective action of the complement (cytase). It is for this reason that the 3 I amboceptor is called by Bordet, Metschnikoff, and Others the sensibilizing substance (substance sensi- bilisatrice) or the fixative (fixateur). Finally, a 10ng and ingenious series of experiments have led Metschnikoff and his associates to believe that there are two forms of cytase, one called macrocytase, formed by the macrocytes (1arge 1ymphocytes derived from the spleen, lymph-nodes and certain endothelial and connective tissue cells) and concerned in the destruction of animal cells, such as red blood cells, leucocytes, spermatozoa, various parenchyma cells, etc.; and microcytase derived from the microcytes (polymorphonuclear 1eucocytes), which is active in the destruction of bacteria. The greatest diversity of view concerning the cytolytic process between Metschnikoff and his followers and the observers of the Ehrlich school relates to the question whether the complement (cytase) does or does not exist free in the blood plasma, for upon the answer to this question depends largely our belief as to the relative signifi- cance of intra- and extracellular cytolysis. This is one of the points concerning which more data is urgently needed. But even now the views of Metschnikoff are not inconsistent with the hypoth- esis of Ehrlich. It is, in fact, most entertaining to observe the mutual felicitations of the exponents of the phago- cytic and the humoral doctrines, each group being congratulated upon finding at last in the cherished fetich of the other the long sought clue to the solution of their common problem. Agglutinative Substances.—But we have not finished with the tale of these curious adaptive resources of the living body when called upon to deal with foreign material of special character introduced in unusual ways into its recesses. The phenomenon of agglutination has been 1, 3 :-->>. 4--. tWO { Intermediary body, toplasmic origin (Cytotoxins), Sºlºins * substances. Substance may elaborate— p :--> Sensibilisatrice, Nephrotoxin, Fixateur Hepatotoxin, etc. l Präperateur, Agglutinins, Copula, Precipitins. \. Desmon. B - Anticomplements. Cytotoxins, may lead to the Anticytoxins Antiamboceptors The “antibodies” Agglutinins, formation of “anti- (anti-immune bodies). Precipitins, antibodies.” Antiagglutinins. Antiprecipitins. tionship with microorganisms or their toxic products in infection or immunization are capable of adapting themselves to the new conditions by the elabora- tion of protective substances—bacteriotoxins—espe- cially potent to harm the invading organisms, may it not also be true that these invading organisms in their turn—being even as the body cells are, living structures capable of adaptation to a harmful environment—have the power of self-protection through the development of substances—cyto- toxins—of exalted potency: substances which are protective to the microorganisms through the damage which they inflict upon the living body cells? Such substances, finding in the body their necessary complements, may not only neutralize and render futile the protective substances which the body cells elaborate by direct union with them, but may also induce toxic effects of the severity of which our studies on the toxins of artificial cultures give us no adequate conception. This ingenious hypoth- esis opens a new field for research which may hold the key to many subtle and obscure problems of infection. If I have succeeded in this hurried survey of these newly opened fields of research, it is clear that we are face to face with problems of the highest theo- retical interest and importance, and that in the achievements of the 1ast decade lies abundant in- spiration for the immediate future. But I fancy that I shall not have fully absolved myself from the exigencies of the practitioner's outlook unless I shall point to some of the direstions in which this new 1ore may become available in the ministrations of the physician. In this I must perforce be brief, for in truth the outlook is chiefly one of promise, and much patient research will be required before the bearings and relationships of these new concep- tions shall have been co-ordinated and rendered available in either the prevention or the treatment of disease. But two or three points I shall venture to touch upon. 4O We are first led back to the great group of infec- tious diseases in which, as it now appears, the protec- tive agencies in both natural and artificial immunity are not chiefly antitoxic but bacteriolytic. In the light of the new knowledge of cytolytic sera, and the conditions under which these may be effective, the promise of serum therapy, so long limited to antitoxic immunity, seems now to be definite and encouraging. Let me develop one of the new possibilities. - We have seen in our review of haemolysis that haemolytic serum heated to 56° C. loses its lytic power owing to the destruction of the very labile complements. We have seen further that this power is restored by the addition of a little fresh serum from a normal animal; that is, a serum-con- taining complement. Now it has been found that this “reactivation” of the serum, as it is called, can often be brought about by the sera of various animals. Thus, for example, the serum of the guinea-pig adapted to the erythrocytes of the rabbit is lytic for these cells of the rabbit. If such sera be heated to 56°C. it is no longer lytic, the activities of the complement are destroyed; but the serum can be reactivated by a little fresh serum not only from a normal rabbit, but from the goat and the rat. The serum of many other animals, however, is inef- fective under these conditions. The reason for this, of course, in accordance with Ehrlich’s hypothesis, is that the complements of the reactivating sera have combining capacity with the special ambo- ceptors, and so can become effective, while in other sera, the 1inking of the complement to the red cells through the amboceptors being impossible, there can be no restoration of the lytic action. It is not difficult to secure immune substances (amboceptors) by the adaptation of animals to various kinds of pathogenic bacteria. These may be formed in such abundance as to be out of pro- portion to the complements. But unless these immune substances when injected into the body 4. I for protective purposes either carry with them or find in the new environment an abundance and appropriate forms of complements, they are not wholly available in destroying bacteria. One of the great problems of the immediate future, then, so far as serum therapy is concerned, seems to be to secure suitable complements to act with immune substances if the former do not exist in the human fluids, or to reinforce these substances from the sera of suitable animals if the human stock be Scanty. - . The search then for bacteriolytic sera takes on a wider range, and it is not unlikely, in the opinion of Ehrlich, that the nice adjustment of immune sub- stances and complements which must characterize the most effective bacteriolytic sera will be secured rather by a combination of sera from various animals than from a single animal. There is, however, much ground for believing that in order to be most effective the complements with which we may seek to reinforce the potency of bacte- riolytic sera should come from species closely allied to our own. - . It is possible, therefore, that our blood relation, the monkey, may win at last, through his contribu- tions to the pharmacopoeia, some measure of esteem which, by reason of the scruples of the church and his objectionable personality, has been so long and so unjustly withheld. If the securing of an appropriate complement is thus of such importance in the attempt to prepare bacteriolytic sera for therapeutic purposes, the maintenance of sufficient complements in the human body must be of the utmost significance in its in- trinsic protective mechanism against infection. That this consideration is not without support in fact is shown by the studies of Abbott, Longcope, and others, who have found that after the continuous adminis- tration of alcohol and in various chronic as well as acute diseases, the amount of complement in the blood may be notably reduced. We have thus a 42 distinct and definite contribution to our knowledge of one of those factors in predisposition to infection which, in a general way, are so fully recognized, but which are, for the most part, but ill-defined and little understood. One other point deserves brief mention here. If it be possible so to adapt an animal to a particular form of cells that its serum shall become specifically lytic for these cells, may we not have here a possible outlook for the preparation of sera which will be effective in limiting the growth if not of 10cally destroying malignant or other tumors? But few experiments have been done with this end in view, and these appear to afford little encouragement. It is possible that with a fuller knowledge of cytolysis the outlook will be more promising. But a great amount of preliminary research must be carried out before the problem can be confidently approached. The hypothesis of Ehrlich, which so closely co- relates the action of toxins with the assimilation of nutrient stuff, has led to new conceptions of the details of the relationship of foods transformed by the preliminary digestive process to the material which is finally placed at the disposal of the cells. It seems not unlikely that through the action of the cell receptors the food material which arrives in the body fluids may not only be adapted to the specific uses of the cells, but that by the formation of count- less varieties of substances analogous to the so-called “antibodies” of immunization, the cells are pro- tected against equally various toxic substances. If this be true, the hope seems justified that following the lines of research suggested by this new technique we may be able ultimately to understand more clearly the details of the so-called internal secretion and those disturbances of chemical adjustment which give rise to many important phases of autointoxication. To sum up the whole matter, then, we have reason to believe from our present knowledge that immu- nity is either antitoxic or bacteriolytic, and that the two protective processes may often act together. 43 It appears probable that bacteriolytic immunity is dependent upon the mutual action of two groups of substances, one amboceptors or immune sub- stances which may be specifically increased by artificial adaptation; the other—complements—ex- isting already formed in the normal animal, either free or within certain types of cells. There seems to be abundant ground for the belief that the protective agencies which are evoked in both natural and artificial immunization are simply those which the body makes use of in its normal metabolism, exaggerated and diverted to different ends, it is true, in the face of emergencies, but giving evidence of the birth of no new physio- 1ogical capacities. The new methods of research and the far-reach- ing conceptions which they have stimulated and fostered seem likely to mark a new era in physio- 1ogical chemistry, and to 1ínk more closely than any other extension of knowledge in our time some of the most subtle and urgent problems of medicine to the wider outlooks of general biology.* *BIBLIOGRAPHY. The records of these recent researches are widely scattered through the German, French, and English technical periodicals. - The summary of Aschoff (Ehrlich's Seitenketten- theorie und ihre Anwendung auf die künstliche Immuni- zierungsprozesse. Zeitschrift für Allgemeine Physiologie. Bd. I. Heft 3; I 902) is most complete and contains a full bibliography. The monograph of v. Dungern (Die Antikörper, 1903) contains much valuable material. In English, the recent Huxley Lecture by Welch (Recent Studies of Immunity. Reprinted in the Medical News, October 18, 1902; and in Science, November 21 and 28, 1902) is admirable, and deals with especial fulness with toxins and their relationship to various important patho- logical processes. - Ritchie's discussion of the subject (The journal of Hygiene. Vol. II, Nos. 2, 3, and 4, 1902) treats in a clear and philosophical fashion the facts and hypotheses involved, and contains many valuable suggestions for further research. * A CASE OF MULTIPLE FIBROMA (FIBRO—NEUROMA) OF THE NERVES OF THE LOWER EXTREMITIES, WITH DIF- FUSE ENLARGEMENT OF THE SCIATICS – COMPLICAT- ING SARCOMA AND METASTASES IN THE LUNGS.* John H. LARKIN, M.D. (Tutor in Pathology, College of Physicians and Surgeons, Columbia University, Mew York.) I. INTRODUCTION. Since the report of two cases of multiple neuroma by Cheseldon in 1740 and Tilesius in 1798, many studies of multiple fibromata of the nerves have been made. The most recent summary by Plebe and Hektoen contains an analysis of such cases, and to this we may refer. - Altogether when the reports are critically reviewed some seventy cases of these so-called multiple neuroma are on record. The purpose of this paper is to record a new case in-which the primary lesion is complicated by the occurrence of sar- coma. This occurrence of sarcoma in connection with mul- tiple fibroma of the nerves has been recorded by Reckling- hausen, Westphalin, and Garrè, and such cases form a group of considerable scientific interest and practical importance. Recklinghausen first drew attention to the relation of fibrous skin tumors to those of the nerves and their subsequent ma- lignant transition; since then Garrè, after a review of the published cases, has collected seventeen of sarcoma of the nerves, which he has divided into two groups, viz.: first, a group characterized by multiple skin tumors, tumors of the peripheral nerves, pigmentation of skin, and functional dis- turbances; and second, a group in which the above condi- tions are absent; but in their stead we have enlargement alone of one or more nerves. Both groups may be compli- cated later by the development of a malignant type of growth and the formation of metatases. * Received for publication March 3, 1903. 217 2 I 8 LARKIN. I shall not touch on the clinical history or the surgical as- pects, because they will be made the subject of a contribu- tion by Dr. Kammerer, to whom I am indebted for the case, except to state that the patient was a female, age 24, who entered St. Francis Hospital on May 9, 1901, with a large tumor on the posterior part of the right thigh, and with a similar but smaller tumor on the posterior part of the left thigh. She was operated upon May 23, 1901. Death occurred June 19, 1901. II. AUTOPSY. The autopsy on June 19, four hours post-mortem. Body is well preserved, but greatly emaciated. On the posterior part of the right thigh is a large, soft tumefaction extending nearly to the trochanter. A long incision resulting from the opera- tion leads directly into a large cavity, the base of which is made by the shaft of the femur, and its walls by the very soft, Spongy, and gelatinous muscles of the leg, There are large areas of blood infiltration mixed with gelatinous tumor mass extending up and down the thigh. The muscles of the pos- terior group are absent. The shaft of the femur shows no necrosis and the sciatic nerve cannot be found at this point. The nerves of the lumbo-sacral plexus on this side show marked enlargement from their exit, from the spinal canal to their junction to form the great sciatic — the diameter of each nerve being I.5 cm. At the sacro-sciatic foramen the diameter of the sciatic nerve is 3.5 cm., while the distance from the foramen to where the nerve ends and is replaced by tumor mass is seventeen centimeters. From the point to the junction of the internal and external popliteal nerves, a dis- tance of twenty centimeters, the great sciatic nerve is absent, its divided ends merging into the surrounding tumor tissue. The internal popliteal nerve is enlarged, its diameter being one and a half centimeters. From here down, the posterior tibial nerve is of about the same diameter till it divides into the internal and external plantar, the diameter of the latter being one centimeter. The anterior tibial nerve is one centimeter in diameter throughout. The anterior crural nerve and its A CASE OF MULTIPLE FIBROMA. 2 IQ branches show no enlargement. The nerves were of firm consistence and presented here and there nodular swellings. For the most part the great sciatic and its branches seemed to be of uniform thickness throughout. The small cutaneous nerves were thickened and presented many nodular swellings along their course, giving them a beaded appearance. This latter condition was especially well marked in the small sciatic and superior gluteal, where the nodular swellings con- nected with these nerves were so abundant as to make a lobulated mass above the tuberosity of the ischium. The dissection of the left sciatic nerve discloses a some- what different condition to that of the right, in that the con- tinuity of this nerve is preserved, although presenting a tumor of much smaller size within the nerve sheath. The nerves of the lumbo-sacral cords at their exit from the spinal canal are 1.5 cm. in diameter; and at their junction to make the great sciatic nerve at the foramen is three centimeters in diameter. From here down, the nerve is of a uniform thick- ness to a point twenty-five centimeters from the sacro-sciatic foramen where it enlarges into a fusiform swelling which is ten centimeters in longitudinal direction six centimeters in diameter, the whole tumor being fourteen centimeters in circumference. The longitudinal section of this fusiform mass discloses a tissue of moderate consistence, irregularly mottled with light yellow and grayish gelatinous areas and encapsulated by the nerve sheath; near the lower part there is commencing cyst formation. The sciatic nerve entering this tumor at its upper pole forms a very delicate plexus at its periphery, when the nerve is again restored at its lower end into its appropriate branches. From this point down the internal popliteal, posterior tibial, plantar, and anterior tibial nerves are enlarged, their diameters being the same as those of the right leg. The small cutaneous nerves are thickened and present many nodular swellings along their course. The brain and spinal cord are normal. On the branches of the cauda equina are a few fusiform swellings. The dissection of the epigastric, sympathetic, cervical, and brachial plexuses show no changes or enlargement of their branches. 22O LARKIN. The heart, liver, kidneys, spleen, and gastro-intestinal tract present no abnormal conditions. In both lungs, however, are numerous tumors, varying in size from a hazelnut to a hen's egg, of grayish color and gelatinous consistence. These are scattered irregularly through the lung paren- chyma. Visceral and parietal pleura are normal. In the diaphragm on the right side are several embedded masses of firm consistence and somewhat sharply circumscribed in the diaphragmatic tissue. In the head of the pancreas are three large nodules of a firm consistence, the rest of the organ being normal. III. MICROSCOPICAL EXAMINATION. Portions of the tumor and thickened nerves were hardened in formalin, ten per cent, Orth's fluid, Flemming's solution, and alcohol, and stained by Weigert-Pal, hematoxylin and eosin, Van Gieson and Weigert's elastic tissue methods. The changes in the heart, liver, kidneys, and spleen were those of parenchymatous degeneration. (a.) Sections of the right sciatic nerve, three centimeters below the foramen, show uniform enlargement and give a picture of diffuse sclerosis. The connective tissue of the perineuron running in between the nerve bundles in narrow and wide bands divides the nerve into smaller and larger compartments. There is marked increase of fibrous tissue of the epineurion, but more so of the endoneurion. The num- ber of nerve fibers is notably diminished and replaced by connective tissue of great density. In some bundles only a few fibers are to be found, while a surprisingly large number of well-preserved fibers could be found in others. Sections stained by Weigert-Pal and Van Gieson methods show well the loss of nerve fibers and increase of connective tissue. The individual nerve fibers, although in general compromised by the new-formed fibrous tissue, are surprisingly free in many sections of the enlarged popliteal and anterior tibial nerves. Sections of the smaller and larger tumors connected with the peripheral nerves present the characteristic structure of loose texture fibroma. The fibrous tumors appear to JOURNAL OF MEDICAL RESEARCH. WoL. IX. PLATE XII. FIG. I. A. – Right sciatic nerve. B. – Left sciatic nerve. Fig. 2. Cross-section of sciatic nerve in neuro-fibroma with sarcoma. LArkin. Fibro-NEUROMA. A CASE OF MULTIPLE FIBROMA. 22 I spring from the nerve sheath which sends in bands of denser connective tissue. These nodes consist of typical fibrillated connective tissue which is extremely tangled and separated by a fluid stroma containing many medium sized degenerated cells. The number of blood vessels varies greatly in many of the nodules. Sections stained by Weigert-Pal method show no nerve tissue in the nerve itself, but generally a few fibers can be seen at the extreme periphery of the sheath. Sections of Large Tumor which had been partially removed from Right Sciatic Merve. — The cells of this tissue are of the polyhedral type and vary in size; in places they are packed together with fibrillated stroma between, while in other areas separated by a fluid stroma staining by hema- toxylin in which are many fine fibrils staining especially well by Flemming's hardened tissue. The blood vessels are numerous and for the most part thin-walled, showing in places cellular proliferation at their periphery. Mitotic figures are abundant and areas of necrosis at the center of the tumor quite large. The tissue has all the characteristics of a typical sarcoma. Sections of the tumor stained by Weigert-Pal show no nerve fibers. (b.) Sections of left sciatic nerve and its branches. As the microscopical picture of the left sciatic nerve beyond the tumor presents an identical appearance to that of the right Sciatic nerve and its branches, which have already been described, it will be unnecessary to repeat. The tumor of this left sciatic nerve, being, however, structurally different, will be here described. * Sections from this tumor are chiefly made up of spindle cells with an abundant intercellular fibrous stroma. In places the stroma is dense with few cells, while in other areas it is very cellular and edematous. The cells run in various direc- tions, but seem to have a more or less close relation to the adventitia of the blood vessels. The nuclei and cells vary considerably in size and shape. Many huge, deeply stained chromatin masses with a proportionate amount of protoplasm 222 LARKIN. are by no means uncommon. The tissue is well supplied with thin-walled blood vessels which seem to be intimately associated with the tumor cells. Sections of Metastatic Tumors in Lungs. – Sections from these nodules present an identical appearance to the large tumor of right sciatic nerve; viz., they are made up of poly- hedral cells, intermingled with fibrillated stroma, and contain blood vessels. It is perfectly evident that histologically these tumors are the direct result of metastatic invasion from the right sciatic nerve tumor. Sections from Tumor Masses in Pancreas and Diaphragm. — Sections from both tumors present an identical appearance. They are surrounded by a fibrous capsule, the tumor ele- ments composed of spindle cells intermingled with stroma and blood vessels. I do not think that these tumors have any connection with the tumors described on the sciatic nerves, although histologically they appear like the tumor of the left sciatic nerve. I am inclined to class them as isolated examples of malignant transitions in the thickened nerves supplying the pancreas and diaphragm, viz., those nerves of the splenic and phrenic plexus. . Anatomical Diagnosis. – Multiple fibromata of both sci- atic nerves and branches — complicating sarcoma — meta- Stases in lungs. The most important features of this case, then, are that it adds another to the list of neuro-fibromatoses of the nerves described by other writers, that it is of especial interest in showing the bilateral condition of tumor growth on the sciatic nerves, and also that it is an admirable example of that class of cases described by Garrè ; viz., that there exists a congenital predisposition for nerves to undergo diffuse sclerosis, to sarcomatous metaplasia, and the formation of InetaStaSeS. THE Journal of MEDICAL RESEARCH, May, 1903, Vol. IX., No. 3. [Reprinted from THE MEDICAL NEws, April 4, 1903.] THE PROGNOSTIC VALUE OF THE DIAZO-REAC- TION IN PULMONARY TUBERCULosis. BY FRANCIS CARTER WOOD, M.D., OF NEW YORK ; INSTRUCTOR IN CLIN ICAL PATHOLOGY: COLLEGE OF PHYSICIANS AND SURGEONS, COLUMBIA UNIVERSITY; PATHOLOGIST To ST. LUKE’s HospitaL. IN one of his early papers on the diazo-reac- tion, published in 1883, Ehrlich speaks of the prognostic value of the reaction both in typhoid and in phthisis. He considered that a marked reac- tion in either disease was an indication of a se- vere infection and that in pulmonary tuberculosis especially, a continuous and strong reaction was a sign of very grave prognostic import. A num- ber of theses and papers by other observers were published about this time, showing that in a large proportion of cases the reaction became more strongly marked as the disease advanced and fre- quently afforded useful prognostic data; but that there were fatal cases which did not show the diazo-reaction before death and that there were mild cases in which, though a strong reaction was present at some time during the course of the disease, yet the patient progressively improved. Since 1884 but little has been published on this subject; but with the present revival of chemical and microscopical diagnosis in internal medicine, the diazo-reaction has again come into promi- nence especially through the writings of Michaelis, Schraeder and Naegelsbach, Becker, Clemens, and others. The more recent writers, On the whole, agree with the original views pub- lished by Ehrlich, but some have been unable to confirm his results. The subject is one of such general interest that it seemed proper to pursue it further and especially on a large series of cases under similar conditions. This opportunity oc- curred in the wards of St. Luke's Hospital, where a large number of cases of progressive phthisis were under observation. A considerable number of these cases came to autopsy and the anatomical findings were of value in controlling the results of the physicial examinations. Owing to the un- favorable conditions under which tuberculous cases are placed when treated in hospitals within city limits, it is the rule at St. Luke's to retain in the hospital for any length of time only such pa- 2 tients as seem beyond the hope of cure by climatic treatment, and to urge all others for whom there is a reasonable hope of recovery to leave the wards and to attempt to find either homes in the country or care in one of the sanatoria for tuber- culosis. In this way the cases are immediately divided into two sharply distinguished groups. In the first are all early, non-progressive cases, which as a rule leave the hospital in a few weeks, Often sooner. In the other group are those cases in which an active and progressive process is un- der way. Many of these cases die in the first few months of hospital treatment; others live for a year or more; but very few leave with any im- provement in the pulmonary condition. Oppor- tunity is thus offered to study a series of severe cases of tuberculosis and to observe the variations in the urinary reactions. The total number of cases examined was 363. In all those patients who remained in the hospital for more than a few days the diagnosis was veri- fied by the finding of tubercle bacilli in the spu- tum. Only in a very few, exceptionally mild cases, were the bacilli not found during the short stay of these patients in the wards. Of these 363 cases, II 7 died. The results of the examination of the urine of the fatal cases showed that in 81 cases, or 69 per cent. of those who died, the diazo-reaction had been continuously positive be- fore death. In 22 cases, or 18 per cent., there had been a positive reaction during a large portion of the time during which the patient had been treated, but this disappeared shortly before death. In other words, about 90 per cent. of the fatal cases show a positive reaction during the last few months of life. Of the I4 remaining cases which at no time showed a reaction, ten had the symp- toms of a chronic diffuse nephritis; and two of the cases undoubtedly died from the kidney le- sion. The kidney lesion seems to interfere with the excretion of the substance producing the diazo-reaction, a point to which Clemens has called attention. Two other cases not giving a reaction died of hemorrhage from the rupture of a large vessel in the lung, though the tuberculous process was not active at the time. The cases which were discharged numbered 246. A certain number of them left the hospital for reasons of discipline or because they did not like the treatment which they received. These numbered 22. Twelve were severe cases and had a continuous positive reaction. Three were posi- 3 tive at some time; seven were negative. Of the remaining 224 cases which left the hospital in good condition, I6 were observed for so short a time that their results should not be included here; some of these cases gave faint positives or a positive on the day after admission and a nega- tive a few days later. Of the 208 cases left after Substracting the above; that is, cases which could be fairly considered as mild non-progressive forms; 188, or 90 per cent. of those examined for over four weeks showed no reaction. The 20 other cases gave alternating reactions; that is, positive one week or month, and negative the next; so that no conclusion could be drawn from their examinations. If we take, then, the cases which may be considered as unfit for hospital treatment and capable of cure, we find that some IO per cent. give occasional diazo-reactions, so that it does not seem possible in this country to apply the rule which is suggested for the German sanatoria by Michaelis and Clemens, to exclude all cases with a positive diazo as unfit for climatic treatment, especially as a good many persons with a very slight lesion give a diazo-reaction on ad- mission, and then, as they improve under the altred hygiene, the good food, and the rest in bed which they obtain in a hospital, never give a reaction during the further period of examina- tion. If we collect all the cases which were admitted to the hospital and examine them to see how many gave a positive reaction at some time dur- ing their residence, we find the number to be 154, or 42 per cent. of the whole. The number would be considerably reduced if we were to exclude those cases which gave a single positive reaction on admission and then a continuous negative. If now we compute the number of fatal cases in the I54 which gave a single positive reaction the number will be found to be 103; that is, 66 per cent. of those who apply for hospital treatment and are found to give a positive reaction will die, and the largest number within six months. These results correspond very well with those reported by others. For example, Clemens found that 87 per cent. of the fatal cases show a posi- tive reaction. Rutim yer found that 85 per cent. of the fatal cases in his series gave a positive re- action; while Michaelis noted that 72 per cent. of the cases giving a positive reaction die within six months. - Of the exact chemical nature of the substance 4 or substances which cause the diazo-reaction, we are as much in the dark as when Ehrlich first published the method nearly twenty years ago. We know only that it is produced by the ac- tions of bacterial toxins and during the course of abdominal metabolism such as goes on in a pa- tient suffering from a gastric carcinoma or from a chronic heart lesion. It may be produced in animals by the injection of the toxins of the tubercle bacilli as they exist in Koch's tuberculin T. R. . It is easily destroyed by the alkaline fer- mentation of theºûrine and disappears from that fluid on long standing or on prolonged boiling, which will distinguish it from the drug reactions especially from that produced by naphthalin or § —naphthol. In general the drug reactions can be recognized by any one familiar with the color of the diazo-reaction, so that there is but little danger of confusion. The foam is never salmon pink, but yellowish or purple. The reaction is not dependent upon temperature, for some of the cases in my series showed a constant and strong positive with a normal or subnormal temperature. The appearance of the diazo-reaction in the urine is not a constant phenomenon. In looking over a series of tests carried out for a long period of weeks or months, it will be noticed that the re- action may be continuous and strong for a month or so and then suddenly disappear for a week or so only to reappear later on in its original inten- sity. During this time the condition of the patient may grow progressively worse, so that it would be expected that the reaction would remain quite constant. The conditions underlying this varia- tion are not completely understood. It was noticed by my assistant, Dr. N. E. Ditman, that the reaction seemed to vary somewhat with the atmospheric conditions; that is, on excessively hot, damp days the reaction in a group of pa- tients was usually more intense than on cool, dry days. This variation is no doubt due to the fact that the condition of the patient was influenced unfavorably by the excessive heat. Perhaps also the question of food and the body metabolism may be concerned. The variation, however, was never very great and any case showing a strong positive reaction, or a negative, continued to give a positive or a negative, the action of the tem- perature not being sufficient to completely sup- press the reaction or to bring it on. An im- portant factor in suppressing the reaction in the urine is due to drugs. As Burghart has shown, 5 Some of the tannic compounds, such as tannalbin, tannigen, or even a decoction of uva ursi which contains tannin, are capable of causing a disap- pearance of the reaction from the urine. Creo- Sote and creosotal also cause a great diminution in the intensity of the reaction or may even sup- press it. Clemens has also shown that the pres– ence of bile, urobilin, or hydrochinon in the urine, may interfere with the reaction. In such cases he recommends the removal of the pigment by the addition of a few drops of lead acetate solution or a little animal charcoal and then filtering. It is often possible to obtain a reaction by shaking out the acidified urine with ether or amyl alcohol, neither of which extracts the diazo-substance. The intense yellow reaction which we often meet with in the urine of phthisical patients has been shown by Burghart to be due to phenol which is often present in excess in the urine of the severe cases and as a rule completely obscures the reaction. The excess of phenol in such urines can be easily demonstrated by adding to a test tube of the urine some strong nitric acid and boiling. After the mixture has cooled some bromine water is added and if a marked turbidity of the urine is produced an excess of phenol compounds may be assumed. This excessive production and excre- tion of phenol compounds is often observed in severe cases of phthisis and often accounts for the frequent absence of the reaction at that time. Cases, however, of excessive phenol excretion have been observed in which the diazo-reaction could still be easily obtained so that other causes must play some part. One possibility is that the excretion of the diazo substance in the urine takes place irregularly as has been noted in the normal excretion curve of other substances, no- tably urea. As is well known the urea of the body is not excreted constantly, but there may be a slight retention for a day or so and then the surplus will suddenly be got rid of through the kidneys. This may also be true of the diazo substances. As an evidence of this we may note that in the fourteen cases of phthisis which died without a positive reaction having been present, ten had an active chronic nephritis with casts and albumin. The urinary excretion in nephritics is known to be exceedingly ir- regular and retention of metabolic products is often seen extending over considerable periods of time and it is not improbable that the diazo bodies are subject to the same conditions. 6 The reagents used were those recommended by Ehrlich in the Charité-Annalen in 1886. Two solutions were employed which were mixed at the time of using: I Sulfanilic Acid ... . . . . . . . . . . . . . . . . . . . I Strong Hydrochloric Acid. . . . . . . . . . . . 50 Aq. ad. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IOOO II Sodium Nitrite ... . . . . . . . . . . . . . . . . . . . 5 Aq. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IOO Two c.c. of II are added to one hundred c.c. of I. The mixture will keep in good condition for about two days in a cool place. Equal parts of urine and the reagent are mixed in a test tube, and one-seventh or one-eighth part of ammonium hydrate is added. The salmon pink color of the foam obtained after shaking is the essential point in a positive reaction. If we sum up in concise form the results of this study they will be as follows: I. If the urine of a case of pulmonary tuber- culosis shows no diazo-reaction and a kidney le- sion can be excluded the prognosis is favorable. Only ten per cent. of the moderately severe cases here recorded gave a reaction and in a number of these the reaction disappeared on treatment. Early cases not ill enough to apply for hospital treatment do not give the diazo-reaction. II. If the urine of a case of pulmonary tuber- culosis shows an Ocasional diazo-reaction the prognosis is not necessarily grave as only some 66 per cent. of the patients showing an occa- sional positive reaction die. III. If the urine of a case of pulmonary tu- berculosis shows a continuous strong diazo-reac- tion the prognosis is very grave since a large pro- portion of such cases die within six months. IV. The presence of a diazo-reaction on the first examination of a patient should not debar the case from a thorough trial of climatic treat- ment in a proper sanatorium. I wish to express my obligation to Dr. Norman E. Ditman, late pathological interne at St. Luke's Hospital, for his valuable assistance in carrying out the reactions and in the preparation of this paper; also to Dr. J. D. Condit and Dr. F. G. Hodgson, who carried out a portion of the tests during their service as pathological internes at St. Luke's Hospital. 7 BIBLIOGRAPHY. I. Asada-Inaug. Diss, Erlangen, 1991. º - * 2. Beck—Prognostic Bedeutung d. Diazo bei Phthisikern, Charité, Annalen. Bd. 19, 1884. 3. Becker—Münch. med. Woch., 1900. º . Burghart—Berl. klin. Woch., 1899; Berl. klin. Woch., 1901, p. 276; Deut. med. Woch., Vereins. Beilage, I90I, p. I95. 5. Clemens-Deut. Arch. f. klin. Med., 1899, p. 74. 6. Damen—Nederl. Tijdschrift voor Geneeskunde, 1900, p. I 185; Abst. in Cent. f. innere. Med., 1901, p. 181. 7. Ehrlich—Zeit. f. klin. Med., 1882, p. 285; Charité An- nalen, 1883. 8. König–Klin...therap.. Woch., 1900. 9. Krokiewicz—Wiener klin. Woch. 1898, p. 703. io. Michaelis—Berl. klin. Woch., 1900. II. Rutimeyer—Corresp. f. Schweitzer, Aerzte, 1890. 12. Schraeder and Naegelsbach–Münch. med. Woch., 1899, D. I.339. [Reprinted from THE MEDICAL NEws, August 8, 1903.] A SIMPLE AND RAPID CHROMATIN STAIN FOR THE MALARIAL PARASITE. BY FRANCIS CARTER WOOD, M.D., OF NEW YORK ; INSTRUCTOR IN CLINICAL PATHOLOGY, COLLEGE OF PHYSICIANS AND STURGEON S. THE recent increase in our knowledge of the alterations in the morphology of the blood which occur in disease is largely due to the simplifica- tion in the technic of staining which has been so pronounced in the last few years. As a result of this technical simplicity the clinician has at his disposal the means for the rapid and easy prep- aration of suitably stained blood smears, and in consequence a large amount of material has been observed and recorded. Believing that any im- provement in the technic of blood staining which tends toward ease and simplicity of manipulation may be of use to students and teachers of blood morphology, the writer has felt that certain meth- ods which he has found valuable may not be with- out interest. The first experiments were based upon an ac- cidental observation that old alcoholic solutions of the eosinate of methylene blue in the form de- vised by Jenner", and republished as an indepen- dent discovery by May and Grünwald” three years later, are capable of giving a fairly good chro- matin stain in the young forms of the malarial parasite. The methylene blue component of the dye is apparently altered by the alkali dissolved from the glass of the container and a portion is transferred into what has been shown by Michaelis" to be closely allied to, if not identical with, methylene azure. If a slide containing ma- larial parasites is stained with this ripened Jenner dye and slowly washed off in distilled water, a moderate chromatin stain is produced showing the larger nuclear masses of the ring forms and also the reddish granules occasionally seen in the bodies of the lymphocytes. Such a selective stain is not, however, produced on Smears which are dropped quickly into distilled water or washed off under the tap, for the peculiar qual- ity of the stain which causes it to color the chro- matic substance appears only when the dye is in a watery solution and depends perhaps upon some phenomenon of dissociation of the compo- 2 ments of the salt in water. The alcoholic Solu- tion is not dissociated, and its pure blue color is immediately altered to a reddish purple by the addition of water with the formation of an abun- dant precipitate. A number of observers have deyised staining combinations in which advantage has been taken of the solubility of the eosinate of methylene blue and methylene azure in methyl alcohol and the power which the mixture has to produce a chro- matin stain when diluted with water, among them Leishman,” Wright," and Reuter." The lat- ter uses the alcoholic stain only as a stock Solu- tion to be diluted with water before use; the others fix and stain at the same time in the methyl alcohol in which is dissolved the precipitate formed by mixing eosin and methylene blue, con- taining some methylene azure produced by pre- viously heating the methylene blue with a dilute alkali. - To obtain a chromatin stain with the Leishman or Wright mixtures it is necessary to dilute the fresh alcoholic dye with distilled water after it has been in contact with the smear for a few min- utes, and allow this combination to act for from three to ten minutes upon the blood spread, but the precipitate which immediately forms when the mixture is diluted is very likely to be depos- ited upon the Smear and give rise to annoying artefacts which can only be removed by treating the slide with strong methyl alcohol and repeating the staining process. The writer has also found after a considerable experience with both meth- ods that it is difficult to obtain constant results even when using smears of the same age and thickness and the same preparation of the stain. The Wright stain especially, when it is suc- cessful, affords an exceedingly instructive smear and demonstrates the chromatin granules of the malarial parasite sharply and well; but as a stain for general class use or for preparing large series of slides for students, the writer has not found it as useful as a much simpler method, using the purified methylene azure as obtained from Grübler. The use of the pure dye, the preparation of which we owe to Giemsa," obviates the uncertain and tedious process of preparing the methylene azure solution by the treatment with alkalies and gives the staining method much of the certainty and simplicity of the old eosin and methylene blue stain so much used before the in- troduction of the Ehrlich and Jenner combina- 3 tions. Giemsa and Nocht” recommended staining slides for some time in a dilute mixture of the eosin and methylene azure, a method which is necessary for staining old slides, but unnecessary for fresh preparations. The writer has found that while this eosin methylene azure combination will stain slides which are from two to three years old and produce under these circumstances an ex- cellent chromatin stain without any of the dif- fuse bluish color which is usually seen in prepa- rations of slides over one month old, a fact of considerable advantage, especially for the pur- pose of restaining faded demonstration slides or old type specimens which show some especially in- teresting features, that an equally good stain of fresh specimens can be obtained as follows. wº Fix the preparation for one minute in strong methyl alcohol, wash off in water and stain the slide for a few seconds with a one-tenth-per-cent. aqueous solution of yellowish eosin. The surplus of the eosin solution should be poured off and a few drops of a one-fifth-per-cent. solution of methylene azure poured over the slide. The stain- ing is complete in from one-half to one minute and the slide is then washed with distilled water for fifteen seconds and dried with blotting paper. The best paper for this purpose is a thick, smooth- surfaced variety used by photographers to dry prints. The close grain and hard finish of this paper prevents the deposition of any fiber on the surface of the blood smear. A longer treatment with the azure solution produces a deeper chromatin stain, a shorter treat- ment produces a stain of less intensity. If a stain of any particular depth is required the whole pro- cess may be easily observed under the microscope, selecting a group of leucocytes as a guide. The stain will be seen to gradually color the nuclei of the polynuclears a deep red, and at the proper point the neutrophile granules will be easily vis- ible with a good four mm. dry lens. At the same time the nuclei of the large and small lympho- cytes assume a deep red color and the cell bodies of the large forms a pale blue, while the bodies of the small cells will be a deep blue. At this stage of the process the malarial parasite is of a pale blue color with deeply red chromatin gran- ules. If the methylene azure has remained in con- tact with the smear but a short time the red cells will be of a bright pink color, but if the staining is prolonged the red cells become bluish or green- ish in color. After prolonged staining with the 4 azure the color of the neutrophile granules gen- erally fades and the bright pink of the eosin is replaced by a bluish tint. For general pur- poses, therefore, it is best not to stain the slide too long with the methylene azure. By increas- ing the strength of the azure solution up to one- half of one per cent. the staining process with that substance may be shortened to a few seconds, but overstaining is much more likely to take place. If the blue color of the red cells is very marked the excess may be removed by treating the smear for about one second with dilute alcohol, say of about 80 per cent., and then washing off in water. Some color is likely to be removed from the nuclear chromatin by this process of decolorization, and the red of the finer chromatin particles of the gametes loses its sharp outline or may be entirely extracted so that the alcohol treatment is not to be recommended for general use. Thus by the use of this process a good chromatin stain can be obtained in three minutes after the blood smear is secured. Such preparations may also be used for differential counting, though for general pur- poses the Jenner mixture is preferable because the eosinophile and basophile granules are on the whole more characteristically stained. The large and small lymphocytes on the other hand are bet- ter demonstrated with the methylene azure. The large mononuclears cannot be constantly sepa- rated from the large lymphocytes by any of the dyes containing methylene blue, and are best classed with the large lymphocytes. Beginners may mistake the blood plates, which often lie in depressions in the red cells, for malarial parasites, especially as they stain red with the chromatin dyes. The distinction must be made by the dif- ferences in morphology. The blood plate is not sharply outlined and the nuclear substance is in the form of an indistinct mass not surrounded by a clear zone, while the small ring-form of the ma- larial parasite is very sharp in its outline, as is also the chromatic substance, the latter being sur- rounded by a clear zone and beyond this by the blue ring of the body of the parasite. The staining of old preparations cannot be achieved in a satisfactory manner with any of the usual blood stains except the Ehrlich triacid, which does not demonstrate the malarial parasite, By the use of dilute mixtures of eosin and methy- lene azure, as suggested by Nocht, it is possible to obtain perfect stains from slides which are months or even years old. The older the slide the 5 more dilute must be the stain, and the longer it must remain in the staining fluid. Two methods may be used. The slide should be fixed for at least half an hour in methyl alcohol. Then it may be allowed to remain for from four to twenty-four hours in a strong mixture containing about 50 c.c. of a one-to-a-thousand methylene azure solu- tion to which has been added from two to five c.c. of a one-to-a-thousand aqueous eosin. The mixed solutions are best placed in a Coplin jar and the slides allowed to stand vertically. As a rule, a precipitate is not formed, but if any should form it is not likely to be deposited on the surface of the smear. In case such a precipitate should be deposited on the smear it can be removed by dip- ping the slide in strong ethyl alcohol for a few seconds and then washing off in water. Slides stained in this manner are usually much over- colored and must be decolorized in 80 per cent. ethyl alcohol and then washed in water. By this means the excess of the blue dye is removed and the chromatin particles in the malarial parasites are well brought out. The same result may be obtained by staining for from twenty-four to thirty-six hours in the above mixture diluted about five times with distilled water. It is not usually necessary to decolorize slides stained in this dilute mixture. BIBLIOGRAPHY. L. Jenner. Lancet, Vol. I, 1899, p. 370. May and Grünwald, Cent. f. innere Med., Bd. XIII, 1902, p. 265. L. Michaelis, Cent. f. Bakt., Bol. 29, 1901, p. 763. Leishman, Brit. Med. Jour., 1901, p. 757. Wright, Jour. of Medical Research, 1902, p. 138. Reuter, Cent. f. Bakt., Bd. 30, 1901, p. 248. Giemsa, Cent. f. Bakt., Bol. 31, 1902, p. 429. Nocht, Encyk. d. mik. Technik, p. 785. THE MIXED TUMCRS OF THE SALIVARY GLANDS. BY FRANCIS CARTER WOOD, M.D., OF NEW YORK, Instructor in Clinical Pathology, College of Physicians and Surgeons, Columbia University; Pathologist to St. Luke's Hospital. (A Study from the Department of Pathology of Columbia University.) It has long been known that the tumors of the salivary glands possess, as a rule, a very peculiar morphology which does not correspond to the structure of the tumors found in other organs. The greater number of the salivary tumors be- long to a class known as mixed or, what is perhaps better, complex tumors; that is, new growths containing a consider- able variety of tissues generally regarded as of mesoblastic origin, such as cartilage, myxomatous tissue, fat, and lymphoid structures. The parenchyma cells, proper, resemble morpho- logically either connective-tissue cells, in which case the tumors are considered as Sarcomata, or endothelial cells, in which case the growths have long been called endotheliomata. Beside these definitely mesoblastic structures these com- plex growths contain cells which resemble epithelial, endothe- lial, or connective-tissue cells, and accordingly the tumors have been considered as of epithelial, endothelial, or of a sarcoma- tous nature. These cells, which for convenience may be desig- nated as parenchymal to distinguish them from the cells of the stroma, are present in greater or less abundance in the new growths, and give them their peculiar morphology. The varia- tions which exist in the proportions between the stroma and the parenchyma and in the morphology of the cells of the stroma and parenchyma have given rise to much confusion in the clas- sification of these tumors, and have rendered difficult the exact determination of their histogenetic relationships. Pathologists have endeavored to escape responsibility by FRANCIS CARTER WOOD. coining compound titles to include all the forms of tissues found in such a complex growth; thus, adeno-myxo-chondro-sarcoma has frequently been used to designate tumors of this group. This additive method of naming tumors is quite unscientific, and gives no just idea of the pathological relations of the growth nor of its clinical character. Hansemann has carried it to an extreme. He proposes the following division of the mixed tumors which are at present generally considered to be of endo- thelial origin: I. Endothelial Carcinoma. 2. Endothelial Sarcoma. 3. Endothelial Carcinosarcoma. - 4. Endothelial tumors with development of special parts of the stroma. (a) Cylindroma; (b) Myxoma; (c) Chon- droma; (d) Scirrhus; (e) Mixed forms with transition into Sarcoma or carcinoma. 5. Endothelial Adenoma. The question immediately arises in connection with the use of such a scheme of classification as to the distinctive morphol- ogy of an endothelial tumor and of endothelium. The orig- inator of the term endothelium was His, in 1865; and by it he denoted the cellular linings of the serous cavities, of the blood- vessels, synovial membranes, lymph spaces, etc. Golgi then applied the term to certain tumors derived from the meninges and called them endotheliomata. Much opposition has arisen of late among embryologists to the use of the word endothelium as designating a particular group of lining cells, and the general trend of opinion is to replace it by epithelium. Thus, Stöhr calls the cells lining the blood-vessels, epithelium; and Hert- wig has shown that the probable derivation of the flat cells lining the coelom cavity is from the hypoblastic layer of the embryo, and that therefore these cells should be called epithelial. This view is shared by Klaatsch in a paper on the classification of tumors on an embryological basis. The French school also are inclined to call all flat covering cells epithelium, and for consistency, therefore, tumors arising from these cells, carcino- mata or adenomata. Minot and Kollman, on the other hand, 2 MIXED TUMC)RS OF SALIVARY GLANDS. consider the lining cells of the coelom cavity and of the blood- vessels and lymph spaces as mesoblastic in origin, and therefore endothelium. Marchand would limit the term endothelium to the vascular linings and call all other lining cells, surface cells (Deckzellen). The name epithelium he considers should be restricted to all cells, without regard to their origin, which line hollow spaces and free surfaces, these cells lying closely to- gether without well-developed interstitial substance. He sug- gests as a suitable word for all the surface layers of cells (Deckschichten) the Greek derivative, epithem. - In the mixed tumors of the salivary glands the parenchy- mal cells are arranged in long anastomosing strands which often form alveoli lined with one or more layers of flat cells. The difficulty which has arisen in the classification of these tumors lies chiefly in the determination of the histogenetic rela- tionships of these cells. For many years they have been assumed to be derived from cells lining the connective-tissue spaces, and solely from morphological considerations have been regarded as endothelial cells and of mesoblastic origin. Now, inasmuch as the embryological status of the cells lining the lymph spaces has never been determined, a double assumption is made in ascribing to the cells of the mixed tumors an endo- thelial origin, for not only must the adherent of the endothelial theory show that the tumor cells are derived from the cells lining the tissue spaces, but he must also show that these cells are of mesoblastic origin, and therefore not epithelium. The chief exponent of the endothelial theory has been Rudolf Volk- mann, who in 1895 published an important monograph on the endothelial tumors. Since that time the interest in this pecu- liar group of tumors has greatly increased, and numerous papers have appeared describing mixed tumors from various portions of the body, the writers accepting, as a rule, the gen- eral classification laid down by Volkmann and the morpho- logical criteria which he assumed for the differentiation of the epithelial and endothelial growths. The following extract from Volkmann's monograph defines his position on the subject: “The characteristic morphological peculiarities of the endothe- 3 FRANCIS CARTER WOOD. lial tumors lie in the arrangement of the tumor cells in Strands and tubules, which distinguishes the growths from the Sarco- mata and gives them a close resemblance to the carcinomata. The cells of the tumors are often in very close relationship with the connective tissue of the spaces in which the cells lie, as evi- denced by the fact that the cells remain attached to the walls of the space and do not retract when the tumor is hardened in fixing fluids, as is usual in the carcinomata. The cells of the endothelial tumors line the walls of the tissue spaces without the intervention of a layer of normal endothelial cells, such as is seen in the carcinomata where metastases extend along the lymph spaces. The cartilage of the endothelial tumor arises from the fibrous connective-tissue stroma by a softening of the intercellular substance, and the production first of a myxoma- tous tissue which later develops cartilage and a homogeneous intercellular tissue. Cell masses of an endothelial nature may also develop from the cartilaginous or myxomatous areas; and in many cases the spindle cells of the connective-tissue stroma must be regarded as genetically equivalent to the endothelial cells of the solid strands and tubules.” According to Volkmann, the cells of the peripheral lymph spaces assume an active part in the growth of the tumor and, by their proliferation, form fresh extensions of the new growth. This has been shown to be doubtful both by Ribbert and Borst, who state that the cells of the tumor either grow into a tissue space and line the walls with flat cells or may grow over the pre-existing endothelial cells and produce the appearance of a proliferation of the endo- thelial cells, a condition also seen in the peripheral growth of the carcinomata. Glandular structures and growths of a car- cinomatous nature are never found in the mixed tumors; all Such appearances are due to modifications in morphology of the endothelial cells of such growths. Pathologists of the French school have, however, never accepted the current view of the essentially mesoblastic origin of the cells of the tumors of the salivary glands, and have regarded the cells as derived from the epithelium of the glands, Or, in the case of the pharyngeal and buccal tumors, as derived 4 MIXED TUMCRS OF SALIVARY GLANDS. from the small glands of the buccal mucosa. The majority of the French observers are satisfied to consider the mixed tumors as adenomata or carcinomata, and have described many of the simpler forms of the mixed tumors under this designation, reserving the name of mixed tumor for the more complicated growths containing bone or cartilage. This theory of the car- cinomatous nature of these growths fails in several particulars. From a morphological point of view, if the tumors are car- cinomata of the salivary glands it should be possible to trace some connection between the glandular structures and the tumor, as is sometimes possible in early growths of other glands of the body; but experience has shown that the tumors of this group are in a very large majority of cases encapsulated, and show no connection with the gland; indeed, they are often at some distance from it. From a clinical point of view, it is diffi- cult to explain why the carcinomata of the salivary glands should differ so much in their clinical features from the car- cinomata of other glands; for it is well known that these tumors may be present for twenty or thirty years without giving rise to cachexia or involving the surrounding structures. The theory of direct epithelial derivation also does not explain the presence of embryonic structures nor of cartilage; the latter tissue being present in a large proportion of these tumors. Finally, true carcinomata of the salivary glands have been ob- served with a morphology corresponding to those arising in the other epithelial glands and with a clinical course which is considered as characteristic of carcinoma; that is, a rapid and progressive involvement of the surrounding structures and an early invasion of the regional lymph nodes. Pitance, in his thesis published in 1897, suggests that the direct derivation of the parenchymal cells of the mixed tumors from the highly differentiated glandular epithelium is improb- able, even from a morphological stand-point, and that it is much more likely that the epithelium forming the cells of the tumor is derived from masses of cells left in or about the glands during the process of development. This mode of derivation might then explain the clinical peculiarities and the presence of 5 FRANCIS CARTER WOOD. cartilage and embryonal tissues; the rudiments of which might have been left at the same time that the deposition of the glan- dular epithelium took place. The fact that the epithelial cells have not been in a position to develop functionally might also account for their alteration into an indifferent type not resem- bling very closely the normal cells of the fully functionating salivary gland. This view, merely suggested as an hypothesis, has not been generally adopted by pathologists. An important paper on the salivary tumors by Hinsberg appeared in 1899, in which this author developed much the same idea as that of Pitance, but in greater detail and as the result of a large amount of embryological research on the ana- tomical conditions underlying the development of the salivary glands. He points out that in all of the nine tumors which he examined definite epithelial structures could be demonstrated. In itself this was not a new observation, as Nasse, Volkmann, Mauclaire, Cavazzani, and others had figured and described pearl formation in tumors, which they, however, regarded as endothelial in nature. Cavazzani even figures spine cells, which, however, he considers as of endothelial origin. The importance of Hinsberg's work lay in the evidence which he adduced to show the very intimate relations of the parotid and the sub- maxillary glands to the mesoblastic structures of the first and second branchial arches. - Wilms, in a recent paper, entirely agrees with Hinsberg in considering the parenchyma of the mixed tumors as of epithelial origin, but differs from him in the embryological interpretation of some of the morphological findings in the tumors. Landsteiner has recently published an article in which he analyzes the results of the examination of twenty-seven tumors, chiefly from the salivary glands, among them an adenoma of the parotid. Of twenty-six mixed tumors examined, he found Squamous epithelium and prickle cells in ten. The epithelial remnants were found in five tumors of the parotid, in one of the submaxillary, in two of the lip, one each of the palate and the neck. All but one of the tumors containing epithelium also contained cartilage. - 6 MIXED TUMCRS OF SALIVARY GLANDS. Ribbert, in his text-book on general pathology, and Lu- barsch, in his numerous critical reviews of the subject of tumors, are both very guarded in the expression of their ideas on the origin of the cells of the so-called endothelial tumors. Ribbert especially considers the salivary mixed tumors as in all probability of epithelial origin, and his views on the general subject of endothelial new growths may be illustrated by the following quotation: “It is customary to make a diagnosis of endothelioma when, in spite of the carcinomatous arrange- ment of the cells, the organ in which the growth is found lacks epithelial cells. But it is to be remembered that developmental remains of epithelial tissue may be found in abnormal places; such as flat epithelium in the deep connective tissues of the neck, derived from the branchial clefts, or fragments of pancreas in the walls of the stomach or duodenum. It is quite probable that many of the so-called endotheliomata are in reality of epithelial nature; for it is exceedingly difficult at times to decide whether the cells occupy a tissue space or a lymphatic vessel. When an endothelial tumor reaches a certain size, new connective tissue and blood-vessels begin to be formed, and the resulting picture may vary greatly from the original growth. The cells then grow in more or less closely packed strands and, with the diminished fibrous tissue, a deceptive alveolar structure may be obtained in sections. A further appearance due to growth is the increasing closeness of relationship to the blood-vessels. The cells may become arranged in layers around these, giving a picture of angiosarcoma.” The remainder of the papers which have appeared during the past six years favor the view that the cells of the paren- chyma are of endothelial derivation. The evident and extreme diversity of opinion concerning the classification of the mixed tumors of the salivary glands, and the peculiar interest which attaches to them from their rarity and complicated structure, led the writer to examine a large number of growths which may properly be included in the class of mixed tumors. In the course of this study complex tumors were examined from the salivary glands, the lips, the 7 FRANCIS CARTER WOOD. palate, the orbit, the antrum of Highmore, the lachrymal gland, the thyroid, kidney, ovary, testicle, and lung, as well as examples of some of the simpler types of endotheliomata from the meninges, the pleura, and the peritoneum. It seemed best to limit the present paper to the consideration of the tumors of the buccal and the salivary group, inasmuch as the embryo- logical conditions underlying the formation of the tumors of other organs, especially the testicle and kidney, differ greatly from those connected with the origin of the salivary mixed tumors. The writer has been indebted for some of his material to the surgeons of several of the New York hospitals, and he wishes to express his obligation to Dr. Robert Abbe in allow- ing him to use the records of six cases; to Dr. Francis H. Markoe for the use of five cases; to Dr. Charles McBurney for the use of eight cases; to Drs. B. F. Curtis and C. L. Gibson for three cases each; to Dr. J. H. Blake for two cases; and to Dr. F. W. Murray for one case. Case XXV was kindly given to me by Dr. F. S. Matthews. Four other cases were put at my disposal by Dr. J. H. Larkin, to whom I am under obliga- tion for slides and material from his large collections of mixed tumors. The other twenty-seven tumors are from the collection of the Department of Pathology, College of Physicians and Surgeons, Columbia University, the large material of which was put at my disposal through the kindness of Professor T. M. Prudden, whom I also wish to thank for much assistance and advice during the progress of this study. Of the total of fifty-nine tumors from the salivary glands, lip and pharynx, selected for this report, fifty-four may be considered as un- doubted mixed tumors of the so-called endothelial type. Two might perhaps be considered by most observers as sarcomata; One may possibly be an adenoma, though it does not resemble Other adenomata in my collection, which are undoubted growths of an epithelial nature. Case II is not a mixed tumor in the strict sense of the term, but is of interest from an embryo- logical point of view. No attempt has been made to give a full clinical and ana- tomical description of each case, as a long series of such reports 8 MIXED TUMIORS OF SALIVARY GLANDS. can be found in Volkmann and the other German writers just mentioned. No morphological description, however complete, will enable one to appreciate as much of the appearance of a tumor as a drawing, so that frequent references have been made to the plates instead of giving the microscopic details in full. It will be frequently noted during the description of the morphology of the individual tumors, that references will be made to drawings from other specimens. This is necessary simply for economy in the number of plates. A considerable number of the drawings are intended as types, and will be referred to as such. For example, Plates I and VI, Fig. I, though not drawn from the same tumor, show a morphology which is characteristic of a large number of the mixed tumors, and which may be designated as typically endothelial in nature. DETAILS OF CASES. CASE I.—St. Luke's, No. 2. Parotid tumor. The growth was removed from the left cheek of a male, aged thirty-nine years. The tumor had been noticed for two years, during which time it had grown slowly. The growth was hard, painless, and freely movable. When removed the tumor was found to be roughly spherical. The surface was slightly lobular and smooth. A distinct capsule is present of from one to three millimetres in thickness. A por- tion of the parotid gland is still adherent to the dorsal aspect of the mass. The tumor measures roughly four by three by two centimetres. The parenchyma of the growth is arranged in solid strands and in alveoli. The strands spread out in all directions through a soft cellular connective tissue and lose themselves in it. (Plates I and VI, Fig. I.) The cells of the strands are spindle or oval in shape with large nuclei. The chromatin of the nuclei is evenly distributed, so that a net-work is not easily made out. From the tapering ends of the strands the cells often give off long filaments to the surrounding connective tissue. The alveoli are small and oval in shape. They often lie in the course of one of the solid strands. The centre of the strands is filled with a mucous mass staining blue with haematoxylin. In some places the alveoli are 9 FRANCIS CARTER WOOD. suggestively like the alveoli of a secreting gland; but the general type of the growth is that which has been called endothelial. There are no pearls or spine cells in the sections examined. There is no cartilage. Elastic tissue is abundant. The remains of the gland which are attached to the growth are normal. The mor- phology of the tumor is similar to that shown in Plate I, Fig. I. No recurrence of the growth has taken place after five years. CASE II.-St. Luke's, No. 727. The tumor was removed from a seven-months-old female infant. The growth was noticed soon after birth, and its progress was slow for two months and then became very rapid. The tumor lies on the left side of face and neck over the parotid gland, with which it is in contact. Numerous dilated veins can be seen on its surface. It is easily reduced in size by pressure. It measures before removal seven by four by five centimetres. On gross examination after removal, the tumor is composed of fat tissue and soft fibrous tissue. The vessels are not very noticeable in the hardened specimen. There is no capsule. Microscopic examination shows a diffuse cellular growth composed chiefly of oval and spindle cells. The tissue resembles embryonic connective tissue rather than sarcoma. In the masses of spindle cells can be seen occasional alveoli of cylindrical epithe- lium entirely distinct from the rest of the growth, which resemble the alveoli seen in the tumors regarded as of endothelial nature. There is a moderate amount of fat scattered through the speci- men. No cartilage is present; no form of degeneration either in cells or stroma. Elastic tissue is not abundant. This tumor, though perhaps not properly classed as an endothelioma, seems of sufficient interest to include here, for it is an excellent example of a congenital tumor of the parotid region containing fibrous tissue of a cellular, embryonic type, and also distinct epithelial remains derived in all probability from rudiments of the parotid. No recurrence in three years. CASE III.-St. Luke's, No. 1675. Male, aged forty-one years, who fifteen years before had noted a small, hard nodule below and behind the right ear. The tumor was quite painless and not tender. The growth had gradually increased in size until it now measures twelve by eight centimetres. The growth has taken place in an anterior direction, so that the tumor now lies chiefly anterior to the ear, extending for some three centimetres 10 MIXED TUMCRS OF SALIVARY GLANDS. along the border of the inferior maxilla. The ear is pushed for- ward and upward by the tumor. The form of the growth is irregularly oval, and presents a number of cartilaginous nodules which project three or four millimetres above the surface. Other portions are cystic. The tumor is not adherent to the surround- ing tissues. The neighboring lymph nodes are normal. The growth was removed by operation without difficulty, and no recur- rence has been reported in two years. The tumor when hardened measured seven by five by five centimetres. The cut section was pale yellow with small bluish nodules of hyaline cartilage scattered through its substance. There are many small, softened areas filled with myxomatous tissue. Some of these have broken down and formed small cysts. The capsule is of fibrous tissue about two millimetres thick and contains a few small vessels. The microscopical examination of the tumor showed it to be very largely composed of cartilage and soft embryonic connective tissue. The cellular portion of the growth was only moderate in amount. The cartilage is in general purely hyaline, with a few oval cartilage cells in the stroma. In places, however, the connective and elastic tissue of the tumor invades the cartilage, and it becomes more cellular, and contains many long branching and Spider cells, the prolongations of which join in with the fibrous tissue of the stroma. In some portions of the tumor and in the capsule the connective tissue is quite dense, but in general it is very loose and softened with mucous degeneration. Hyaline degeneration of the connective tissue is also seen to a very considerable extent, often extending over large areas, in which all trace of fibrillation is lost. Weigert’s stain for elastic fibres shows the whole growth to be traversed by a fine net-work of elastic fibres penetrating the cartilage and surrounding the alveoli of the parenchyma cells. The parenchyma cells proper are of the so-called endothelial type. That is, they are oval or polyhedral cells with oval, deeply staining nuclei, which line small alveoli and tissue spaces. They are in intimate relation with the surrounding tissues, and may give off prolongations which are lost in the connective-tissue fibrillar among which the cells lie. There is no connective tissue between the cells when they lie in compact masses, nor does the elastic tissue pass between them under these conditions. Some of the alveoli contain masses of homogeneous hyaline material. 11 FRANCIS CARTER WOOD. CASE IV.-St. Luke's, No. 1745. The tumor was removed from the left parotid region of a man of fifty-nine years, who first noticed the tumor fifty-three years ago. It was then the size of a Small nut and freely movable. For many years the growth was exceedingly slow, but in the last few months it has been very rapid. The ear is pushed back by the growth. The tumor is very hard and adherent both to the skin and the deeper tissues. It cannot be moved in any direction. Its surface is smooth. In the neck are a few large hard nodes which have appeared recently. The patient's general condition is good. The tumor was removed with some difficulty, as it was adherent to the deeper tissues, and a large number of the cervical lymph nodes had to be removed. When hardened, the mass measured six by seven by twelve centimetres. It is pear-shaped, with the larger portion above over the parotid. The cervical fat is filled with hard, enlarged lymph nodes. The internal surface, directed towards the parotid, is rough from separation of the tumor from the underlying tissues. The outer portion, directed towards the surface is smoothly encapsulated, and there is a thin capsule between the tumor and the remnants of parotid tissue. The cut section shows two different appearances. The superficial portions are transparent, with faint yellow strands running through them, the deeper are opaque and white. The nodes are also opaque and white. Microscopical examination of the growth shows a similar variability in the tumor. The peripheral portions possess the morphology designated as endothelial (Plate VI, Fig. I), with a soft, fibrous stroma and long branching strands and alveoli, some of the latter filled with hyaline material. The deeper portions are composed of the same endothelial structures infiltrated with car- cinoma. The carcinomatous growth resembles that of an infil- trating epithelioma rather than that of a glandular carcinoma, such as one would expect in a carcinoma of the parotid. The cells are large and flat, staining deeply with eosin. In some areas intercellular spines can be seen. Mitotic figures are fairly abun- dant. The parotid, which is separated from the growth by a fibrous capsule, is normal in appearance, and contains but little carcinomatous infiltration. The nodes are filled with the car- cinomatous new growth and very little lymphoid tissue remains. In the opinion of the writer, the best explanation of the con- 12 CASE IV.-Epithelioma arising in a mixed tumor of the parotid gland. MIXED TUMCRS OF SALIVARY GLANDS. dition is that the patient had since childhood a tumor of the endothelial morphology, and that the recent rapid growth is the result of the carcinomatous change which has taken place in the epithelial cells of the so-called endothelial new growth. That the carcinoma is not derived from the parotid seems probable, for the gland is not extensively invaded, as it would be if the carcinoma were primary. Landsteiner describes a similar case in which malignant changes had taken place in a chondromatous tumor of the submaxillary gland, with the formation of growths of an epitheliomatous character which had broken through the tumor capsule and infiltrated the surrounding tissues. The regional lymph nodes were not invaded, in which point the case differs from the above. CASE V.—St. Luke's, No. 1906. The patient was a male, aged forty-four years; has always been well and strong. Six years before admission to the hospital he first noticed a swelling of the right cheek just above and anterior to the parotid gland. The tumor was freely movable in a vertical direction, but not laterally. It was not painful and was soft, feeling a good deal like a sebaceous cyst. The skin and mucous membrane of the cheek are not adherent to the growth. The removal of the tumor was carried out without difficulty, the growth being encapsulated and not involving any of the deeper tissues. On removal it was found to be roughly oval, measuring five by four by five centimetres. The surface was covered with small nodules about five millimetres high. The cut section was uni- formly opaque and of a dull yellow and white color. No carti- laginous areas could be made out. Microscopical examination of the tumor shows a growth composed of dense fibrous tissue with only a small amount of parenchymal substance. The fibrous tissue is rather cellular in a few places, the cells being fusiform, with large oval nuclei and resembling the fibroblasts seen in granulation tissue. Scattered irregularly through the fibrous tissue are many open spaces lined in general with a single layer of flattened cells which stain deeply with eosin, much as is seen in cornified epithelium. The contents of all of these alveoli have not been preserved in Zenker prepara- tion or in specimens hardened in alcohol and formalin, though hyaline masses are present in a few. Evidently the fluid was of a serous nature and contained but little solid matter. Some of 13 FRANCIS CARTER WOOD. the spaces, however, can be explained by the fact that the cells which originally filled them have fallen out during the manipula- tion of the sections, for solid masses of cells resembling cornified epithelium are present in parts of the tumor. Some of these areas are quite extensive, and spine cells with epithelial fibrillations can be easily made out by suitable staining and the use of high powers. The central spaces of other of the alveoli are filled with large flat cells with faint nuclei and poorly staining cell bodies, which are much like the large flat epithelial cells seen in the alveoli of mammary adenomata in which the milk-ducts have been occluded by the tumor growth. It is evident that a con- siderable proportion of the parenchymal cells are of epithelial origin, though the morphology of the tumor is that described as endothelial. No recurrence in a year. CASE VI.-The specimen was removed from a female seventy years of age. She had noticed a tumor on the inner surface of the left cheek, near the opening of Steno's duct, for at least twenty years. It was oval, movable, and quite painless. The mass had increased in size very slowly. The tumor was easily shelled out and measured about two by one and one-half centimetres. It was smoothly encapsulated and the surface was lobular. On cross section the texture was fine and the surface a yellowish white. A few scattered islands of cartilage, none over three millimetres in diameter, could be recognized by their transparency. Micro- scopically, the growth consisted chiefly of cartilage, soft em- bryonic connective tissue with spider cells, and abundant mucous degeneration, typical endothelial strands and alveoli, and, finally, well-formed epithelial pearls. The cartilage is of the hyaline variety and contains a good deal of elastic tissue. The cartilage passes imperceptibly into either the soft connective tissue or the closely packed cells of the anastomosing endothelial strands. In- deed, there are no sharp boundaries between the various tissues of the tumor. The connective tissue in the centre of the growth is very soft and contains but few cells, and these send out long fibrillae which join with those from other cells and form a net- work, in the meshes of which is found the mucous substance staining deep blue with haematoxylin. In this tissue are seen occasional spherical cells with one or two nuclei. They exactly resemble the cells of hyaline cartilage. Epithelial masses are also present in the mucous tissue. They have no connection with the 14 MIXED TUMCRS OF SALIVARY GLANDS. surrounding structure, but lie isolated as small spherical groups of flattened cells. The so-called endothelial strands and alveoli possess the morphology common to these structures. (Plate II, Fig. I.) The alveoli are filled with hyaline masses. In other portions of the tumor the fibrous tissue is more dense though still very cellular. The epithelial pearls are found chiefly in these areas. The pearls are either quite distinct from the so-called endothelial portions of the growth, or they lie in the course of one of the long branching alveoli of endothelial cells, or they can be seen lying in the solid strands of the endothelial cells. Finally, the epithelial cells may be seen lining the walls of small cavities in the tumor. A few cells in all these masses when stained by Kromayer's method and examined in glycerin or even in balsam show intracellular bridges and the fibrillations characteristic of epithelial cells. Only a certain number of cells show this mor- phology, and they are chiefly the cells which take on a deep eosin stain. The cells of the so-called endothelial type do not show any such structures. (Plate I, Figs. I and 2; Plate V, Figs. I and 2.) - - CASE VII.-St. Luke's, No. 1592. The tumor was removed from a female aged fifty-six years. Three years before her admis- sion to the hospital she noted a small nodule in left parotid region. For two years there was no increase in size, but of late the patient thinks there has been a slow but steady increase in the size of the tumor. Health perfectly good at present. The tumor measures after removal about two by three centimetres. It is lobulated and surrounded by a thin capsule. It is soft in texture, and there are a few areas of softening. Microscopically the tumor is very cellular, with the strands of cells separated from each other by a delicate fibrous stroma which has undergone hyaline degeneration, especially along the blood-vessels, which are moderately abundant. Some of the strands contain alveoli filled with hyaline material. No cartilage is present, nor definite epithelial structures, but an abundance of elastic tissue. No recurrence after a year. CASE VIII.-St. Luke's, No. 522. Tumor was removed from the left submaxillary region of a woman of forty-seven years. She had noticed the growth for eleven years. It was removed, and about three months after the operation she noted a recurrence, which has grown slowly for two years and for the past month quite rapidly. It is now about two centimetres in diameter and 15 FRANCIS CARTER WOOD, causes pain and difficulty in swallowing. The tumor is but slightly movable and has no sharp outline. The patient's condition is good. The tumor when removed was found not to be encapsulated and measures roughly some two centimetres in diameter. It is em- bedded in a mass of fat, into which it merges imperceptibly. The cut section is rather soft and translucent, without any marked macroscopic characteristics. Microscopically the growth is composed of alveoli in a con- nective-tissue stroma. The centre of each alveolus is filled with mucus; the periphery lined with flattened cells. No cartilage, and but very little elastic tissue is present in the recurrence, though it is usually very abundant in the primary growths of this group of tumors. The elastic tissue in a recurrent mixed tumor appears to be practically all derived from that pre-existing in the connective or other tissue invaded by the growth. CASE IX.-St. Luke's, No. II31. Male, forty-five years of age. A year previous to operation had noticed a small lump size of a marble just behind and below angle of right inferior maxilla. The mass was not tender or inflamed. Soon after the growth began to affect the patient's speech, and gradually the tumor was noticed to protrude more and more into the pharynx. For the past six months growth has been rapid. The tumor at present is the size of an orange, but is freely movable, and does not interfere with the motions of the lower jaw. No recurrence after two years. Macroscopical examination of the growth shows it to measure eight by seven by six centimetres, the shape being roughly oval. The surface is slightly lobular. The consistency is firm and elastic in general, but there are harder and softer areas. On section the tumor is seen to be composed largely of cartilage lying in a soft fibrous matrix and containing a few small cysts. The whole is sur- rounded by a thin even capsule not over a few millimetres in thick- ness. The cartilage is of the hyaline variety and is quite trans- parent and of a bluish tint. The cellular portions of the tumor are opaque and yellow. Microscopical examination of the growth shows that it is made up very largely of hyaline cartilage containing in different portions a variable number of cells. The cells have the mor- phology of those seen in normal cartilage. About the edges of the cartilaginous masses the cells are more abundant and lose their characteristic shape. They are often spindle-shape or even epithe- 16 CAsF IX.-Large tumor containing cartilage chiefly. MIXED TUMC)RS OF SALIVARY GLANDS. lioid in form, and are arranged in long branching strands and spherical alveoli forming the so-called endothelial structure. The alveoli are filled with hyaline masses. Elastic tissue is not very abundant in the tumor as a whole, but in some portions of the cartilage there is a fine diffuse net-work of very fine fibres. In the cellular portions the strands are coarser and outline the alveoli. CASE X.—The tumor is a recurrent parotid tumor from a male of about fifty years. The original growth was a small tumor of the left parotid about the size of an English walnut. It was re- ported as a mixed tumor. One year later there was a considerable diffuse local recurrence over the parotid region which was excised. The recurrence penetrated between the lobules of the parotid and could be distinguished from it by the yellow color of the tumor. The whole mass was removed, and no recurrence has taken place at the end of three years. The material removed was of the en- dothelial type with anastomosing strands of flat cells forming alveoli containing hyaline material. No cartilage was present; no epithelial pearls; no large amount of elastic tissue. CASE XI.-St. Luke's, No. 4II. The tumor was removed from a female aged twenty-six years. Five years previous to the operation a tumor the size of a pea appeared in the right parotid region; has grown slowly to the size of an English walnut. None of the lymph nodes of the neck are swollen. When removed, the tumor measured two and one-half by three by two centimetres. It is a flattened, encapsulated mass with a broad, irregular base. A small additional fragment is attached to the main mass by a pedicle. Microscopically the tumor is chiefly made up of a soft cellular fibrous tissue with many nuclei and marked mucous and hyaline degeneration, the latter confined to the walls of the blood-vessels. Scattered unevenly through the whole are a few alveoli lined with flattened cells. A few of the alveoli contain hyaline material. No pearls or well-marked epithelial alveoli present. No recurrence in four years. - CASE XII.-St. Luke's, No. 630. The tumor was removed from a female fifty-six years of age. Five and a half years before admission to hospital she noticed a tender spot on side of neck with a small lump. There was some pain in tumor. The growth has been very slow and gradual. The tumor is hard and lobular with a smooth surface. It lies below the angle of the jaw on the right side and is the size of a hen’s egg. Skin is movable over tumor, 17 FRANCIS CARTER WOOD. and the latter is movable on the deeper tissues. The tumor was easily enucleated, except for one point which was adherent to the digastric muscle. It measures after removal five by three by two and one-half centimetres. The surface is lobular but smoothly encapsulated, except for the area which was adherent to the digas- tric. On section the periphery of the growth is firm and very cellular; the centre, however, has softened, and there is a ragged cavity from which the degenerated tissue has escaped. No areas of cartilage can be seen. Microscopically the growth is very cellular, the cells being arranged in tubules and long strands. The cellular areas in places have been distended by collections of mucus forming a single cyst, or the mucus may have collected in a large number of separate areas forming a large number of cysts with cellular walls, the whole walled about with trabeculae of soft connective tissue. The general type of the growth corresponds to what has been called cylindroma. (Plate IV, Fig. 2.) Though blood-vessels are not numerous, there are many ex- travasations of blood in the tumor, probably the result of the opera- tive handling of the soft tumor. No cartilage and no pearls are to be seen. Elastic tissue abundant. CASE XIII.-Recurrence of the above in about a year. The recurrence is local and in the form of a diffuse infiltration of the tissues with cells of the same type as before described: that is, flat or oval epithelial-like cells arranged in alveoli with a mass of mucous secretion in the centre. In many years the alveoli are not developed perfectly, and the cells have penetrated the tissues in long strands and masses of small cells retaining the morphology of the original tumor. No further recurrence after two years has been noticed since the thorough removal of this recurrence. The regional lymph nodes are not invaded. Very little elastic tissue in the recurrence. CASE XIV.-St. Luke's, No. 659. Parotid tumor removed from a male aged thirty-three years. Eleven years ago first noticed a lump about the size of a small nut, one and one-half centi- metres in diameter. It increased very slowly in size and was ex- tirpated eight years ago. After this there was some induration of the scar, but nothing was seen until two years ago, when a small recurrence was noticed, slowly increasing to present size. There is no pain nor tenderness. No swelling of the neighboring lymph nodes. - 18 MIXED TUMCRS OF SALIVARY GLANDS. The tumor occupies the right parotid region. It is adherent to the deeper tissues and to the skin. Its consistency is hard; it is quite lobular. The form is irregularly oval, somewhat larger below than above, and flattened from before backward. There is no distinct, smooth capsule, but the tumor substance proper is em- bedded in dense fibrous tissue, and does not ramify among the sur- rounding tissues. On section it is quite tough and firm, with harder and softer areas. The soft areas are quite translucent and resemble cartilage. The wound did not heal primarily, but showed a low grade infection with a corresponding rise of temperature up to Io2.5° F. - The microscopical examination shows the tumor to be of a very variable structure. The whole growth is divided into lobules by connective-tissue bands, which are rather dense and contain a few spindle cells. Inside of these lobules are the following struc- tures: (a) In some of the lobules the centre is made up of very fine mucous tissue containing only a few branching cells. These are the transparent areas resembling cartilage. (b) Areas with some mucous tissue, but in addition ramifying strands of spindle and oval cells. (c) Other lobules are filled with closely packed cellular masses resembling sarcoma, except that there is little or no connective tissue between the cells. (d) Areas which bear a strong resemblance to atrophied parotid gland with a few excretory ducts lined with cylindrical epithelium scattered through the mass of alveoli. Some of these alveoli contain hyaline material. No carti- lage is found in the tumor and no pearls, but epithelial tubules lined with cylindrical epithelium are abundant, as is elastic tissue. (Plate VII, Fig. I, and Plate I, Fig. 2.) CASE XV.--St. Luke's, No. IoS5. Tumor removed from parotid region of a female aged forty-eight years. Nine years be- fore the patient's admission to the hospital she noticed a small nodule behind the lobe of the left ear, which gradually increased in size for five years and then began to grow more rapidly. The tumor is now about the size of a small orange. No pain, no gen- eral symptoms, no loss of flesh have been noticed. The tumor feels somewhat elastic when palpated and is slightly movable in all directions. It lies just below the lobe of the left ear and has pushed the lobe upward by its growth. At operation, the tumor was easily shelled out from the tissues of the neck. No enlarged lymph nodes were found. The tumor when removed was an 19 FRANCIS CARTER WOOD. irregularly oval mass with lobular surface and a capsule. Its dimensions were about four by five centimetres. On cross section the tissue is firm, not cystic, and contains no cartilage. In the clear, transparent stroma can be seen the opaque, yellowish areas of the cells of the parenchyma. The microscopical examination shows the growth to be composed of tubular acini lined with large round and oval cells. The lumen of the alveoli contain hyaline masses staining red with eosin and yellow with acid fuchsin. The stroma is not very cellular and shows advanced mucous degenera- tion. No cartilage and no true pearls can be found. Elastic tissue fairly abundant. No recurrence in two years. CASE XVI.-College, No. 6314. The tumor was removed from the parotid region of a male fifty-seven years of age. One year before operation he noticed a small nodule the size of a pea in front of the left ear. The mass is now the size of a hen's egg. It is movable under the skin and on the deeper tissues and the jaw. In some portions of the growth the consistence is soft and fluctuat- ing; in others, hard. The tumor was shelled out of the substance of the parotid gland without difficulty. It contained a cyst of con- siderable size. Microscopically the tumor shows a diffuse growth of oval and spindle cells arranged to form alveoli. These alveoli and branching strands of cells lie in a dense connective-tissue stroma. No cartilage is present and no epithelial structures. Tumor recurred locally in six months. No recurrence after second removal. The mass removed at the second operation includes por- tions of the muscles of the face and neck and lymph nodes from the anterior triangle. The morphology of the recurrence is dif- ferent from that of the primary growth. It resembles angio- sarcoma in the fact that the cellular masses surround the blood- vessels. There are, however, portions of the growth in which the alveolar arrangement is preserved with long branching strands of cells lying in the tissue spaces. The lymph nodes show chronic hyperplasia, but no invasion by the cells of the tumor. * . CASE XVII.-St. Luke's, No. 918. Parotid tumor. The patient was a man aged fifty-five years, who had always enjoyed good health, with the exception that five years previous to his admission to the hospital he noticed a slight swelling on the left cheek above and anterior to the parotid. At that time it was about the size of a bean. Growth has been slow except for the past year, during which the increase has been rapid. At present it is about 20 MIXED TUMCRS OF SALIVARY GLANDS. the size of a walnut and very hard. It is adherent to the skin, but movable over the deeper tissues. Tumor was easily removed at operation, and was found to lie upon the anterior border of the parotid gland. When hardened, the growth measures four by three by three and one-half centimetres. It is slightly lobular, and the cut section is pale and shows a large amount of fibrous tissue. Microscopically the growth contains a large amount of fibrous tis- sue, which is dense and contains only a very small amount of hyaline degeneration. The parenchyma is of a distinctly adenoma- tous type with papillary outgrowths into the alveoli. The papil- lary projections and the alveoli are lined with high cylindrical epithelium, and the lumina are filled with mucus staining blue. In some areas the alveoli are closely filled with cells. At the periphery of the tumor the cells form long strands lining the lymph spaces and resemble closely the so-called endothelial type. No cartilage is present and no epithelial pearls. The tumor would be classed as an adenoma from a morphological stand-point, and yet certain por- tions are of the same appearance as is seen in the endothelial tumors. No recurrence. - CASE XVIII.-College, No. 1399. The tumor was removed from a woman about forty-five years of age. One year previous some enlarged nodes had been removed from the neck just below the ear and behind (?) the sternomastoid muscle. They were con- sidered by the operator to be tuberculous, but no examination was made. One year later there was a recurrence at the same place, and a more thorough excision was carried out. The material con- sists of a few fragments of a tumor which measured about two centimetres in diameter. There is a distinct capsule on portions of the fragments. Microscopic examination shows the growth to be of a cylindromatous type, with alveoli filled with mucus and sur- rounded by a connective-tissue stroma very poor in cells. (Plate IV, Fig. I.) No recurrence after three years. CASE XIX.-College, No. 5673. The tumor was removed from a male aged sixty years. It has been present in the side of the neck just below inferior maxilla for five years, increasing slowly in size. Specimen is a small, roughly spherical, lobulated tumor, about 2.5 centimetres in diameter. The section of the fresh tumor shows it to be enclosed in a rough fibrous capsule from two to five millimetres thick. There is no evidence that the growth has extended beyond the capsule into the surrounding tissues. The 21 FRANCIS CARTER WOOD. cut section is pale, irregularly lobular, and the tissue has much the same consistence and elasticity as cartilage, though it is not trans- lucent. There are a few small hemorrhages into the centre of the tumor, but no necrosis or softening. - Microscopically the growth is composed of a rather dense fibrous stroma, in which are seen larger and smaller areas of large flat cells of an epithelioid type. These cells have large oval nuclei with a well-marked chromatin net-work. They lie in close contact with the smaller blood-vessels and capillaries, and the general alveolar arrangement is determined by this relationship. No con- nective tissue can be made out between the cells and the elastic tissue fibrillae, which are rather scant in the tumor, only surround the cell areas, and lie along the vessels, but do not penetrate be- tween the cells as in so many of the mixed tumors. (Plate III, Fig. 3.) No spine cells or pearls are present, no cartilage, embryonic, connective, or lymphoid tissue. This tumor, which was originally considered as a primary endothelioma of a lymph node, seems more properly defined as a sarcoma of the alveolar type despite the fact that there is no con- nective tissue between the cells. Ribbert has called attention to the fact that connective tissue is missing between the cells of a considerable number of the sarcomata, especially in the group of angiosarcomata with large cells. The reasons for considering the tumor as arising in a lymph node are based solely upon the opinion of the operator who removed the growth; under such conditions any small spherical tumor is liable to be classed as a lymph node. It is true that the morphology of the growth slightly resembles the endothelial hyperplasias of the spleen recently described by Bovaird, Grancher, and others, but the numerous mitotic figures which are present in the growth under consideration—four or five often being visible in a single field—would point rather to a rapidly growing malignant tumor than to a chronic hyperplasia as in the splenic tumors mentioned above. The absence of any embryonic tissues and cartilage certainly renders doubtful the possibility of the tumor being of congenital origin. No recurrence is recorded, though two years have elapsed since the operation. - A few cases similar to this tumor have been described (Puti- ata, Böttcher, Chambard, Zahn, Hoffman, Volkmann), but their morphology is not easily determined from the descriptions given. Ziegler figures a similar tumor in his text-book. 22 MIXED TUMC)RS OF SALIVARY GLANDS. CASE XX.-College, No. 273. Mixed tumor of the pharynx. No history is recorded except that the growth was a very large one, filling the cavity of the pharynx so that the point of origin could not be determined. The mass when removed was irregular in shape and measured about seven by five by four centimetres. The consistence of the tumor was soft, and areas of mucous tissue. could be distinguished by their transparent appearance. Micro- scopical examination of the growth reveals three groups of tis- sues, connective tissue with advanced mucous degeneration; fat tissue; and soft connective tissue in which lie branching strands of large flat and polygonal cells. The centres of some of the strands contain hyaline material staining red with eosin. Elastic tissue and mucous degeneration abundantly present in all portions of the tumor. The arrangement of the large epithelial-like cells is of especial interest and will be considered in detail. The walls of the long tube-like alveoli are lined with two dis- tinct layers of cells (Plate II, Fig. 2), one, the small, flat endo- thelial-like cells with deeply staining nuclei and small cell bodies such as are seen lining the tissue lymph spaces and smaller vessels; the other cells are large epithelial-like cells with large pale nuclei and a well-marked nuclear net-work. These cells form a single layer over the above-mentioned endothelial cells and are not very firmly adherent to them, for in the process of hardening a long strand of these cells can frequently be noticed to have become de- tached from the underlying layer of endothelium. (Plate II, Fig. 2.) Two interpretations are possible; first, that the cells are produced by a new growth of the underlying layer of endo- thelium; and, second, that they are tumor cells which have grown into the lymph spaces and more or less completely filled them, just as one can easily observe in the periphery of a lymph node during the early stage of the invasion by a carcinoma. (Plate III, Fig. I.) The cells of the carcinoma may be seen in the lymph spaces as a single layer of cells lying on the normal endothelium lining. - - w r - The first explanation is that given by Volkmann and those who believe in the endothelial origin of the large flat cells of these tumors. The second point of view has two facts to support it; one is that when the endothelial cells proliferate, as can be seen in various places along the wall of the alveoli, they form masses of small cells with the same deeply staining nuclei as can be seen 23. FRANCIS CARTER WOOD. lining normal lymph spaces, and these masses thrust aside the large flat cells and form small protrusions into the lumen of the alveolus. (Plate III, Fig. 2.) The second fact is that alveoli are found which can be traced for some distance as tubes lined with cells and then end in a small compact mass with closely packed, concentrically arranged cells which resemble epithelial pearls and contain spine cells. According to the second idea, then, the large flat cells are probably epithelial in origin and spread out through the tissues along preformed lymph spaces, leaving the normal en- dothelial lining intact. Taking the tumor as a whole, the cellular portion forms but a small part of the growth, the alveoli and pearls are scattered throughout a very abundant mucous and fibrous tissue which often contains fat cells. The growth is to be . interpreted, in the opinion of the writer, as a tumor arising from a congenital remnant left during the formation of the pharyngeal space, and containing epithelial cells which were destined to form glands, but the normal differentiation did not take place, and the epithelium still retained an indifferent type, with a tendency to revert to the type of squamous epithelium such as lines the pharynx. CASE XXI.—College, No. 2136. Tumor of the vault of pharynx. The growth was removed from a male aged sixty- five years. The tumor had been noticed for a year, gave few symptoms, and was the size of a hen's egg. It was removed by wire snare. The tumor is made up of anastomosing strands of cells which lie in a loose stroma. This distribution of the cells gives the section a reticulated appearance. In other portions of the growth the cells are crowded together in larger and smaller alveoli. The centres of these cell masses are degenerated, con- tain mucin and hyaline material. The connective tissue is also the site of hyaline degeneration. No record of recurrence. No cartilage; no epithelial pearls. - CASE XXII.-College, No. 2079. The tumor was removed from a female thirty years of age. It was encapsulated and oval in form, measuring four by three by three centimetres. It had been present in the soft palate for two years. The cross-section of the growth shows a fine, even surface, with a few trabeculae crossing it and small cartilaginous areas scattered through of irregular size; the largest, perhaps, three millimetres. Microscopically the tumor is composed of a diffuse cellular 24 MIXED TUMC)RS OF SALIVARY GLANDS. growth of spindle and flat cells with but little connective tissue. The cells show but little alveolar arrangement, but in a few places there are distinct alveoli with high cylindrical epithelium lining the walls. The cartilage is in small amount and contains a good deal of elastic tissue. No epithelial pearls. CASE XXIII.-College, No. 7315. The tumor was removed from a female twenty-six years old. It was situated on the roof of the mouth near the median line and just in front of the folds of the soft palate. The size is roughly two by two by one centi- metre. The surface is smooth and slightly lobular. The mor- phology is endothelial in type, with rather abundant cells and only a moderate amount of stroma. No cartilage, but abundant and elastic tissue. - - CASE XXIV.-College, No. Io,042. Recurrence in a year. Tumor had been noticed by the patient for only ten days before its second removal, so that the growth must have been slow. Its dimensions are two by two by five centimetres. Sections show the same morphology as before. The strands of cell masses have invaded some of the mucous glands of the soft palate and have produced a very curious picture; the tumor cells forming alveoli which lie in close contact with those of the gland, and are at times difficult to distinguish from the latter, the resemblance is so close. No cartilage in the recurrence. Elastic tissue very abundant, outlining the tumor alveoli and forming a dense net- work throughout the whole growth, though there is less than in the primary. CASE XXV.-The specimens were removed from a woman about forty-five years of age, who had had a submaxillary new growth in the side of the neck some twenty years before. This was imperfectly removed, and numerous recurrences have taken place which have necessitated operative interference almost every year since. The growth still remains confined to the lateral aspect of the neck, and the general condition of the woman is good. The specimens under consideration represent the last three removals. The numerous fragments show traces of a cap- sule surrounding them and on cross section a uniform pale cellu- lar surface. The microscopic picture varies somewhat in the earlier and later recurrences. In the earlier ones the type is still distinctly that which has been considered endothelial with anas- tomosing strands of cells lying in a connective-tisssue stroma. 25 FRANCIS CARTER WOOD. The later recurrences have lost their characteristic morphology and resemble to a certain extent the true sarcomata with flattened cells. There still remains, however, something in the aspect of these sections under the microscope that is suggestive to any one who has seen a number of the mixed tumors, something in the shape and even staining of the cell nuclei and the arrangement of the cells that differs from the ordinary picture of sarcoma. A few spindle-shaped connective-tissue cells lie between the masses of tumor cells, and these have been stained and figured by a number of observers, notably Barth, as an evidence of con- nective tissue between the cells of the mixed tumors. The case is chiefly interesting because of the large number of recurrences and the comparatively benign character of the growth. No in- volvement of the cervical lymph nodes has taken place. (Plate VI, Figs. I, 2, and 3.) CASE XXVI.-Old No. 1598. The patient was a male aged forty-four years, who had a small nodule on the upper lip for two years. The tumor was removed and recurred in two years. The tumor under consideration is this recurrence. It measures about one centimetre in diameter. It lies in the tissues of the lip near the surface, but covered by the superficial epithelium. The sebaceous glands near the surface are compressed and lie flat- tened out on the capsule of the tumor. The latter does not in- filtrate the deeper tissues. It is composed of small nodules of mucous tissue and hyaline cartilage. The mucous-tissue areas contain branching strands of cells and alveoli. The latter are in general lined with flattened cells, but in a few areas the cells are high and cylindrical, with the nuclei near the basement mem- brane. The cartilage is hyaline. Elastic tissue is abundantly present in the tumor. CASE XXVII.-College, No. 2009. The tumor is recorded as having been present for a number of years in the upper lip. It was neatly encapsulated, and was shelled out of its bed by incising the mucous membrane of the inner surface of the lip. The tumor as removed is about the size of a bean, oval in form, and smooth of surface. It contains no visible areas of cartilage. No record has been preserved of a recurrence. Microscopically the tumor is composed of anastomosing strands of cells in a soft cellular stroma, the latter containing areas of mucous degenera- tion. Large alveoli filled with hyaline material are present 26 MIXED TUMC)RS OF SALIVARY GLANDS. in portions of the growth. No pearls or distinct epithelial structures are present. A moderate amount of elastic tissue is present. - CASE XXVIII.-College, No. 24II. Female aged thirty- four years. The tumor was first noticed when patient was thirty- two years of age. It was under the ramus of the jaw on the left side. Removed after six months’ fairly rapid growth. Re- currence after ten months. Second recurrence after eight months. Third recurrence after six months, involving all the left parotid region and the deeper tissues of the neck. Fourth recurrence six months later. Lymph nodes not invaded. Only portions of the various tumors have been preserved. The microscopic mor- phology is the same in all. The parenchyma is composed of oval or spindle cells with an alveolar arrangement, the alveoli often containing hyaline masses. The stroma shows some mucous de- generation. In the later recurrences the alveolar arrangement is less marked, and at first glance the morphology is suggestive of sarcoma, but occasional areas still show traces of the alveolar types. No cartilage in the growth, and no epithelial structures. CASE XXIX.-College, No. 6560. The tumor was removed from the left parotid region of a male aged thirty-one years. It had been present for four years. During the last year growth had been rapid. The tumor is now the size of a small orange, is elastic and movable on the deeper tissues and under the skin. Easily removed at operation by shelling out the encapsulated tumor from the parotid substance. Microscopically the growth consists chiefly of dense connective tissue containing a few spindle cells with marked hyaline degeneration of the stroma. In a few areas there are masses of spindle and stellate cells lying in soft connective tissue which has undergone mucous degenera- tion. Towards the periphery of the growth, close under the fibrous capsule, the morphology is that of the endothelial type, that is, alveoli and a few solid strands of cells. Alveoli contain hyaline masses. No pearls or flat epithelium are present in the tumor, but there are a few ducts lined with high cylindrical epithelium and containing masses of flat cells evidently derived from the degeneration of the lining epithelium. No cartilage is present and no lymphoid tissue, but there are a few areas of fat tissue. The tumor has not recurred in the two years since the operation. - - - * . 27 FRANCIS CARTER WOOD. CASE XXX.-College, No. 7479. The tumor was removed from the right side of the neck of a male patient aged sixty-five years. For two years a mass had been noticed protruding into the pharynx just in front of tonsil. It has slowly grown during that time, and now projects more externally than internally, and at about the angle of the jaw. The tumor was movable in the deeper tissues and of oval form, about the size of a hen's egg. When removed, the growth was found to be a lobular encapsu- lated mass with an area of softening in the centre. Sections of the growth show it to be a typical tumor of the endothelial type with a great deal of mucous degeneration of the connective-tissue stroma. No cartilage or epithelial structures can be found and no lymphoid tissue. The absence of the lymphoid tissue shows that this tumor is not derived from one of the branchial clefts, though its softened centre and position were suggestive. No recurrence of the growth in two years. CASE XXXI.—College, No. 214.I. Submaxillary tumor. The tumor was removed from a female aged forty-five years. The tumor was first noticed as a growth under the body of the left inferior maxilla some nine years before. The growth was slow. The tumor measures three by four by three centimetres. Attached to it is a large mass of the submaxillary gland. It is partially encapsulated. Microscopically the growth is an ex- quisite example of what has been designated cylindroma. The tumor is composed of alveoli, which may be either solid or par- tially distended with mucus or wholly distended, so that the cells form a flattened ring at the periphery. The most frequent con- dition is, however, a partial distention, with many small spherical areas of mucus in each alveolus. This gives the tumor its peculiar morphology. (Plate VII, Fig. 2.) No cartilage or definite epi- thelial structures are present. The submaxillary gland has been invaded by the tumor, and the compressed alveoli of the gland are with difficulty distinguished from the alveoli of the tumor. The tumor recurred locally in two years, but after a second removal there has been no recurrence in three and a half years. CASE XXXII.-College, No. 9565. Tumor from submaxil- lary region. The growth is hard, lobular, and roughly spherical. Its greatest diameter is about four centimetres. Cross section is pale, smooth, and rather homogeneous. The capsule is very thin. No remnants of submaxillary tissue present. No macroscopic 28 MIXED TUMoRS OF SALIVARY GLANDS. areas of cartilage. Microscopically small masses of cartilage are fairly abundant, and also a good deal of myxomatous con- nective tissue. The main portion of the tumor follows the con- ventional endothelial type, but scattered throughout the whole growth are a considerable number of epithelial pearls. These are either quite discrete or are in intimate connection with the tubules and strands of the endothelial portions. Spine cells can be seen in a few of the pearls, especially those in the myxomatous tissue (Plate IV, Fig. I.) Elastic tissue is very abundant. No recurrence in six months. This case resembles the one described by Wilms, in which he showed that epithelial structures are not confined to the parotid group of tumors as Hinsberg had thought. This observation renders doubtful Hinsberg's supposition that the epithelial structures might be derived from the epithelial anlagen destined to form the ear-drum. CASE XXXIII.-No record is preserved of this tumor, the fragments of which are in the collection of the Department of Pathology, College of Physicians and Surgeons. The label records only that the tumor is a mixed enchondroma of the parotid. Sections from the fragments of tumor which remain show that the growth is in the main of an alveolar type with scattered areas of cartilage and mucous degeneration. The alveoli are lined with flat cells and contain masses of hyaline material. In several places in the sections well-formed pearls can be seen with epithelial spine cells. Elastic tissue is very abundant throughout the growth, especially about the alveoli. CASE XXXIV-V. C. 9. Tumor of parotid. No history is recorded of this specimen. The material consists of a few blocks of tissue evidently from a tumor of six or eight centimetres in diameter. Some of the fragments show a distinct capsule. Micro- scopically the tumor is divided by trabeculae into alveoli of irregu- lar size. These alveoli are lined and filled by a cellular mass which also forms alveoli. These smaller alveoli are often surrounded by a single layer of cells and contain masses of mucus which stain deep blue with haematoxylin. No cartilage is present and no great amount of elastic tissue. At one side of the growth is a remnant of much compressed parotid gland, but no invasion of the gland tissue by the tumor has occurred. (Plate VII, Fig. 3.) CASE XXXV.-The tumor was removed from the outer sur- face of the lower lip of a male patient of thirty-three years. The 29 FRANCIS CARTER wooD. growth was oval in form and measured two by one and one-half by one and one-half centimetres. It had been noticed as a small mass about the size of a pea since childhood, and had grown slowly ever since. The only discomfort noticeable was due to the mechanical interference with swallowing and speaking. An ulcerated surface had been noticed for the past month on the most prominent portion of the tumor. The tumor was movable in the tissues of the lip but quite closely adherent to the skin, so that it was impossible to shell out the growth. Microscopic examination of the tumor shows it to be of the endothelial type with a large amount of mucous degeneration of the connective- tissue stroma. Very few alveoli are to be seen, the cells lying, as a rule, in anastomosing strands embedded in the mucous tissue. There is no cartilage present and no epithelial structures. Elastic tissue is very abundant. About the periphery of the growth, but external to the capsule, are many mucous glands, in some of which the alveoli have become atrophied by pressure, in others are still normal. The tumor has not recurred in six years. CASE XXXVI.-Old Nos. 232, 24I, 242. The material con- sists of a tumor from the parotid region with the fragments re- moved from a series of five recurrences, extending over some five years' time. There is no other history connected with the speci- mens. The primary growth is of the conventional endothelial type with abundant strands of cells occasionally forming alveoli which are filled with hyaline material. No cartilage is present, and only a moderate amount of elastic tissue. The various recurrences are of interest only as they show the gradual tendency of the growth to assume a sarcomatous type with only occasional suggestions of the alveolar structure of the original. In the last recurrence, however, the morphology is quite markedly alveolar, though not quite as evident as in the original. The case is of interest only to illustrate the possibilities of frequent local recurrences without involvement of the internal organs or the neighboring lymph nodes. & CASE XXXVII.-Old No. 465. The tumor was removed from a male thirty-eight years of age. He had had a small parotid tumor for fifteen years. For the last six months the tumor has grown rapidly and is now of oval form, measuring six by four by four centimetres. There is a thick, rough, fibrous capsule which surrounds the periphery of the tumor, and a mass of the 30 MIXED TUMC)RS OF SALIVARY GLANDS. parotid gland is embedded on the internal surface of the growth. The cut surface is divided by trabeculae with numerous lobules measuring from five to fifteen millimetres in diameter. No car- tilage can be seen. Microscopically the tumor is cellular and follows generally the cylindromatous type, though a few areas are composed of solid masses of cells. Very little fat is present in the tumor and but little mucous or hyaline degeneration. Elastic tissue is also scarce. The parotid is normal. No pearls or epithe- lial structures are to be seen. - CASE XXXVIII.-The tumor was from a male of about thirty years of age, and had been present on the inner aspect of the upper lip in the median line for many years. It was easily shelled out by incising the mucous membrane over it. When removed it was an encapsulated mass of oval form with a diame- ter of eight millimetres. Sections show the tumor to be com- posed of branching strands of oval cells in a stroma of soft con- nective tissue with much mucous degeneration. No epithelial structures, no cartilage, but a good deal of elastic tissue is present. CASE XXXIX.-Old No. 573. An adenoma of the parotid gland. No history has been preserved of the tumor and the entire specimen consists of four fragments each about two centimetres square and five millimetres thick. The microscopical appearance of the growth is that of a true adenoma or adenocarcinoma of the parotid. It is composed of a dense fibrous stroma, the trabeculae of which outline areas of cell alveoli. These alveoli are lined with one or two layers of cylindrical cells with large oval nuclei and a well-marked chromatin net-work, differing in this latter detail from the so-called endothelial cells. The alveoli, as a rule, contain no secretion. Remains of the ducts of the parotid are scattered through the tumor, and at one portion of the section normal parotid gland tissue can be found in close contact with the tumor cells. Another portion of the tumor shows the morphology of what has been called the endothelial type. There are long tubular alveoli lined with flattened cells; smaller alveoli contain- ing hyaline secretion, and also alveoli containing double rows of lining cells such as have been figured in Plate II, Fig. 2. No car- tilage and no pearls are to be found. Elastic tissue is fairly abun- dant. The case is reported here merely to show that morphology counts for little in the final decision as to the histogenesis of the cells of the salivary tumors, because in a growth evidently epithe- 31 FRANCIS CARTER WOOD. lial in nature, areas can be found in which the morphology alone would lead one to consider the tumor as endothelial, just as in many cases of the small congenital epithelial tumors of the skin a cylindromatous arrangement of the tumor cells is not infrequent. CASE XL.-Old No. 639. The tumor is from a woman aged fifty-eight years. Eighteen years before the present operation, a Small growth over the parotid was removed by the action of caustic. The tumor has been growing ever since. The growth is of oval form with the inner surface flattened, and measures eight by seven by five centimetres. It is encapsulated on its outer sur- face and somewhat lobular. The inner aspect is rough from the severing of adhesions to the deeper tissues. The cut section is fine in texture and pale with transparent areas of mucous tissue and cartilage. There are a few haemorrhages into the substance of the growth. No well-marked trabeculae are present. Micro- scopically the growth is of the cylindromatous type with few cysts. The fibrous tissue is rather dense and the cells few and spindle in form. A few areas of hyaline cartilage are to be seen, and in Some places a large amount of mucous degeneration has taken place in the connective tissue. In one portion of the growth are a few alveoli lined with high cylindrical epithelium, but these are all confined to a limited area and are not generally found. There is also an area which resembles a cellular sarcoma with cells closely packed about the vessels. The tumor contains a moderate amount of fat and is abundantly supplied with elastic tissue. There is no history of a recurrence after removal. . CASE XLI.-St. Luke's, No. 696. Recurrence after removal of a parotid tumor. The patient was a male from whose parotid region a large mixed tumor had been removed six months before. No local recurrence was noticed after the operation, but the patient gave symptoms of pulmonary and other metastases. A small subcutaneous nodule was removed from the side and showed a large-celled sarcoma with a tendency to alveolar arrangement of the cells. No further record of the case has been obtainable. The morphology of the growth is like that of the following case. CASE XLII.-St. Luke's, No. 1590. Recurrent sarcoma of parotid. The patient was a male twenty-eight years of age, who had had a tumor removed from the parotid region four months before. The original growth had been noticed as a small nodule for several years. Several months before its removal the growth 32 CASE XLII.-Recurrent sarcoma of parotid, MIXED TUMCRS OF SALIVARY GLANDS. had become more rapid. No sections of the original tumor were obtainable, but the patient said the tumor had been hard and easily movable. The recurrence is a conical mass projecting from the surface of the left parotid gland. It is about four centimetres high and six centimetres in diameter at the base. The skin over the tumor is reddish, adherent, and the superficial veins are well marked. The growth is fixed to the parotid. On section, the tumor is firm, except for one area of softening, and contains no cartilage. Microscopic examination of the growth showed it to be a large polyhedral celled sarcoma of the conventional type, with a more or less marked alveolar arrangement of the tumor cells. The only noteworthy features are the enormous number of mitotic figures present in the cells of the tumor and the slight amount of invasion of the parotid tissue. The parotid glandular tissue is atrophied in many places and replaced by connective tissue in which is a large amount of mucous degeneration. The mor- phology of the tumor is the same as in the recurrent Case XLI. The patient died from operation shock. CASE XLIII.-Old No. 760. The specimen is a tumor of three years’ duration, which is situated at the angle of the jaw superficial to the deep cervical fascia and quite movable under the skin and on the deeper tissues. It is oval, measuring about three by three by four centimetres, and smoothly encapsulated. The cut section shows a uniform pale cellular surface with small islands of hyaline cartilage scattered through it. The microscopic examination of the growth shows it to be composed largely of a cartilaginous matrix in which lie numerous strands of cells and alveoli. The alveolar walls are lined in many places with high cylindrical epithelium, and a few areas are present in which faint intercellular bridges can be made out. No pearls are present. There is a very abundant elastic tissue net- work in the cartilage. The fibres also pass between the cell masses and form walls about the alveoli. This large amount of elastic tissue seems to be rather characteristic of these tumors, for Spuler has not found much elastica in the chondromata. CASE XLIV.-Old No. 1400. Growth was removed from a male aged thirty years, after a slow growth of two years. The tumor was situated at the angle of the jaw, lying on the parotid gland. It measures after hardening four by five and one-half 33 FRANCIS CARTER WOOD. by three and one-half centimetres. There is a fairly complete capsule over the various portions of the tumor, and this capsule sends off trabeculae which divide the tumor into a series of lobules resembling those seen in a section of the pancreas. Small areas of cartilage can be seen in one portion of the growth. The microscopical examination shows the tumor to be of the ordinary morphology generally known as endothelial with alveoli filled with hyaline material and walled with flattened cells. There are also long branching strands of cells of an epithelial type which form a meshwork throughout the soft fibrous stroma. In places the solid strands form pearl-like structures with concentric layers of closely packed cells which stain strongly with eosin and show no nuclear structure. At the periphery of these pearls and in some of the strands of flat cells intracellular bridges and fibres can be made out. A small fragment of compressed parotid and some skin are to be seen in some of the sections. Neither tissue is invaded by the growth, but both lie outside the fibrous tissue capsule of the tumor. The tumor is evidently a mixed tumor containing both epiblastic and mesoblastic tissues. No record of patient's further condition. - CASE XLV.-Old No. 1679. The specimen was removed from a male of sixty-three years. The tumor had grown for a year from the surface of the superior maxilla. No other history is recorded. The material as preserved in the collection of the Pathological Department is composed of a number of irregular masses which altogether form a tumor the size of a small lemon. The skin is adherent to the external surface of the growth, but is not invaded by the cells of the neoplasm. The cut surface is smooth and of fine texture. No cartilage can be seen. A thin capsule surrounds the peripheral portions of the fragments. Microscopically the tumor contains three distinct structures,-a cylindromatous type of growth, a simple tubular endothelioma, and a form in which closely packed cell areas are surrounded by septa of connective tissue which have undergone hyaline degenera- tion. (Plate VIII, Fig. I.) There is a moderate amount of elastic tissue present, but no cartilage and no epithelial structures. CASE XLVI.-College of Physicians and Surgeons, Old Series No. 1727. The tumor was removed from the parotid region of a man aged fifty-seven years. No history has been recorded except that the tumor had been present for a number of 34 MIXED TUMCRS OF SALIVARY GLANDS. years. The material which had been preserved consists of numer- ous fragments which together measure some six by five by four centimetres. The outer surface of the tumor is covered with nodular elevations. The whole is surrounded by a fibrous tissue capsule. Portions of the parotid gland are adherent to the periphery of the growth. The cut section is opaque and uniform in appearance except for the presence of small areas of softening and haemorrhage. No cartilage is visible. Microscopically the growth resembles a large-celled sarcoma, but at the peripheral portions of Some of the cell masses an alveolar arrangement can be worked out. The alveoli are filled with mucus which takes a blue color with haematoxylin. There is no connective tissue be- tween the individual cells. Elastic tissue is not abundant. No cartilage is present. No invasion of neighboring lymph nodes. The parotid gland is not involved. It is difficult to exactly class this tumor, for which the original diagnosis was adenosarcoma. Some observers might with perfect right insist that the mor- phology more closely resembles the sarcomatous type than that called endothelial. The vessels are more numerous than in the purely endothelial tumors, and in certain areas the cells are in very close relation to the vascular endothelium, giving an appear- ance such as is found in the angiosarcomata. CASE XLVII.-Old No. 1728. The specimen is a tumor from the parotid region. No other record has been preserved. The material consists of a few blocks of tissue evidently removed from a large mass. Microscopically the growth is largely com- posed of myxomatous tissue with strands of cells lying in it. These cells form alveoli containing hyaline material. Small areas of cartilage are scattered throughout, not in the form of clean cut nodules, but rather in diffuse masses with large oval cells, shading off gradually into the myxomatous areas with spindle cells. The transition is very gradual, and it is often impossible to decide just where the point of separation occurs. In portions of the growth the stroma cells are arranged in well-formed alveoli with high cylindrical cells lining the walls. No pearls are present. Elastic tissue is very abundant. CASE XLVIII.-Old No. 1844. The patient was a female fifty years old. Three and one-half years previous to operation a swelling appeared in front and below the right ear. The tumor has grown rapidly and is now of large size. It is slightly 35 FRANCIS CARTER WOOD. movable on the deeper parts. The skin is red and adherent. The tumor is oval in form and measures thirteen by nine and a half centimetres. It is rather soft and elastic to palpation. Macro- scopically the cut surface of the growth is of fine texture with numerous clear areas of mucous degeneration. There are faintly marked fibrous tissue trabeculae throughout the tumor, and it is surrounded on its outer aspect by a thin capsule. Microscopically the growth is composed chiefly of soft fibrous tissue, which is the site of advanced mucous degeneration and closely packed cells of spindle and oval form resembling sarcoma. In portions of the growth where the fibrous tissue is more dense, however, the cells are arranged in alveoli and strands as in the endothelial type. No pearls are to be found. Elastic tissue is abundant. No record of a recurrence. CASE XLIX.—Old No. 1858. The tumor was removed from the submaxillary region of a male aged forty, after a slow growth of seven years. The tumor measures seven by five by five centi- metres and weighs IIo grammes. It is encapsulated. The sur- face is lobular. Trabeculae run into the growth from the capsule and divide it into numerous secondary lobules which are filled with soft cellular masses cutting like cheese. The color of the cellular areas is pale, and there are numerous transparent spots due to mucous degeneration. Microscopically the growth varies a good deal in different portions. Some of the areas resemble an angiosarcoma, except that the cells do not lie in close contact with the vessels, but are separated from them by a thick vascular wall. In other areas the morphology is cylindromatous, with long anas- tomosing strands of cells separated from each other by connective tissue altered by advanced hyaline degeneration. In certain por- tions there is hyaline degeneration of the connective tissue with mucous degeneration of the cells bordering the degenerated con- nective tissue. The cells over large areas are replaced by mucous secretion. No definite epithelial structures can be found. Elastic tissue is abundant. No cartilage is present. The further history of the case is not know. - - CASE L.—The tumor was removed from the left parotid region of a male of twenty-seven years. He had had a small movable tumor in this region since childhood. Growth has been very slow and constant since that time. The tumor measures about two centimetres in diameter, is easily movable in all direc- 36 CAsk L.-Sketch from a photograph of a patient with a parotid tumor and congenital malformations of the ear and face. MIXED TUMCRS OF SALIVARY GLANDS. tions, and is very hard. About two centimetres posterior to the tumor is a small pedunculated growth which was congenital. It is roughly one centimetre long and one-half centimetre in diame- ter. The form is cylindrical and the surface is covered with fine hair. The growth is attached by a pedicle to the deep fascia. The ear also shows an auricular fistula, which was noticed at the time of birth. The tumor over the parotid is lobular and encap- sulated; its cut section is white and close grained, with a few Small islands of cartilage scattered irregularly throughout. Micro- scopically the tumor is of the conventional endothelial type with islands of cartilage and much mucous degeneration. There is nothing peculiar about the growth except that there are numerous pearls in the more cellular portions. These pearls are formed of masses of flattened cells with a central area in which no structure can be made out. At the periphery of the pearls the epithelial cells show characteristic fibrillations and intracellular bridges. Keratohyalin granules can be seen in a few cells. In the neigh- borhood of the pearl containing tissue are also a few cysts and branching tubules lined with a single layer of flat cells which stain deeply with eosin. These cells are of the same appearance as the epithelial cells forming the periphery of the pearls, and though no fibrillae can be made out, they resemble in morphology the other epithelial cells in the sections. Elastic tissue is abundant in the tumor. - - The small pedunculated mass posterior to the tumor contains a spur of fibro-elastic cartilage, and seems most easily explained as an accessory tragus. This assumption is rendered more prob- able by the presence of another congenital malformation, the auricular fistula. The case is of especial interest because of the connection of a tumor containing embryonal tissues with other congenital malformations of the facial region, as evidenced by the accessory tragus and auricular fistula. No recurrence has taken place in two years. - - CASE LI.-Old No. 1959. Tumor removed from side of neck. No other history concerning the specimen has been recorded. The portions of the tumor which have been preserved are two flat sections measuring some three by five centimetres and one to five centimetres thick. The external surfaces are lobular and encapsulated. The cut surface is smooth and compact and contains scattered areas of cartilage. Microscopically the growth 37 FRANCIS CARTER WOOD. is of the endothelial type with strands and alveoli and cartilaginous areas merging imperceptibly into the cellular masses. There is but little mucous degeneration. Well-formed pearls are present formed of flattened cells at the outer portions, and in the centre a mass of degenerated cells which stain red with eosin. In the peripheral portions of the pearls the cells can be seen to be con- nected by intercellular bridges, the fibres from which pass over into the cell bodies. The pearl-forming cells are in contact with those of the solid strands and alveoli of the endothelial type and continuous with the latter. The pearls do not occupy any limited area, but are scattered throughout the more cellular portions of the growth. Elastic tissue is very abundant in all portions of the growth. - CASE LII.—College, No. 2017. The tumor was removed from the parotid region of a female aged twenty-four years. The growth had been noticed for two years previous. The tumor is a lobular, roughly spherical mass, two centimetres in diameter. It is encapsulated. Microscopically the tumor is composed of branching strands of cells lying in a matrix of connective tissue, much of which has undergone mucous degeneration. No cartilage nor evident epithelial structures are present. Elastic tissue is moderately abundant. No recurrence. CASE LIII.-College, No. 2397. The patient was a female fifty years of age. The tumor was situated in the parotid region and had been noticed for one year. The growth is a pear-shaped, encapsulated mass, measuring seven by six by four centimetres. It is surrounded by a dense fibrous capsule. Microscopically the growth is of a mixed type, showing solid masses of cells in cer- tain areas and alveoli in others. Some of the alveoli are lined with high cylindrical epithelium. Elastic tissue is abundant. No in- vasion of the lymph nodes. No cartilage present; has not recurred. - CASE LIV.-No record is preserved of this tumor. The specimen is designated as a myxochondroma of the parotid. The tumor measures three by three by five centimetres and is encap- sulated. Ragged fragments of the parotid gland are attached to the periphery of the growth together with portions of the masseter muscle. One pole of the oval tumor is firm and shows a smooth homogeneous cut surface. The other pole has softened and broken down to form an irregular cyst cavity. The capsule is 38 MIXED TUMCRS OF SALIVARY GLANDS. Some three millimetres thick and composed of dense fibrous tissue. Microscopically the growth is largely composed of mucous tissue with many spider cells. Scattered through this tissue are long anastomosing strands of cells which occasionally form alveoli. Parotid tissue is compressed by the tumor and somewhat atro- phied, but the tumor structures have not invaded the salivary gland. Small islands of cartilage are scattered through the growth and occasional epithelial pearls can be seen. CASE LV.—The original tumor was removed from a man of fifty-five years of age. It was on the left side of the neck just under the body of the inferior maxilla. The growth had been noted for about two years. It was adherent to the deeper tissues and partly encapsulated. Microscopically it was an exceedingly cellular growth with branching strands of cells, some of which form alveoli with hyaline contents. No cartilage is present and no epithelial structures; elastic tissue is abundant. Recurrence in three months with invasion of the tissues of the neck and the periosteum of the inferior maxilla. The recurrence retains to a certain extent the alveolar arrangement of the primary growth, but is more diffuse and cellular, and invades the muscle and con- nective tissues of the neck. No invasion of the regional lymph nodes. The patient died a few months after the second operation. CASE LVI.-The tumor is in the collection of the Patho- logical Department, labelled myxosarcoma of parotid. No other record has been preserved. The growth is an oval encapsulated mass measuring about seven by five by four centimetres. In the centre is a ragged cavity forſhed by necrosis of the tumor cells. The tumor is divided into large alveoli by trabeculae of connective tissue. The contents of the alveoli are translucent and soft. These areas are seen to be made up of mucous tissue when examined microscopically. In this mucous tissue lie strands of cells occa- sionally arrranged to form alveoli which contain hyaline material. No cartilage or epithelial structures are present. Elastic tissue is very abundant. CASE LVII.-St. Luke's, No. 1792. The tumor was removed from left submaxillary region of a young woman of twenty-eight years. She had noticed the tumor for four years, and thought that the growth had veen very slow until three months before the operation. It is now the size of an English walnut, and is freely movable in the lax tissues of the neck. The tumor when removed 39 FRANCIS CARTER WOOD. and hardened measured three by three and one-half by two centi- metres. The surface was smooth but slightly lobular, and the tumor was inclosed in a thin capsule. Cross-section of the growth shows a firm white mass without visible trabeculae. Microscopi- cally the tumor is composed of interlacing strands of cells with an occasional dilatation forming an alveolus which is often filled with hyaline material. No cartilage and very little degenera- tion are present. No epithelial structures. No recurrence in six months. - . CASE LVIII.-A small tumor removed from right submaxil- lary region of a woman of thirty years. No history was obtained, except that no recurrence has taken place in the two years since the operation. The tumor measures four by three by two centi- metres. The surface is rough and covered with fragments of the submaxillary gland, and fibrous tissue. On section the surface is very pale and translucent. Some of the more cellular areas are yellow and opaque from the fatty degeneration of the cells. There are a few small cysts, none over three millimetres in diameter. No cartilage is visible. Microscopically the growth is of the ordinary endothelial type, with no peculiar characteristics except that scat- tered through the tumor are well-defined alveoli lined with high cylindrical epithelium which is often fatty. These alveoli bear a striking resemblance to the atrophic ducts of the submaxillary tissue which are adherent to the periphery of the tumor. At one portion of the growth near the periphery the alveoli are filled with flat epithelial cells which are arranged to form imperfect pearls. No cartilage is present in the growth, and very little mucous degeneration of the connective tissue is to be found. Fragments of this tumor were examined for glycogen both before and after hardening, but none could be demonstrated. Elastic tissue abun- dant. (Plate VII, Fig. I.) Some areas of the tumor show ex- cellently the production of the morphological appearances, which have been called cylindromatous by the softening of the connective- tissue trabeculae, many of which have undergone hyaline degen- eration. (Plate VII, Fig. 2.) CASE LIX-The specimen has no record except that it was removed from the parotid gland. The amount of material in- dicates that the tumor must have been of the size of a man’s fist. The larger portion of the growth is well encapsulated and endothelial in type. No cartilage is present. Other portions 40 PLATE [. Fig. 1.-Case VI. Endothelial portion of the tumor from the cheek which contained epithelial pearls. Fig. 2.-Case VI. Epithelial tubules, cartilage, and alveoli of the so-called endothelial type. PLATE II. º- * tº Nº._\º § - - º - º Q º º € * º -- º - ºxº - - - - º º gº *** FIG. I.-Case VI. Endothelial type of alveoli with flat cells lining the lymph spaces and spreading out into the surrounding connective tissues. Fig. 2.-Case XX. Invasion of lymph spaces in a pharyngeal tumor by epithelial cells which leave the endothelial lining of the lymph space in its normal condition. PLATE III. FIG. I.-Lymph channel from the periphery of an axillary lymph node invaded by the cells - from a carcinoma of the breast. - - - - - - - - ----- - - - - --- - - - - - - - - - - -- - -- - FIG. 2.-Case XX. A lymph space from tumor of the pharynx showing the proliferation of the endothelium displacing the epithelial cells of the tumor from their position on the walls of the space. FIG, 3–Case XIX. Section showing one of the alveoli and the close relationship of the cells to the blood-vessels, PLATE IV. r A. --- º Cº. #3. º º - º : º, º * \tº º, sº º & FIG. I.-Cases VI and XXXII. Nest of epithelial cells in embryonic gelatinous tissue. 3 º i. º- A. * - i. º "ºº- *, - * º, * * - * §º ; § § ** - § º º Fig. 2–Case XVIII. Cylindromatous type of tumor from the neck. PLATE V. FIG. I.-Case VI. Epithelial pearls in a tumor from the cheek near the parotid gland. Other portions of the growth are of the endothelial type. C - - º- ºº ºº * º - 2. º, º, sº . - º sº ºS- sº sº - º º - - - º º - * - - - - Fic. 2.-Case VI. Detail from one of the epithelial alveoli showing prickle cells and fibrillae in the epithelial cells. Kromayer's modification of Weigert's fibrin stain. PLATE VI. --~~~~) -- º - º º º - º ºgº º º º: §: ' , ſº º * ... ººs & º - Q º gº º º, . Pºº º “ Aº 2%,” º º sº º - - 0. - g º º º ſº - gºogº 2 tº gº º º, º żº - - º º --> --> *. ef * * º FIG. I.-Case XXV. First recurrence showing alveolar arrangement of the cells of the tumor. Yºººººº. *-sºº --- º” º **ś Sºzº - º “… º. Sº - º º º sº º sºº" - Sºº º - º, a gºod I --- º º -> *Cº. ". º *** Tº A 3. sº wrºsº, - sº e.º.º.º. eºsº | º ºea’s Q º a i sº - º - º --~~~~ FIG. 2.-Case XXV. Second recurrence. Fig. 3.-Case XXV. Third recurrence with approach to an indifferent type resembling sarcoma. Small connective-tissue spindle-cells, can be seen lying between the cells of the tumor. PLATE VII. FIG. I.-Case LVIII. Elastic tissue net-work stained by Weigert's method. ºº: 2%) tºº º º ºf ºa º ż. § ºft - §§ º *º & - - %. º º º Öğ) §º º Óð 33 º, as 2.3% º Qāºšº Çº sº FIG. 2.-Case LVIII. Alveoli filled with colloid FIG. 3.-Case XXXIV. Alveoli filled with mucus. material. PLATE VIII. Fig. 1.-Case XLV. In this portion of the growth the parenchyma is more abundant than the stroma, which remains as a small hyaline layer along the course of the blood- vessels. Fig. 2.-Case LIX. Portion of tumor with a large amount of mucous degeneration in the stroma. MIXED TUMCRS OF SALIVARY GLANDS. show a peculiar alveolar structure resembling a gelatinous carci- noma which appears to be due to a large production of mucus crowding the cells of the tumor into strands. No flat epithelial cells with spines were found in the growth, which is chiefly interesting for its peculiar morphology; the transition between the endothelial portions and those resembling carcinoma being quite gradual. (Plate VIII, Fig. 2.) SUMMARY OF THE CLINICAL COURSE AND GENERAL MORPHOLOGY OF THE TUMCRS. Before summarizing the details of the results of the study of the cases just given, it may be well to note the relative fre- quency of occurrence of tumors in the salivary glands. Böhme, in a collection of a large number of tumors from this region, found that they occur in the parotid ten times more frequently than in the submaxillary, while the sublingual gland is the site of tumor formation only in I per cent. of the whole number examined. Küttner also gives the proportion as eleven to one. This proportion is not to be observed in my own series, for, taking those in which the source of the tumor is definitely recorded, the parotid growths number twenty-six, while the submaxillary number thirteen, or exactly half as many. That this is probably due to the fact that in preserving these tumors the rarer submaxillary growths were more often saved than the more abundant parotid growths is evident from the Patho- logical Records of St. Luke's Hospital, where the parotid growths operated upon are five times as numerous as the Sub- maxillary. Of the fifty-nine tumors recorded in this paper only four are from the lip, four from the pharynx or palate, two from the neck, and one from the cheek. Even this cheek tumor was in contact with Steno's duct, and should probably be considered as connected genetically with the parotid. If we consider the tumors from the point of view of the frequency of the mixed growths as compared to the other neo- plasms of the salivary glands, we find that out of the fifty-nine cases one may possibly be considered as an adenoma and two as sarcomata, a third is a fragment of embryonic connective 41 FRANCIS CARTER WOOD. tissue, fat, and blood-vessels, the remainder may be considered as belonging to the mixed tumors or the so-called endothelio- mata. The author has in his possession three carcinomata and One pure adenoma, while a pure myxoma, a rhabdomyoma, and a round-cell sarcoma are in the collection of the Pathologi- cal Department. Evidently the mixed tumors form a large proportion of the tumors arising in the salivary glands. Cases reported by other observers show the same proportion. Kauf- mann in fifteen cases of Salivary gland tumors saw one angio- sarcoma; Nasse in thirty-six cases saw two carcinomata and two sarcomata; Volkmann in thirty-three cases saw one carci- noma, one myxoma which he is inclined to regard as of endo- thelial origin, and one fibrosarcoma. Landsteiner, who has already been quoted, saw only two purely epithelial tumors out of twenty-seven examined, the remainder being of the complex variety. - Volkmann also quotes Marchand as saying the greater proportion of the salivary tumors which the latter had seen were to be classed under the general group of endotheliomata, and that purely epithelial growths were of very great rarity. The tumors described by the French observers as adenomata or epitheliomata are usually of the complex type, and very few are either pure sarcomata or carcinomata. . Küttner in ninety-seven submaxillary tumors which he was able to collect from the literature found sixty-four endo- theliomata, six sarcomata, three doubtful cases of adenoma, five carcinomata, and nineteen cases in which the description was so imperfect that no determination of the nature of the growth could be made. It is evident, therefore, that the mixed tumors form a very large proportion of the tumors of the salivary glands. As regards the pharynx and lips, it is a matter of common experience that the mixed tumors are exceedingly rare in these situations, and that the purely epithelial tumors are exceedingly frequent. General Morphology.—The mixed tumors of the salivary glands are found, as a rule, to be encapsulated, lobular growths, 42 MIXED TUMIORS OF SALIVARY GLANDS. with harder and softer areas, the denser portions being due, as a rule, to the presence of cartilage or firm connective tissue. They can be divided macroscopically into three great rough groups with characteristic morphology, and to a certain extent with a definite clinical course. - - - First. Very fibrous tumors with very little cellular struct- ure and with but little mucous degeneration and no cartilage. Second. Very hard tumors containing large masses of cartilage and but little connective tissue or cellular parenchyma. Third. Soft, very cellular growths with transparent tra- beculae of mucous tissue surrounding areas which are opaque and yellow, which on microscopical examination will be found to be dense cellular areas, the color being occasionally, though not always, due to fatty degeneration or necrosis of the cells. The first and second forms are likely to be benign in their clinical course, while the third form is likely to recur locally or to run an exceedingly malignant course. A few of the parotid tumors examined and one from the pharynx belong to the first group; about one-fourth of the growths contain more or less cartilage and fall into the second; while the majority of the mixed tumors belong to the third. It must be remembered, however, that a great amount of varia- tion is present in different parts of the same tumor, and while one portion may be cellular the other may be composed largely of cartilage. The cellular tumors without cartilage, of the third class, have much the same consistency as an inflamed lymph node. The cut section of one of these tumors may show either a large amount of clear hyaline cartilage with a moderate amount of yellowish cellular substance lying between the pale blue trans- lucent nodules of the cartilage, or the whole growth may be cellular and of a pale yellow or grayish hue, or, finally, the cartilage may be scattered throughout the tumor in the form of small nodules. Some of the tumors resemble very closely soft fibromata, especially if much mucous degeneration is pres- ent. If the surface of the more cellular growths is scraped with 43 FRANCIS CARTER WOOD. a knife, a milky fluid can be obtained which is due to the removal of the alveolar cells by the mechanical action of the knife. The development of these tumors may take place before birth, in which case they are often associated with other con- genital defects of the facial region. (Case L.) As a rule, however, the actual development of these growths becomes noticeable most frequently during the second decade of life. Time of Occurrence.—In my own series the age at the time of the operation was as follows: one tumor under one year, seven between the ages of twenty and thirty, ten between thirty and forty, ten between forty and fifty, ten between fifty and sixty, and four over sixty. If the cases are subdivided according to the decade in which the tumor appeared, the age periods are somewhat different. Four tumors appeared in the first decade, one in the second, thirteen in the third, eight in the fourth, nine in the fifth, four in the sixth, while three still remain in the seventh. - The average age in these cases is then a little higher than that found by Kaufmann and Küttner, according to whom the majority of growths appeared in the second decade. The num- ber of tumors from other regions of the face is too small to furnish useful statistics. - The average duration of these tumors until they cause the patient to seek surgical aid is in my series eight years and nine months, thus varying only slightly from the time given by other observers, eight years. A long clinical course is not uncommon; thus, Case VI gave a history of the presence of a tumor for twenty years, Case IV for fifty-three years. The Social status of the patient may also determine a variation in the time which a tumor is allowed to remain, for among the working-classes these tumors often have a longer history than among the wealthy. Relations of Tumors to Surrounding Structures.—The anatomical position and the relations of these tumors to the surrounding structures are often characteristic of growths of 44 MIXED TUMCRS OF SALIVARY GLANDS. this type. They are seen in the pharynx attached to the bony structures of the hard palate, often indenting the substance of the bone, but not involving the bone substance. There is usually a thick fibrous capsule which separates the tumor from the periosteum, or the capsule of the tumor may be formed by the periosteum. They also occur on the inner surface of the cheeks, generally near the openings of Steno's duct, and in the substance of the lip, most often near the median line. They also occur in the cervical region in the upper triangle of the neck and in front of the sternocleidomastoid muscle. The most frequent situation, however, is near or in the substance of the parotid and submaxillary glands. Their relations in this case are most variable. The tumors may not be in any way con- nected with the gland but merely lie near it; they may be attached to the gland by a pedicle or be adherent to a small accessory gland; they may indent the surface of the gland, being separated from the acini by a fibrous-tissue capsule, or, finally, the tumors may be multiple and dot the substance of the gland itself. Usually in large tumors the remains of the glandular tissue are stretched out over the surface of the tumor and undergo pressure atrophy, but are separated from the tumor substance by a capsule. In the pharynx, the lips, and the cheeks the tumors usually lie just below the surface of the mucous mem- brane; but the parotid, submaxillary, and cervical new growths are below the deep fascia. They are, however, unless of a large size, easily movable in the tissues. The larger tumors are fixed, not so much by invasion of the structures surrounding them as by the tight stretching of the fascia which they cause during their growth. In those cases in which the tumor becomes malignant, the immobility is due, of course, to the invasion of the surrounding tissues. The skin is freely movable over the benign growths. The ear may be distorted or pressed entirely backward by a large tumor in the parotid region, especially if the growth has exten- sions behind the angle of the jaw, and is therefore unable to expand anteriorly, 45 FRANCIS CARTER WOOD. Clinical Course.—Considering the mixed tumors from the face and neck as a group, we may say that an average of the cases in the literature shows that some 25 per cent. undergo changes which express themselves in a clinically malignant course, while about 30 per cent. recur after operative removal, though some of these recurrences may be stayed by a second and more complete removal. The records of my own series are somewhat different. Out of thirty-seven cases in which notes of the patient’s condition after operation have been re- corded, seventeen, or 45 per cent., recurred locally. In four of the cases either internal metastases were soon apparent or the local condition was so serious that the patients either died of operative shock, as in two cases, or became inoperable from the invasion of the bone and deep tissues of the neck. Of the remaining thirteen cases one is still alive after many operations, but will probably soon succumb to the growth, which is now beyond removal. In twelve cases, then, or nearly 33 per cent., there were local recurrences which were checked by operation, So that a second removal should always be attempted if the anatomical conditions permit of a complete removal of the new growth. The malignancy of these growths can be judged to a cer- tain extent by their slowness of growth and their physical characters. The hard fibrous and cartilaginous tumors are apt to be benign, while the soft cellular types are likely to prove malignant. But frequently a tumor which has remained for a long time quiescent will begin a most rapid growth (Case XXXVII), and in a few months increase in size more than during its entire previous existence. This sudden and rapid growth is accompanied by the clinical and microscopical evi- dences of malignancy, and the tumor spreads through the sur- rounding tissues, involves the skin and the salivary glands, and may form metastases, changes which are illustrated by Cases XXV, XXVIII, and XXXVI. Twenty cases out of thirty- Seven, or 55 per cent., were permanently cured by operation. No sufficient explanation has been adduced for the sudden change which takes place in the benign growths and causes 46 MIXED TUMCRS OF SALIVARY GLANDS. them to become malignant. Kaufmann regards the constant traumatic irritation of the tumor by the muscles of mastication as a possible inciting agent. It is quite probable that the dense fibrous capsule which surrounds these tumors contributes to their prolonged benign character by preventing the tumor cells from easily spreading through the tissue; for when this capsule is removed by opera- tion on a quiescent tumor and small portions of the tumor cells are left behind in the scar, a rapid recurrence is the rule, and the new growth may reach a bulk in a few weeks greater than that of the original tumor, which may have taken years to develop. The increased vascularity of the tissues may have something to do with this rapidity of growth, for the compara- tive absence of blood-vessels in the primary mixed tumors is noteworthy. This poor blood supply in the original growths and the lack of close relationship between the cells of the tumors and the vessels may also be contributive to the comparative in- nocence of the mixed tumors, while in the sarcomata, with their very abundant blood supply and close relationship between cells and vessels, extreme malignancy is the rule. There is still a third element which must be considered as a possible factor in the development of malignancy in these growths, and that is the abundant embryonal connective tissue which is found in many of them. In two of the writer's cases of very malignant growths, the metastasis was purely sarcomatous; there was no alveolar arrangement of the cells and none of the other elements of the original tumor were present. It seems likely that the embryonal connective tissue of the original growth underwent a change which expressed itself in the formation of a rapidly growing sarcomatous tissue, and that the other ele- ments of the tumor took no part in the process. The cartilage of the original tumor does not, as a rule, appear in the recur- rences, and the amount of elastic tissue is in general much less than in the primary growth. In those cases in which a recur- rence does contain cartilage, as in Case XXVI, the removal of the primary growth has been very incomplete, so as to leave portions of the original cartilage in situ. 47 FRANCIS CARTER WOOD. In only one of the cases examined by the writer has there been metastatic invasions of the lymph nodes. This case (IV), however, was complicated by what is probably to be classed as an epithelial tumor arising in a mixed growth of the parotid. Both the remains of the mixed tumor structure and the epithe- lial alveoli are to be found side by side in the same section. None of the other cases of the mixed type of tumor furnished metastases involving the regional lymph nodes. Tanaka, how- ever, considers the invasion of the regional nodes as a very frequent occurrence with the endothelial tumors, but his cases are to be viewed with considerable doubt, as none of them were from the salivary glands, and the histories and the pathological reports would incline one to regard the cases which he describes as sarcomata, especially those arising in the skin. The case of Griffini and Trombetta may be an exception to the rule of the non-invasion of the lymph nodes. The tumor was cartilagi- nous and of sixteen years' duration. At the close of that time it began to grow rapidly, and on the death of the patient was found to involve the lung, the pleura, the cervical, and the bronchial lymph nodes. In the tumor masses in the pleura and the lung some cartilage was detected, but none in the lymph nodes. The tumor had invaded the bone marrow and the ves- sels of the neck, and sent long strands down along the channels of the neck, so that it is possible that the metastases were by direct extension from the primary growth. The authors class it as a chondrocarcinoma of the submaxillary gland, so that it may fall under the same class as Case IV, the drawing ap- pended to the paper showing a growth of carcinomatous type. The metastases of the mixed tumors are of two varieties, local and remote. In the first form the recurrence is usually of the same type as the original tumor; that is, the general morphology of the original growth is retained and the recur- rence may contain cartilage, mucus, and hyaline tissues. When the recurrences extend over a long period the tumor may grad- ually lose its characteristic morphology and resemble a sarcoma very closely (Case XXV). This change should be care- fully distinguished from that in which the tumor stroma under- 48 MIXED TUMCRS OF SALIVARY GLANDS. goes a Sarcomatous change with the production of a typical sarcoma with rapidly growing and wide-spread metastases. In the secondary growths of this type the stroma retains no trace of the embryonic structures, no cartilage or hyaline tissue is present, and the tumor cells do not show their original alveolar arrangement. Cases XLI and XLII are examples of such changes. In a case reported by Barozzi and Lesné, in which a recurrent tumor occupied the entire left side of the face and neck, the cervical and tracheal lymph nodes were invaded by the new growth. The original tumor appears to have been not a cylindroma, as it is designated by the writers, but an epithelioma with mucous degeneration of the connective-tissue stroma. In fact, the term cylindroma as used by the French writers has not the same significance as when used by Sattler and other German writers, but in general may be considered as designating an epithelial tumor with degeneration of the Stroma. Occasionally the growth invades the wall of one of the large veins of the neck, and emboli are carried in this way throughout the body, as occurred in Case XLI of my series; but this is rare, and the growth, as a rule, remains fairly well localized, local recurrences being not infrequent unless the re- moval of the original growth has been very thorough, a condi- tion often difficult of realization in the parotid growths owing to the close contact of many of the growths with the facial nerve. Case XXV is an excellent example of repeated recur- rences extending over a long period of time, while Case X is an example of a prompt recurrence with a final radical removal of the growth. Cases XXVII and XXXVI are examples of frequent recurrences with the final death of the patient. Billroth reports a case in which, during the course of twenty-three years, nine operations were performed to remove local recurrences. The cases of primary endotheliomata of the cervical lymph nodes which exist in the literature (Zahn, Hoffmann, and Volk- mann) ran an exceedingly malignant course, while the single case which the writer has had opportunity to observe was quite benign and did not recur after removal (Case XIX). 49 FRANCIS CARTER WOOD. RÉSUME OF THE MORPHOLOGY OF THE TUMCRS. The details of the microscopic description of the mixed tumors are best subdivided into two groups, one embodying the structure of the stroma; the other, the structure of the parenchyma. - The connective tissue of the tumors is very variable in structure and amount. In the denser forms, especially in the tumors from the pharynx and in Some of the parotid tumors, the connective tissue is arranged in very firm, dense bands trav- ersing the growth, and also surrounding it, in the form of a firm fibrous capsule. The bands traversing the substance of the tumor divide it into a series of trabeculae in which the cellular structures lie. These bands, as well as the capsule, stain intensely with the Van Gieson stain, and also take a very clear and sharp stain with the Mallory phosphotungstic acid haematoxylin, especially if, as Ribbert recommends, the mate- rial has been fixed in Zenker's fluid. There is only occasionally evidence of a basement membrane of connective tissue about the alveoli of the parenchyma cells such as exists in the normal parotid or submaxillary. The finer fibrillae cannot be demon- strated between the cells of the alveoli nor between the cells of the solid strands of the so-called endothelial cells, even when the tissues are stained deeply with the phosphotungstic mixture or Van Gieson’s stain. The absence of connective-tissue fibrillae between the cells of a tumor is usually regarded as indicating the epithelial origin of the tumor cells, since they are usually present between the cells in sarcomata. Some writers, notably Barth, claim to have been able to demonstrate the presence of fine fibrillae of connective tissue between the cells of the so-called endothelial masses by means of special stains. Barth recommends Rosin's stain, which is a modified Ehrlich-Biondi tissue stain. The staining solution is made by taking four centigrammes of the dry Ehrlich-Biondi mixture, dissolving it in Ioo cubic centimetres of water, and adding to the solution seven cubic centimetres of a half per cent. aqueous solution of acid fuchsin. Paraffin sections remain in this for five minutes and are then washed for two minutes in water, differen- 50 MIXED TUMCRS OF SALIVARY GLANDS. tiated for ten seconds in I to 2000 acetic acid, dehydrated in alcohol, cleared in xylol, and mounted in dammar. The nuclei are green, the connective tissues are red. Barth's drawings are excellent, but not absolutely con- clusive; and it seems possible that he has stained some of the spindle-shaped connective-tissue cells which often lie between the flat endothelial cells, and that true fibrillae are not demon- strated. The writer has tried this method on a number of tumors of different types, and has not been able to demonstrate this condition; nor has it been possible to do so with the phosphotungstic acid method, though the fibrillae can be easily shown with this stain between the cells of many of the sar- comata which have been used as a control for the method. In the softer tumors the connective tissue is correspond- ingly looser in texture; it is often oedematous, and very fre- quently is replaced by myxomatous tissue. The myxomatous changes are frequently seen, but do not seem always to be due to a degenerative process, as was considered probable by Volk- mann, but rather to an actual presence in the tumor of embry- onic tissue of the myxomatous type. The evidences for this assumption lie in the nature of the tissue, which does not resemble the ordinary myxomatous degeneration in connective tissue, but is more like the tissue of the embryo with its long branching cells. The quite constant association of the myxo- matous tissue with cartilage is another reason for assuming its origin to be embryonic, for, as will be seen later, there are many reasons for supposing that the cartilage is derived from the Meckel's or Reichert’s cartilage of the developing embryo. Embryonic connective tissue without myxomatous struct- ure is frequently found in these tumors. It consists of a loose fibrous tissue stroma with spindle cells scattered through it in small numbers. The spindle cells form a portion of the fibrous net—work of the connective tissue by giving off from their cytoplasm long fibres which, in properly stained specimens, can easily be traced for a considerable distance. The cells of the stroma have been described more or less while discussing the structure of the connective-tissue elements, so that only a review is necessary here. These cells vary in 51 FRANCIS CARTER WOOD. number and morphology in the different tumors, depending largely upon the condition of maturity of the connective tissues present. The stroma of the dense growths contain but few cells, and these are of the familiar connective-tissue type; that is, of long spindle-shaped cells with small, evenly staining nuclei without definite chromatin net-work, and rarely showing mitotic figures. The tumors in which connective tissue of a more embryonic type is present, and the tumors of the previous group in which portions of the stroma have undergone a sar- comatous change, show cells with large nuclei and well-marked chromatin net-work. The cell body is more prominent and more sharply outlined and mitotic figures are occasionally seen. In the myxomatous and cartilaginous portions the cells are either the branching cells regularly seen in myxomatous tissue or they are the encapsulated cells of hyaline cartilage with their double nuclei. In the tumors examined by the writer no evi- dences of proliferation were found in either the cells of the myxomatous tissue or in those of the cartilage, though such appearances in the cartilage of these tumors have been noted by von Ohlen. The large flat cells of the tumor parenchyma are also found scattered through the myxomatous portions of the new growth and about the periphery of the cartilage, which fact has given rise to speculations concerning the origin of the car- tilage from the flat so-called endothelial cells, or of the flat cells from the cartilage. These flat cells may be few in number and diffusely scattered throughout the stroma, or they may be collected into strands spreading out diffusely through the myx- omatous or fibrous tissue and also entering into close rela- tionship with the cartilaginous areas. Some of these cells in the myxomatous portions may be observed to be closely con- nected with the stroma by giving off the fibrillae into its sub- stance, others retain the oval or polyhedral form of the flat epithelial cells lining the alveoli. Giant cells are not uncommon, especially on the cellular borders of the myxomatous areas into which they often project. The mechanical conditions here which permit the free growth of the cells into the soft myxomatous areas may possibly ac- 32 MIXED TUMCRS OF SALIVARY GLANDS. count for the condition, especially as similar pictures have been seen in pleural growths where the same freedom of growth is possible. Cartilage is found in about one-fourth of the mixed tumors of the salivary glands and of the facial region which the writer has been able to examine; other observers have found it in half the tumors examined. Osteoid tissue is fairly frequent, while bone is rare. Bone was present in no one of the tumors examined by the writer. The cartilage is either hyaline or fibro-elastic. As a rule, there are numerous fine elastic tissue fibrillae scattered through even the more typical hyaline cartilage in the form of a fine mesh-work easily demonstrable by Weigert’s method for elastic fibres. The cells of the cartilage are usually similar to those of hyaline cartilage under normal conditions; they may be encap- sulated or free in the homogeneous stroma. Occasionally, however, there are variations in the arrangement and the num- ber of the cartilage cells and tumor cells which render it easy to distinguish from the normal tissue. The most frequent anomaly is the presence of a large number of cells without the capsule which usually surrounds them. A number of cells may collect in small masses in the substance of the cartilage and calcification may take place, the salt deposited being most frequently the sulphate of calcium. Finally, the substance of the cartilage may soften and produce small cysts containing a clear brownish fluid. The cartilage may form a greater part of the tumor, or it may be very scanty in amount and merely scattered through the tumor in small islets. In all cases it is in very intimate connection with the connective-tissue stroma and cells, and also with the so-called endothelial cells of the tumor. The elastic tissue content of these tumors is not mentioned in detail in any of the monographs on the subject. Indeed, the literature on elastic tissue in tumors is still meagre and contra- dictory in its statements. Melnikow-Raswedenkow found no newly formed elastic fibres in tumors, all those present were derived from the organs in which the growths occurred. He examined connective-tissue tumors, epithelial tumors, teratoid 53 FRANCIS CARTER WOOD. tumors and cysts, and concludes that there is but little elastic tissue in any one of these growths, and also that the elastic tissue of the organs invaded is only mechanically influenced by the tumor cells in their growth. Williams, on the other hand, found that a few fine fibres are Occasionally seen in the newly formed stroma of the carcino- mata, but only in tumors in which the stroma was especially abundant. - - Alice Hamilton describes a fibrosarcoma of the brain in which numerous elastic fibres could be demonstrated which she considered as newly formed, inasmuch as they were not con- fined to the region of the blood-vessels, which is the only site of elastic fibres in the brain substance. Huguenin, in an ade- noma (?) of the testicle containing cartilage, glands, and epi- thelial masses, says that the cartilage in his preparations con- tained no elastic fibres, but admits that the stain was not entirely Satisfactory, owing to the method of preserving the material. In some of my own specimens of mixed tumors of the testicle elastic tissue was very abundant throughout. Gliuski found little or no elastic tissue in a mixed growth from the oesopha- gus. K. Landsteiner is the only observer who mentions the abundance of the elastic fibres in the salivary tumors. The elastic tissue in the mixed tumors is very much more abundant than in the sarcomata or carcinomata. In the latter it is largely confined to the pre-existent stroma, and there is only a moderate new production, and this in the region of the vascular system. In the mixed tumors, however, the stroma of the growth is penetrated by a fine mesh-work of fine and coarse fibres of elastic tissue. These are specially abundant in the dense fibrous tissue of the capsule and the trabeculae of the more fibrous growths. In the soft embryonic fibrous tissue of the more cellular tumors the elastic tissue is in the form of very fine branching fibrillae having no connection with the blood-vessels. In the cartilaginous parts the elastic tissue pene- trates all portions as a fine net-work, not so dense, however, as in true fibro-elastic cartilage. The elastic fibres usually encircle the alveoli in considerable numbers, but do not penetrate between the cells lining the alveoli. When the stroma cells 54 MIXED TUMC)RS OF SALIVARY GLANDS. diffuse out through the connective tissue, or even when they remain in Solid strands, the finer elastic fibres often penetrate between the cells. The fibre net-work may surround a single cell or it may surround a group of several cells. Blood-vessels are infrequent in these tumors, and when present seem to bear little or no relation to the cells of the parenchyma. They are usually seen in the connective-tissue trabeculae of the more fibrous growths and present no pecu- liarities. - Small areas of fat tissue are to be found in a number of the mixed tumors inside the capsule of the growth and lying in the connective-tissue stroma. They are quite independent of the fat of the surrounding connective tissue, the gland with which the tumor is in relation, and are probably derived from the embryonic connective tissue which forms the stroma of the growth. - The same is probably true of small lymphoid areas which are not uncommon in the mixed tumors. Their derivation is in all probability the lymphoid tissue of the buccal and pharyn- geal mucosa, and is especially well marked in the cysts and clefts from the branchial pouches which have persisted until adult life, and still retain a well-developed envelop of lymphoid tissue about the epithelium lining the cavity of the cyst or sinus. The Cells of the Parenchyma.-The cells of the paren- chyma of the mixed tumors are arranged either in alveoli or in solid strands, which are often connected with alveoli at some portion of their extent. When arranged in alveoli they are large and flat, resembling epithelial cells in their general grouping, but differing from ordinary epithelium in having an oval homogeneous nucleus without a well-marked nuclear net—work, the entire nucleus staining diffusely. The cells under ordinary conditions do not show intercellular bridges nor spines, such as are frequently seen in epithelial cells. These cells may be high and cylindrical; they may be cuboidal or polygonal; and, finally, if the lumen of the alveolus be filled with accumulated material, they may be flattened out into low cells resembling the endothelium lining the blood-vessels. In a number of the tumors examined, instead of alveoli being 55 FRANCIS CARTER WOOD. present the parenchymal cells are in long strands, as above indicated, extending through the tissues closely attached to the connective tissue of the tumor and often containing long spindle-shaped cells of an evidently connective-tissue nature. The cells of the strands may also give off long fibrillar pro- longations to the connective tissue. This can be best seen in the myxomatous portions of the tumor, for there the fibrillae can be more easily traced. As was stated under the heading of the fibrous stroma, no connective tissue can be made out between the parenchyma cells, either when they are in alveoli or in strands. When sec- tions are stained either with Van Gieson or with Mallory's connective-tissue stain, fibrillae can occasionally be demon- strated between the cells, but this appearance is due to the spindle cells lying between the others, and is not of constant Occurrence. A basement membrane is not usually present about the alveoli, but the elastic fibres are occasionally very abundant around them. In a few tumors the cells lining the original alveoli are crowded together by connective-tissue growth and degenera- tion, so that the lumen is lost and the whole structure distorted in much the same way as in an intracanalicular fibroma of the breast. Case LIX is an excellent example of this form of distortion. (Plate VIII, Fig. 2.) An important point of difference between the cells lining the alveoli of this form of tumor and the cells of an epithelial new growth is that the normal endothelium lining the lymph spaces can generally be seen under a layer of epithelial cells, while it cannot be so seen in the alveoli of the mixed tumor cells. That this is not always true can be seen in Case XX, where the so-called endothelial cells of the tumor can be made out advancing over the normal endothelium lining a lymph space, just as the cells from a carcinoma may be seen to advance over the endothelial cells of the lining of the lymph channels of a lymph node. (Plate II, Fig. 2.) As previously stated, Volkmann claims that the so-called endothelial cells lining the alveoli do not shrink away from the connective tissue with which they are in contact, as is the case 56 MIXED TUMIORS OF SALIVARY GLANDS. in carcinomata, even when the tissue is considerably shrunken by fixing agents. Borst and others have shown, however, that this distinction cannot be strictly held, for in some of the endo- thelial tumors the cells do shrink away from the connective- tissue walls of the alveloli, and in carcinomata it is easy to find areas where the epithelial cells remain in contact, even though there is considerable shrinkage during the hardening process. It is true, however, that the separation occurs more frequently in the carcinomata and is rare in the mixed tumors. Epithelial cells with well-marked spines are more or less abundant in the parenchyma of eight cases in my series of fifty- nine. They are arranged very frequently in pearls, the centre of the pearl staining strongly with eosin, as in the pearls from the ordinary epitheliomata of the skin. When stained with Kromayer's modification of Weigert's fibrin method, well- marked keratohyaline granules can be made out in Some of the cells at the periphery of the pearls, and also a moderately well- marked fibrillation which passes from one cell to the other. (Plate V, Fig. 2.) The intercellular bridges are well marked in many of the pearls and in some of the alveoli in which no pearl formation has taken place. In six of the other tumors examined, which did not con- tain pearls, cell alveoli were found which bore a very close resemblance to the ducts of an atrophic salivary gland. These alveoli were lined with a single layer of cuboidal cells and pos- sessed a well-marked basement membrane. (Plate I, Fig. 2.) Three of the smaller tumors were embedded in paraffin and sectioned in a complete series. In no one of them were well-marked epithelial spine cells demonstrated. A large num- per of sections from the other tumors were examined, but no spine cells were found; so the presence of epithelial spine cells cannot be regarded as a constant or even a frequent occurrence, as Hinsberg is inclined to believe. Connected with the pearls, branching strands of cells are often seen, which lack the characteristic morphology of epi- thelial cells and resemble the strands considered by Volkmann as endothelial in nature. 57 FRANCIS CARTER WOOD. DISCUSSION OF RESULTs. If we look over the case histories of a series of tumors of the type which we have described, in order to obtain some clue to the conditions which surround their origin, we are struck, in the first place, with the fact that all these new growths arise in regions of the body which are characterized by two conditions,—one, the presence of complex organs either of epithelial or mesoblastic origin, and, secondly, from an embryological point of view, by an extremely complicated course of development. • Both of these conditions obtain in the facial region and in the region of the development of the kidney, ovary, and testicle, to which areas these complex tumors are practically confined, if we except the site of the postanal gut, where complex forms also occur under precisely the same conditions. If we consider these tumors as a whole, we must notice that they contain elements which are derived from the meso- blast at a comparatively early stage of differentiation. Evi- dently, cells of mesoblastic origin must have been left in these situations at a period when they contained all these tissue possi- bilities and before a final differentiation took place, for the change into a definitive form of cell precludes the possibility of further differentiation; as, for example, the final changes of undifferentiated embryonic mesoblast into the highly special- ized cell forms, such as muscle or cartilage, preclude further changes except of a regressive type. - It is interesting to observe more closely the association of these tumors and malformations which result from imperfect adjustment of the tissues which go to form the facial region. These malformations are the dermoids of the orbit and the eye, the cysts and fistulae which arise from the imperfect closure of the thyroglossal duct or the branchial clefts, the imperfections from lack of union of the lateral processes forming the mouth and cheeks, and, finally, the malformations of the ear, with either an accessory tragus or mandibular or cervical tubercle. In Case L, a small congenital parotid tumor was present in the upper part of the parotid, while just behind it was a super- - 58 MIXED TUMCRS OF SALIVARY GLANDS. numerary Spur of cartilage derived from the non-union of the spina helicis with the helix. There was present also an auricu- lar fistula in the helix resulting from the same lack of union between the helix and the spina. (See Schwalbe and Grade- nigo.) The tumor was a complicated growth, containing among other things a considerable amount of hyaline cartilage; the spur, on the other hand, contained yellow elastic cartilage, and was therefore not, in all probability, derived from the same Source as that in the tumor; but the connection of the tumor of the parotid with the other congenital lesions is strongly sug- gestive of the simultaneous displacement of both sets of remnantS. The derivation of the cartilage which these complex tumors contain has always interested pathologists. There are two chief possibilities to be considered; one, which has long been held, is that the cartilage is derived by a process of meta- plasia from the connective tissue of the tumor or from the tumor cells; the other is that the cartilage is either a remnant left in the formation of the branchial arches, or that certain cells are left from that portion of the embryo which is to finally form these arches, and that these cells are capable of developing into either an undifferentiated fibrous or mucous tissue or into the more highly differentiated form of cartilaginous tissue. Virchow's idea was that the cartilage originated in the connective tissue, much as cartilage arises from the periosteum of bone, and this view is developed in Volkmann's paper. The latter holds that the process is “exceedingly simple,” and con- sists in a chemical alteration of the intercellular substance by which the fibrillae of the connective tissue are softened and dissolved into a more or less homogeneous mass of basement substance, and thus forms either cartilage or myxomatous tissue. The flat cells of the tissues then undergo a metamor- phosis in response to that which has taken place in the con- nective tissue, and become either encapsulated cartilage cells or emit fibrillae and become the branching cells of the myxoma- tous tissues. All transitions can be observed between the cells and the connective tissue; but the most frequent process is the transformation into myxomatous tissue first, and then an 59 FRANCIS CARTER WOOD. alteration of the cells and the stroma into cartilage. Volkmann and von Ohlen find that the cartilaginous portions are often the site of very active proliferation, and that they form long strands of cells similar to those seen in other parts of the growth. Against all these views are the facts that the carti- laginous tumors are of extremely slow growth clinically; that mitotic figures have not been found in the cells of the cartilage by the writer and many other observers; that the elastic tissue is not arranged in the cartilaginous portions as it would be if the pre-existing connective tissue had merely become softened, but is in fine fibrils scattered throughout; that the cartilage is not found in the metastases, but only in local recurrences; and, finally, that these degenerative changes in connective tissue with the formation of cartilage do not take place in other tumors which contain abundant connective tissue, but only appear in a certain narrow group of tumors which are in all probability congenital in their origin. Cohnheim, in his lectures on general pathology, said long ago that “the germs of the cartilaginous tumors of the parotid region are unused remnants of the cartilaginous portions of the branchial arches.” Klebs believes that when the organism has reached a certain physiological point of development and the cells possess a degree of specific stability, that cartilage can only be produced from embryonic remnants. Orth also holds the same view of the origin of the car- tilage, while most of the special writers on the subject support Volkmann in deriving the cartilage from the connective tissue. In general, it may be said that the idea of the change of one tissue into a closely related form is gradually losing the general acceptance which it once possessed, and such changes are more and more being considered as due to embryonic errors of development. - Ribbert, in his latest publication, when speaking of the metaplasia of connective tissue into cartilage or myxomatous tissue, says that such changes even within these narrow limits are very infrequent in their occurrence, and that there is con- siderable reason to suppose that these appearances may be due 60 MIXED TUMIORS OF SALIVARY GLANDS. to embryonal misplacements of tissues. In another section he considers it probable that most, if not all, of the pure myxomata are due to some congenital remnant of mesoblast displaced at a stage when it contained the possibilities of myxomatous devel- opment. Lubarsch states that while it is undoubted that the “ epithelia” lining the serous cavities and the blood-vessels can produce connective tissue, which can best be seen in the or- ganization of thrombi and the inflammation and foreign body irritation of the serous cavities, the question of the metaplasia of endothelial cells into cartilage cells is more than doubtful. Benecke considers that this change is possible in tumors, but the condition may be due to an embryonic misplacement of cells. Certainly the fact cannot be regarded as proven. The possibility of the metaplasia of true ento- or ectodermal cells into connective tissue has never been positively shown. Kro- mayer's interpretation of the pictures which he obtained in the examination of the soft naevi of the skin has never been widely accepted, and the view of Unna that the naevus cells are epithelial in origin is now generally received. It is especially difficult to prove the metaplasia of epithelium into connective tissue in the skin because the epithelial cells become flattened and drawn out into long spindle-shaped forms which often lie in close connection with the connective-tissue fibrillae of the pre-existing stroma, So that this apparent change into con- nective-tissue cells is to be regarded as the result of mechanical action, and true connective-tissue cells in the biological sense have not been produced. This mechanical change may also occur in the epithelial masses of the salivary mixed tumors. The main weight of the evidence seems to be on the side of the congenital misplacement of cells which have the power to form either cartilage or myxomatous tissue. The same ex- planation would account most simply for the presence of fat and lymphoid tissue in these tumors. Fat tissue is not ordi- narily seen in tumors as a portion of the growth itself, but usually as a part of the tissue through which the tumor may be growing. But in many of the congenital tumors, for example, those from the kidney, whose origin from embry- onic misplacements is more definitely settled than in the case (51 FRANCIS CARTER WOOD. of these complex tumors of the parotid, the presence of fat tissue and striated muscle in intimate relationship is regarded as a valuable evidence of the derivation of the tumor stroma from the mesoblast at a time when the kidney rudiments were ii, close contact with the muscle-plates of the middle layer of the embryo. The examination of sections of human embryos of from six to twelve weeks of age shows that such embryonic displace- ments may easily occur in the region of the salivary glands. In transverse sections through the head of embryos of this age, the parotid and the submaxillary may be seen as small outgrowths from the floor of the mouth, lying almost in contact with the first and second branchial arches, or rather with the cartilage which these arches contain. Very often tubules of the gland may be found to be surrounded by the cells of the perichondrium. Inasmuch as the arches reach their development at or about the fourth week, and at that time contain the cartilagi- nous tissue which forms Meckel’s and Reichert’s cartilages, while the parotid does not appear until the sixth or eighth week, for any embryonic remnant to contain both parotid and carti- lage, the tissue must have been displaced before the fourth week. It is not therefore a portion of parotid tissue which is displaced, but a portion of the epiblast which is to line the buccal cavity, together with some of the underlying mesoblast, the latter carrying with it the whole group of mesoblastic possibilities, cartilage, myxomatous tissue, fat, and even mus- cle." This necessarily early displacement of the tissues would make up the tumor accounts for the great variability of the mesoblastic and epiblastic structures present in the growth and the close intermingling of cartilage, myxomatous tissue, hya- line, fat, muscle, and bone, together with cells of both epithelial and connective-tissue types. This early derivation of the remnants is also a point upon which Wilms has rightly insisted as rendering superfluous the * See report of a rhabdomyoma of the parotid by Prudden, American Journal of the Medical Sciences, 1883, page 438. 62 MIXED TUMIORS OF SALIVARY GLANDS. explanation which Hinsberg finds necessary to make in order to account for the presence of cornified epithelium in the midst of these complex tumors. Hinsberg has to assume a metaplasia of the parotid glandular epithelium into epidermal cells; but if it be remembered that the displacement takes place before the buccal epithelium has become differentiated into parotid, and while it still retains the qualities inherent in that epithe- lium, that is, of forming epithelial cells for the lining of the buccal cavity, the assumption of metaplasia to account for the presence of the epidermal cells is quite unnecessary. Hinsberg's further assumption that the epithelium might possibly be derived from the rudiments of the tympanum, which at this stage lies quite close to the parotid, is rendered doubtful by the fact that epithelium has been found in the submaxillary tumors, but the distance of the submaxillary gland from the tympanum is so great at all stages of development that such a displacement of tympanic epithelium is quite out of the question. . If now we turn from the elements of the stroma to the parenchyma of these tumors, the most striking feature is the presence of numerous cellular structures arranged either in solid strands or in the form of alveoli lined with flattened cells. Some of these cells are evidently epithelial in origin from their morphology, which is that of spine cells with well- developed intercellular bridges and keratohyaline granules or of tubules lined with cylindrical epithelium. The latter forms occur, so far as the tumors examined show, but rarely in any of the new growths from the facial region; they are frequent, however, in the tumors of the testicle. The other type of cells is that which has long been designated endothelial, because these cells possess a morphological aspect strikingly similar to the endothelia lining the lymph spaces. Three views are possible as to the origin of the cells. One, that they are endothelial in nature and derived from the endo- thelium of the lymph spaces; second, that they are epithelial and derived either from a misplaced portion of the parotid or from a misplacement of the mesoblast and epiblast at a period when the cells have not as yet developed into the highly differ- 63 FRANCIS CARTER WOOD. entiated gland cells, but contain the possibilities of such devel- opment only, as is explained above in the discussion of the probable origin of the cartilage; third, that the tumors contain both endothelial and epithelial elements. We will first take up the endothelial view as presented by Volkmann and his successors. I. Is the endothelium capable of producing tumors? That this is probable is shown by the cases of tumors occurring in lymph nodes where epithelium cannot be present, and also in tumors of the bone marrow and the spleen, such as have been recently collected by Bovaird. The splenic tumors are, however, more probably hyperplastic growths than true tumor formations. The tumor reported by Sailer must also be re- garded as a growth from the endothelium of the vein in which it was found. The tumors of the dura are no doubt of an endothelial character. 2. Are the parenchyma cells of the complex salivary tumors derived from the endothelium ? The various facts which have been adduced to substantiate this view are: (a) The cells resemble endothelial cells from a purely morphological aspect. That this is of slight value in the determination is shown by Case VI of my own series, in which typical epithelial cells can be found in strands of cells otherwise conforming in all morphological points to endothe- lium. The same is true of cases reported by Hinsberg, Land- steiner, and Wilms in which spine cells have been found. (b) The cells of the tumor are continuous with the cells of the peripheral lymph spaces in the capsule of the tumor, and that the endothelium of these spaces takes part in the growth of the tumor. Ribbert, Borst, and others have shown, how- ever, that the lymph-space endothelium does not take part in the tumor growth, and that the spreading of cells of a new growth along lymph spaces, either with or without active growth of the lymph-space endothelium, is possible in carci- noma as well as in the so-called endothelial tumors. Ribbert has also called attention to the fact that the spaces which Volk- mann regarded as normal lymph spaces in the fibrous tissue were not in fact lymph spaces, but tumor alveoli. The apparent . 64 MIXED TUMCRS OF SALIVARY GLANDS. new growth of endothelium in the spaces is due to the fact that the cells of the tumor grow over the endothelial cells already present and give the appearance of an hyperplasia of the pre-existing cells. - • * * (c) The argument based upon the supposition that the cells of the endothelial tumors do not shrink away from the walls of the lymph spaces, and that the original endothelial cells cannot be seen under them, has been shown, by the obser- vations of Borst, Lubarsch, and many others, to be incorrect. One of my own cases shows this shrinkage in many parts of the tumor. (Case XX, Plate II, Fig. 2.) (d) Ribbert, Lubarsch, and others have shown that the fact upon which so much stress has been laid, that the solid strands of cells are in intimate connection with ceils of un- doubted mesoblastic origin, has but slight value in deciding that the two groups of cells are of identical origin, for such secondary unions of true epithelial and connective tissues are not infrequently observed in the epithelial tumors of the skin. (e) The fact that the tumors are not closely connected with the salivary glands, and that the cells of the new growth are not connected with those of the gland, but are separated from it by a capsule, does not prove that the cells of the tumor are not of epithelial origin, but only that they were separated from the glandular rudiments before the latter were enclosed in the capsule normally surrounding the gland. The formation of this capsule takes place about the fourth month of foetal life, while up to this time the acini of the gland are in close relationship with the periosteum of the inferior maxilla. - If we take up, on the other hand, the considerations which point to the view that the cells of these complex tumors are derived from the outer layer of the embryo, we find that they are based upon the following facts: - (a) Epithelial cells with intercellular bridges and kerato- hyaline granules are found in a few of these tumors, which are otherwise of a typical endothelial morphology. The epithelial cells lie in masses which would under ordinary circumstances be considered as very closely related to the surrounding con- nective tissues. - .65 FRANCIS CARTER WOOD. (b) Well-formed tubular glands with a basement mem- brane and a lumen filled with a homogeneous secretion are found in a few of these tumors, which are otherwise typically endothelial according to Volkmann. g - . (c) A great proportion of these tumors contain alveoli filled with a material resembling the colloid secretion of the thyroid or the material which may be found in the alveoli of an atrophic salivary gland. r (d) There is no connective tissue between the cells, as is generally the case in tumors of a mesoblastic origin. This is seen in stained preparations and in sections which have been either shaken out or pencilled to remove the cells of the paren- chyma. - (e) These tumors occur only in or near epithelial struct- ures and in regions in which congenital inclusions in the form of dermoids and cartilaginous remnants are fairly frequent. The intimate mixture of tissues of widely different type in a single tumor is a strong presumptive evidence that the growth is of congenital nature. It is difficult to apply a suitable name to this group of tumors, and yet it is advisable to do so, if for no other reason than to call attention to the fact that they cannot be classed as adenomata, nor as epitheliomata or carcinomata, nor yet as sarcomata. From each of these groups of tumors the compli- cated growths which we have studied are separated by their morphology alone. The writer prefers to return to the old name of “mixed tumors,” or, perhaps better, complex tumors, though the introduction of the latter might be justly criti- cised as adding a new name to our already complicated nomenclature. - The problem of the exact nature of these growths cannot be definitely settled so long as we must rest our distinctions upon morphological or histogenetic differences. The experi- ments which Ribbert, Lubarsch, and many others have carried out on the results of the transplantation of normal glandular tissues have given us only studies upon the atrophy and partial regeneration of fragments as grafted upon other tissues and under the most unfavorable conditions for permanent growth. 66 MIXED TUMCRS OF SALIVARY GLANDS. We cannot expect too much of the transplantation of mature, highly differentiated tissues into an unfavorable environment, nor have the results of the transplantation of embryos or frag- ments of embryos afforded the results for which we hoped. The change of place and environment must be made at an earlier date and in a manner somewhat less crude than has been practised heretofore. Experimental teratology has at present only given results when applied to the lower forms of animal life. - CONCLUSIONS. To sum up the results of this study in a few words: I. There is a group of extremely complicated tumors oc- curring in the facial region which contain elements from both epi- and mesoblast in most intimate relation to each other. II. The complicated structure of the stroma, containing as it does elements such as embryonic connective tissue, carti- lage, bone, fat, and lymphoid tissue, and very rarely striated muscle, is explained most easily by the assumption of an em- bryonic misplacement of mesoblast. III. The structure of the parenchyma is so slightly char- acteristic in morphology that its epithelial nature in all cases can only be considered as probable; yet in about 24 per cent. of the tumors examined the presence of epithelium is un- doubted. The form and relationships of the cells of the paren- chyma do not furnish sufficient data to justify these cells being regarded as of endothelial origin. IV. The theory of early embryonic displacement of epi- blastic tissue during the process of formation of the parotid and submaxillary glands and the branchial arches may account for many of the morphological peculiarities of the cells of these tumors, especially the lack of many typical features which we associate with epithelium. The same condition may be seen in the epithelial cells of the congenital moles, in which the epi- thelium is with difficulty distinguished from connective-tissue cells, owing to its close connection with the stroma of the tumors and its undifferentiated type. V. The mixed tumors of the salivary glands run a clinical course strikingly different from the sarcomata and carcinomata 67 FRANCIS CARTER wooD. in that they are slow growing and generally benign. The regional lymph nodes are not invaded, and recurrences are likely to remain local in a considerable proportion of the cases. BIBLIOGRAPHY. Only those papers which are directly concerned with the salivary group of tumors are included here, and chiefly those which have appeared since 1896. Volkmann’s “Monograph” contains a very complete collection of the articles before that date. Barozzi et Lesne. Cylindrome de la region sous-maxillaire. Soc. Anat. de Paris, March 19, 1897, p. 266. Barth, T. Lymphangio-Sarkome des Mundbodens. Diss., Jena, 1896. Also in Beit. v. Ziegler. Beadles, C. T. Myxomatous Tumor of Submaxillary. Trans, of London Path. Soc., Vol. xlviii, 1897, p. 66. * Benecke. Artikel Chondrom in Drasches Bibliothek der med. Wissen- schaft, 1900. Ueber freies Wachstum metast. Geschwulstelemente in serösen Höhlen. Arch. f. klin. Med., Band lxiv, p. 237. Berger. Rev. de Chir., 1897, p. 361. Tumeurs mixtes du voile du palais. Biondi. Sei casi di chirurgia della glandola parotide. Clinica chirurgica, 1899, No. 1. : Billroth. Virch. Archiv, Band xvii, 1859. Bland-Sutton. Tumors, Innocent and Malignant. Böhme. Zur Kasuistik d. Speicheldrüsengeschwulste. Diss., Berlin, 1892. Borrmann. Ein Blutgefässendotheliom, etc. Virch. Arch., Band cli. Sup- plementheft, 1898, p. 151. Wachsthum d. Gefässgeschwulste. Virch. Archiv, Band clvii, 1899. - Borst, Max. Das Verhalten der Endothelien bei der acuten und chro- nischen Entzündung. Würzburg, 1897. Verhand. der Physik-Medizin- Gesellschaft zu Würzburg. Bosc et Jeanbrau. Parotid Tumors. Nouveau Montpellier Médical, 1899. Böttcher. Virch. Archiv, Band xxxviii, p. 400. Bottini. Chir, des Halses, 1898. Bovaird. American Journ. Med. Sciences, 1900, p. 377. Briddon. ANNALs of SURGERY, Vol. xxx, 1899, p. 219. Brosch, A. Zur Frage der Entstehung der Nierenzellen aus Endothelien. Virch. Archiv, Band czliv, p. 289. - Du Castel. Cylindrome de la région parotidienne. Annales de Derm. et Syph., 1897, p. 187. Cavazzani, T. Angiosarcoma. Rivista Veneta di Scienze Mediche. Tome xxvi, 1896, p. 405. Chambard. Revue de Méd., 1880. - Chaput. Bull. et Mém. Soc. Anat. de Paris, 1900, ii, pp. 906–908. Collet, André. Des Tumeurs mixtes des glandes salivaires des levres. Paris, 1895. Abstract in Cent. f. Path., 1896. Cone, S. M. Bull. Johns Hopkins Hospital, May, 1898, p. 114. Cordes, H. Ein Fall von Endothel. myxomatodes palati mollis. Deut. med. Woch., 1900, No. 35. (38 MIXED TUMCRS OF SALIVARY GLANDS. Curtis, B. F. The Post-Graduate, 1898, p. 25. Curtis et Phocas. Arch. prov. Méd., 1899. Degen, D. Ein doppelseitiges Sarkom der Parotis. Diss., Freiburg, April, I900. Delabarberie. Arch. gén. d. Méd., 1890. - Dunbar and Potel. Tumeurs bénigne du voile du palais (Fibroma). Echo. méd. du Nord., 1900, p. 588. Dutton, T. E. Liverpool Medico-Chir. Journal, 1898, xviii, p. 369. Eberth and Spude. Familiáre Endothelioma. (In white mice.) Virch. Archiv, Band cliii, 1898. I Eisenmenger. Die plexiformen Sarkome des Gaumens. Deut. Zeit. f. Chir, 1894, Band xxxix, p. 1. Elsching, A. Haemangioendothelioma tuberosum multiplex. Wiener med. Presse, 1896, p. 169, No. 5. (Skin Tumor.) - Ernst. Ueber diffuse multiple Angio-Lipombildung mit Endothel- Wucherungen. Diss., 'Freiburg, 1896. Finger, J. Endothel. 'fibrosum der weichen Hirnhaut und Bemerkungen ūber Endothel und Endotheliome tiberhaupt. Diss., Würzburg, 1897. Fürstenheim, W. Kiemengangshautauswächse mit knorpeligem Gerüst. Jahrb. f. Kinderheilk., 1895, Band xl. - Gaucher. Epithel. prim. de la rate. Thése de Paris, i882. Gevaert. Tumors of Palate. Belgique Médicale, 31 mars, 1898. (Unim- portant paper, chiefly clinical.) Gliuski. Virch. Archiv, 1902, p. 381. Glockner. Riesenzellen u. endotheliale Geschwälste. Beit. v. Ziegler, xxvi, 1899, p. 73. - Gouguenheine et Ripatilt. Contribution a l'étude des tumeurs benignes du voile du palais. Annales des Maladies de l’oreille, du larynx, 1896, xxii, pp. 1-9. Gradenigo. Missbildungen der Ohrmuschel. Archiv f. Ohrenheilk., Band xxxiv, 1893. Griffini and Trombetta. Atti. d. R. Accad. di Turino, Tome xviii, 1883. Gruenhagen, E. G. Beiträge zur Lehre von den Speicheldrüsen Tumoren. Diss., Berlin, 1898. Hamilton. Journ. of Experimental Medicine, 1899. Hammer. Tumoren des Oberkiefers und angrenzender Gegenden. Virch. Archiv, Band czlii, p. 503. Hansemann. Path. Anat. et Hist, des Carcinoms. Deut. med. Woch, No. 33, 1901, p 554. Ueber die Bezeichnung Endotheliom. Deut. med. Woch., 1896, Nos. 4, 14. Verh. d. Naturforscher Versam., Frankfort, 1896. Bösartige Geschwälste, Igo2, p. 62. Haug. Arch. f. Ohrenheilkunde, Band xliii, p. Io, 1897. (Ear cases.) Henkel, Max. Beiträge zur Histiogenese der Parotisgeschwillste. Inaug. Diss., Berlin, 1896. Hess, F. Beiträge zur Geschwälstlehre. Inaug. Diss., Bonn, 1896. Hinsberg. Beit zur Entwicklungsgeschichte u. Natur d. Mundspeichel- drüsengeschwälste. Deut. Zeit. f. Chir., Band li, 1899, p. 281. Beit. z. klin. Chir., 1899, Band xxiv, p. 275. Von Hippel. Ziegler's Beit., 1893. 69 FRANCIS CARTER WOOD. Homp, G. Angiomyxsarkom d. Thränendrüse. Diss., Königsberg, 1896. . Huguenin. Virch. Archiv, 1902, p. 396. Hutchinson, J., Jr. Adenochondroma of the Submaxillary Gland. Trans. London Path. Soc., Vol. xlviii, p. 63. Janssen. Virch. Archiv, 1895, Band czxxix. Jawenko. Mischgeschwälste d. Parotis. Diss., Würzburg, 1897. Jolly, J. Etude Anatomo-Pathologique d'un angiome sarcomateux. Arch. de Méd., Exp., Tome vii, 1895, p. 621. Josuran, C. Veränderung d. Epithels in chron. entziándeten Parotis. Diss., Zürich, 1898. Kaufmann, Parotissarkome. Arch. f. klin. Chir., Band xxvi, p. 672. Klaatsch. Ueber d. jetzigen Stand der Keimblattfrage mit Rücksicht auf die Pathologie. Münch. med. Woch., 1899, No. 6, p. 169. Klein, G. Endotheliale Metastasen des Carcinoms. Verh. d. Deutsch. Naturf. und Aerzte, 1896, p. 18. Koch, U. Ueber Parotis Tumoren. Diss., Freiburg, 1897. Kolossow. Biol. Cent., Band xii, 1892, p. 87. Arch. f. Mik. Anat, Band xlii, p. 318. * - König, F., Jr. Cent. f. Chir., No. 27, 1899, p. 19. Verh. Deut. Gesell. f. Chir. Koslowski, B. S. Chirurgie, 1898, p. 447. (Russian.) Abstract in Cen- tralbl. f. Chir., 1899, No. 9, p. 283. Kretz. Neoplasms in same Individual with Metastases. Wiener klin. Woch., 1892, No. II. Krohn, Wilhelm. Ueber Geschwälste der Submax. Speicheldrüse. Inaug. Diss., Halle, 1899. Kromayer, E. Zwei Fälle von Endothelioma tuben. Coll. des, etc. Virch. Archiv, Band czxxix, p. 282. Zur Histiogenese der weichen Haut- naevi, Metaplasia von Epith. zu Endothel. Dermat. Zeit., Band iii, p. 265 . - Kühne. Mischgeschwälste der Parotis. Diss., Freiburg, 1895. Küttner. Beit. z. klin. Chir., 1896, Band xvi. - Lambrecht et Gaudier. Echo Méd. du Nord., Tome i, p. 19. Landsteiner. Mischgeschwälste d. Speicheldrüsen. Zeit. f. Heilkunde, I90I, p. I. - - - - Langhans. Casuistische Beit. z. Lehre von den Gefässgeschwälsten. Virch. Archiv, 1875. - Limacher, F. Ueber Blutgefässendotheliome der Struma, etc. Virch. Archiv, Band cli, Supplementheft, 1898, p. II.3. . Linser. Verkalkte Epitheliome u. Endotheliome. Beit. v. Bruns., xxvi, I900, p. 595. - Lotheissen, G. Ueber die Geschwillste der Gland Submax. Beit. v. Bruns., 1897, Vol. xix. - - Löwenbach. Z. Kennt. der Geschw. der Submax. Speicheldrüsen. Virch. Archiv, I 50. * * * ... • - . Lübarsch. Ergebnisse. Jahrg, i, Abt. 2, S. 366. Jahrg. ii, S. 592. Zur Lehre v. d. Geschwälsten, 1899, p. 233. Verh. d. Ges. Deut. Naturf. und Aerzte, Band ii, 1895. Arbeiten aus der Hygien. Instituts zu Posen, 1901, p. 205. Zeit. f. klin. Med., 1902, p. 491. 70 MIXED TUMCRS OF SALIVARY GLANDS. Marchand. Verh. d. Deut. Gesell. f. Path., 1899. Mauclaire and Durrieux. Tumeur mixte du voile du palais. Bull. Soc. Anat. de Paris, 1897, p. 609. McWeeny, E. J. Endothelioma (Tumor contained bone and cartilage.) Trans. of Royal Acad. of Medicine in Ireland, Vol. xii, 1895, p. 366. , Melnikow-Raswedenkow. Elastische Gewebe. Beit. v. Ziegler, Band xxvi, 1899, p. 546. Meneau. Un cas d'hematolymphangioma circonscrit. Mém. et Bull. de la Soc. de Méd., etc., de Bordeaux, 1894-1895, p. 334. Mikulicz and Kümmel. Krankheiten d. Mundes, 1898, p. 180. Minot, C. S. Classification of Tissues. Boston Society of Medical Sciences, December 19, 1899. Mixter. Boston Med. and Surg. Journal, 1896, p. 137. Morestin. Traité de Chirurgie, par Dentu et Delbet, 1898, pp. 296 and 406. Morpurgo. Endothelioma. Atti della Accad. delle scienze med. e nat. di Ferrara, Vol. lxix., 1895. Mourie, Pierre. Endothelioma ( ?). Contribution à l'étude du cancer primitif du voile du palais et de la luette. Bordeaux. Thése, 1895. Mulert, Willy. Die gutartigen Tumoren der Zungenbasis. Diss., Würz- burg, 1895. Nasse. Arch. f. klin. Chir., Band xliv, p. 233. Nicoladoni, C. Adenom aus der Wange. Deut. Zeit. f. Chir, Vol. i., 1872, D. 432. v. Ohlen. Beit. v. Ziegler, Band xii. Otto, Ernst. Diffuse multiple Angio-Lipombildung mit Endothelwucher- ung. Diss., Berlin, 1895. - Parmeijer. Mixed Tumor of Parotid. Diss., Leyden, Igor. Peters. Ueber Path. Cölomepithel Einstulpungen bei menschlichen Em- bryonen. Verh. d. Deut. Gesell. f. Gynak, 1897, p. 524. Picou et Ramond. Splenomegalie primitive. Arch. d. méd. exp., 1896, p. I69. Pitance. Thése de Paris, 1897. (Quoted by Morestin.) Polak, Daniels. Stroma in Sarcoma. Virch. Archiv, Band clzv, p. 238. Pollack. Arbeiten aus der Hygienischen Instituts zu Posen, 1901, p. 160. Pollman, L. Ein Endotheliom der Pleura und des Peritoneums. Ziegler's Beiträge, 1899, p. 37. - Ponsot, R. Tumeurs de la glands sous maxillaire. Thése, Lyon, 1894– I895. Pupovac, D. Ein Beitrag zur Casuistik und Histologie der sogen. Endoth. Deut. Zeit. f. Chir., Band xlix, 1898, p. 77. Pussacq-Larcebau, T. B. Etude sur les tumeurs bénignes du Pharynx buc- cal et inferior. Thése, Bordeaux, 1895. Putiata. Virch. Archiv, Band lxix, p. 245. - - Ribbert. Deut. med. Woch., 1896, No. 1. Mallory Stain for Fibrous Tis- sue. Cent. f. Path., 1896, p. 427. Vierteljahresschrift d. Naturf. Ge- sell. Zürich, 1896. * . r Riedel, W. Pap. Endothelioma. Diss., Greifswald, 1898. 7, 1 FRANCIS CARTER WOOD. Ritter, C. Fettgehalt in den Endotheliomen des Knochens. Deut. Zeit. f. Chir., Band l, 1899, p. 349. Roughton. Tumors of the Palate. (Nothing new.) British Journal of Dental Science, Vol. xxxix., 1896. Sailer. Primary Endothelioma of the Left Superior Pulmonary Vein. Contribution from the William Pepper Laboratory of Clinical Medicine, I900. Schellong, C. Ueber Geschwiilste des weichen Gaumens. Diss., Göttingen, 1897. Schwalbe, E. Endothelioma. Vich. Archiv, Band xlix, p. 451. Schwalbe, G. Das äussere Ohr. Bardeleben's Handbuch der Anatomie, 1898, p. 149. . Scriba, M. Endothelioma. (Paper read at the XIIth Congress, Moscow.) Centralbl. f. Chir., 1897, p. Io99. Spuler. Ueber den feineren Bau der Chondrome. Beit. v. Zieg., 1902, p. 263. - SSobolen. Zur Lehre von d. Endothel. Neubildungen. Virch. Archiv, Band clxi, p. 50. (Doubtful case from intestine.) Steinhaus. Mischgeschwälste der Mundspeicheldrüsen. Virch. Archiv, Band clxviii, 1902, p. 233. Stengel, Hans. Ein Fall von Endothelioma Sarcomatosen des Oberkiefers. Diss., Würzburg, 1896. Sutke, H. Beit. z. Keimt. d. Metastasen d. prime. Nieren-Carcinom. Virch. Archiv, Band clzviii, p. 305. Tanaka. Ueber die klinische Diagnose von Endotheliomen und ihre eigen- thūmliche Metastasenbildung. Deut. Zeit. f. Chir., 1899, p. 209. Tervaert and de Jong. Arch. f. Ohrenheilk, Band xliii, 1897, p. 53. (Ear cases.) Thorel. Ergebnisse, Lubarsch-Ostertag, 1898, p. 235. Tusini, G. Endothelioma. (Two parotid tumors.) Festschrift for Du- rante, Vol. ii, p. 85. Unna. Epithel. Abkunft der Naevuszellen. Virch. Archiv, Band cycliii. Vincentiis, Carlo. Endothelioma adiposa. Rivista Clinica, 1893, No. 7. Volkmann. Ueber endotheliale Geschwälste. (Very complete literature up to 1896.) Deut. Zeit. f. Chir., Bol. xli., 1895, p. 1. Waldeyer. Endothel. und Epithel. Cinquantenaire de la Soc. de Biol. Vol. Jubilaire, 1899, p. 531. Arch. f. mik. Anatomie, Band li, 1900, p. 1. Williams, Alice. Journal of Experimental Medicine, Vol. v., 1900, p. 131. Williams, H. N. Festschrift for Welch, 1901. - Wilms. Embryome und Embryoide Tumoren des Hodens. Deut. Zeit. f. Chir., Band xlix, 1898, p. I. Mischgeschwälste, Heft 3. Winkler, Karl. Betheiligung des Lymphgefässsystems an der Verschlep- pung boartiger Geschwillste. Diss., Breslau, 1898. Wlasso. Sarcome angioplastique. Virch. Archiv, Band clzix, p. 220, 1902. Wojtkiewicz, S. Knorpelhaltigen Auswächse am Halse. Diss., Zürich, 1900. (Contains a good discussion of the literature of the subject.) Ziegler, H. E. Coelemfrage. Verh. d. Deut. Zoolog. Gesell, 1898. 72 THE AGGLUTINATION OF THE PNEUMOCOCCUS WITH CERTAIN NORMAL AND IMMUNE SERA AUGUSTUS WADSWORTH, M.D. (Alumni Fellow in Pathology, Assistant in Bacteriology and Hygiene, College of Physicians and Surgeons, Columbia University, N.Y.) (From the Department of Pathology, College of Physicians and Surgeons, Columbia University, N.Y.) THE AGGLUTINATION OF THE PNEUMOCOCCUS WITH CER— TAIN NORMAL AND IMMUNE SERA. * AUGUSTUs WADSworTH, M.D. (Alumni Fellow in Pathology, Assistant in Bacteriology and Hygiene, Col- Żege of Physicians and Surgeons, Columbia University, M. Y.) (From the Department oſ Pathology, College of Physicians and Surgeons, Columbia University, N.Y.) In the course of some experiments on pneumococcus in- fection in the immunized rabbit, it was important to have some criterion, such as the agglutination reaction, of the degree of immunization developed in the animals. The methods used by other observers to determine the pneu- mococcus agglutination were tried. Some of these were wholly unreliable; others were suggestive, but uncertain; none were found trustworthy or sufficiently delicate when practised as a routine. After a long series of experiments, the chief faults in these previous methods were determined and in a large measure eliminated. As a result of these studies, I devised a new technic which has given more pre- cise reactions than any of the methods hitherto used and which has proved sufficiently delicate to be, I believe, of broader and more practical service. Previous methods. – The susceptibility of some animals, such as the rabbit, to pneumococcus infection as compared with the relative insusceptibility of certain other animals, such as the dog, has been determined from the results of animal inoculation. These observations were accompanied and followed by more accurate studies of the action of the sera of different animals on the growth of the pneumococcus. Metchnikoff, Behring and Nissen, Kruse and Pansini, Wash- bourne and Eyre, and others have noted that in general the blood serum of the rabbit has practically no bactericidal action on the pneumococcus; that of the guinea-pig varies, 1 Read before the American Association of Pathologists and Bacteriologists, May, I903. Received for publication July 9, 1903. (228) N. THE AGGLUTINATION OF THE PNEUMOCOCCUS. 229 some exerting slight action; while the blood Sera of cats, dogs, rats, squirrels, sheep, and horses are bactericidal in varying degrees, and less favorable to the growth of pneu- mococci. Kruse and Pansini noted especially the more marked formation of chains and bunches in the sera of the less susceptible animals. These variations in the growth of the pneumococcus have been further studied by Bezancon and Griffon, and also by Huber. The growth in bunches and chains was considered so marked and characteristic in the sera of specifically immunized animals that these ob- servers describe it as a pneumococcus agglutination. Bezan- con and Griffon inoculated one to two cubic centimeters of clear undiluted blood serum with pneumococci and observed the character of the growth at the end of fifteen or sixteen hours' incubation. The growth formed a membrane at the bottom of the tube, composed of bunches and chains, and this so-called agglutination was easily seen with the unaided eye; or the growth clouded the medium, and the small bunches and short chains could only be seen with the microscope. This method was applied clinically by them to one hundred and eighty-six patients — sixty-four of undoubted pneu- mococcus infection, twenty-two uncertain cases, and one hundred controls. In the serum from cases of pneumococ- cus infection the growth in bunches and chains, such as was found in the immune sera, could be distinguished easily and constantly from the fine clouding produced by the growth in the controls of normal serum. They, therefore, considered this so-called agglutination of practical diagnostic value. Although these observers noted this reaction in dilutions of immune sera up to one to fifty, they found that the results were very uncertain. Neufeld studied the agglutinative action of the serum of immunized animals on pneumococcus cultures in broth by the usual methods, and was thus the first to note typical agglutin- ation reactions in various dilutions, both microscopically and macroscopically. The maximum observed was one to fifty. He found that the agglutination took place in one-quarter to One-half hour after the mixtures were made. In order to 230 WADSWORTH. obtain good reactions the serum must be drawn from highly immunized animals and must be fresh. If allowed to stand for two weeks or even less he found that it was no longer active. He also found the activity of the sera from different animals to vary greatly and the reaction was often uncertain. In Neufeld’s experiments different cultures of pneumococci agglutinated equally well. Compared with the typhoid ag- glutination, Neufeld found that the pneumococcus reaction differed in some respects. Thus he claimed that heated (IOO" C. for 3 hours) pneumococcus cultures agglutinated as well as the fresh. Furthermore, normal sera from the rabbit or man, in his observations, always failed to agglutinate pneumococcus cultures. These differences, as will subse- quently be shown, are only apparent, not real, and are due to the lack of delicacy in the method. Criticism and corroboration of the work of other observers. — However accurate may be the observations previous to Neufeld on the above noted growth of the pneumococcus in chains and bunches, and however much these may suggest the existence of special immune substances in certain sera, they were, nevertheless, based on rather crude experiments, allowing of much latitude in their interpretation. In all such experiments with bacteria the special selection attained by the micro-organism from its previous environment and its ability to adjust itself to sudden changes, are variable and disturbing factors not easily eliminated. But, aside from this, there are other and often indeterminate conditions, such as the amount of oxygen absorbed by the serum, its sugar con- tent, reaction, and so on — conditions to which the pneumo- coccus is extremely sensitive, and which in the ordinary artificial media suffice to bring about many of the above noted phenomena in the growth of the pneumococcus. I have grown the pneumococcus in the fresh, undiluted sera of the cat, dog, sheep, bullock, horse, and man, and in various ascitic and pleuritic fluids, and have been unable to note any consistent variation in the character of the growth which I could safely correlate with any special resistance or lack of susceptibility to infection in the animal. Furthermore, I have THE AGGLUTINATION OF THE PNEUMOCOCCUs. 231 grown the pneumococcus in sera from immunized rabbits and obtained the fine clouding noted in other sera. This, how- ever, may be considered more or less exceptional, for it is certain that short chains or small clumps were more often found in the sera of highly immunized animals. On the other hand, a rapid growth in normal rabbits' (or other) serum under most favorable conditions — as to oxygen, sugar, and so forth—frequently produced a flocculent precipitate which gave a voluminous sediment, similar to the appearances Bezan- con and Griffon describe as occurring only in immune serum. These facts suffice to indicate the extreme care necessary in making observations in this field and the great chance of error. Neufeld’s observations of the typical agglutination of pneumococcus cultures by immune sera was easily confirmed in my experiments. For example, the serum from two im- munized rabbits" when mixed with forty-eight to seventy-two- hour broth cultures of pneumococci in dilutions of one to ten and one to twenty, showed a clumping not present in the one to sixty or control tubes; but in dilutions of one to one the rapid lysis obscured the agglutination phenomena. The lysis occurring in dilutions of one to one or less was studied in the hanging drop and by stained microscopic preparations. With methylene blue the cells stained very faintly, and with the gentian violet copper sulphate stain recommended by Hiss, empty capsules, degenerated frag- ments of cells in capsules, in clumps or singly were found. In short, these were the changes first described by Radziew- sky in his studies of the Pfeiffer phenomenon with the pneu- mococcus, and later by Neufeld. The swelling of the cells which one sees in the hanging drop and noted by the above cited observers as a first stage in the degeneration, it seems probable, is the usual normal development of the capsule in the presence of the serum described by Hiss, who brought out the fact that capsules may be demonstrated on pneumo- Cocci growing in artificial media, if only a drop of serum be * These animals, II6 and II'7, were immunized in the course of two weeks by inocu- lation with boiled cultures of pneumococci, 5 cc. subcutaneous and 2% ce. intravenous, and with living cultures, I cc. subcutaneous and I ce. intravenous. At the site of inoc- ulation one animal developed a subcutaneous abscess from which it recovered. 232 WADSWORTH. added in preparing the coverglass. This lysis of the pneu- mococcus cells obscuring the agglutination was, I found, easily eliminated by heating the serum to 56° C. for one hour, and thus destroying the complement. With old sera this was un- necessary, as the complement had already degenerated. I have thus been able to corroborate the general observa- tions of others as to the bacteriolytic and agglutinative action of specific immune Sera upon the pneumococcus; but as a practical routine procedure for determining with delicacy the activity of different normal or immune sera, the methods were complete failures in my experience. This was due to many variable and often indeterminate conditions, but chiefly to the difficulty of distinguishing slight reactions and to the very considerable immunity which had to be developed by the animals before the presence of agglutinins in the blood serum could be detected. The great drawbacks to precise observation were the biological condition of the pneumo- cocci and the small quantity of growth present in the usual broth cultures. It was hoped that these difficulties would in a measure be overcome, and that better and more precise results might be obtained, by concentrating the cultures at an early stage of the growth so that a greater number of fresh cells would be present in the same amount of fluid. The new technic. — A peptone broth made from meat infusion which had been carefully neutralized before boiling has, in my experience, proved an optimum artificial fluid medium for growing the pneumococcus, especially if allowed to stand in the cold and absorb oxygen. In a flask contain- ing approximately two hundred cubic centimeters the maxi- II) Ul Ill growth is usually reached in from twenty-four to thirty-six hours." At this stage the culture is centrifugalized, * If glucose be added to the media, this maximum growth is reached much earlier, and there is an increasing quantity of precipitate formed as the growth con- tinues. The precipitate increases the amount of sediment, and carries down many of the pneumococcus cells. Such cultures should be centrifugalized before much sedi- ment appears, and only the top whitish layer used, as this is composed chiefly of pneu- mococcus cells. My experience with the glucose growth as compared with the growth in plain broth seems to indicate that if taken at the best stage the glucose growth gives the more delicate reactions; but the growth in plain broth, without glucose, gives THE AGGLUTINATION OF THE PNEUMOCOCCUs. 233 the clear fluid decanted, and the sediment shaken with about fifteen cubic centimeters of isotonic (.85 per cent) salt solution. The few clumps not broken up sink rapidly, leaving a dense, finely divided suspension of pneumococcus cells less than forty-eight hours old." Dilutions with the sera to be studied may then be made in small, slender tubes, and observed for some twelve to eighteen hours at 37° C. The more marked reactions may be complete in five to six hours or less, but twelve or more hours are often required to bring out the more delicate tests. The serum was at first heated to prevent lysis, but as the reactions were studied in so much higher dilutions, this was abandoned and the fresh serum used, as is customary in typhoid and other agglutinations. Sera stored in the ice- box were found to be active for four or more months; the lytic power was gone, and the agglutination did not appear in high dilution, but in low dilution immediate and complete reactions took place. Previous methods were not sufficiently accurate to establish this, and Neufeld drew the erroneous conclusion that in two weeks the pneumococcus agglutinin had degenerated. In order to eliminate the growth of the organisms, the salt Solution suspension was heated for one hour at 56° C.” Comparative tests of this procedure were made with strongly active sera. In low dilutions the reactions were marked and there was little difference apparent; but in high dilution the heated cells did not show the clumping as well as the un- heated cells. In the low dilution the agglutination appeared early, before much growth had developed, and in the high dilutions the amount of growth was slight, so that it rarely in- terfered with accurate observation. Occasionally, however, the pneumococcus cells, when very fresh and in large numbers, developed with great rapidity, clouding the mix- tures, and forming a flocculent precipitate which gave in a possibly the more stable and better average. This is doubtless due to the rapid degen- eration of the cells in glucose cultures. - 1 If much older growths or suspensions be used, the cells settle too rapidly, render- ing the comparison of tests with controls difficult and uncertain. *An exposure of twenty minutes to 52° C. usually kills the pneumococcus. 234 WADSWORTH. few hours a voluminous sediment." This might easily be taken for an agglutination, were it not that the growth is always marked and the control mixtures with normal serum have the same appearance. The fact that this may occur is important, as suggesting the chances of error and as indicat- ing the necessity of control mixtures with normal sera, if ob- servations are to be made after much growth has taken place. Results obtained with the new technic. — I have studied the agglutination of the pneumococcus in the blood serum from various normal animals and from a large number of specifically immunized rabbits by means of this new technic. It is not necessary to give in detail all the experiments; only such have been selected as serve to demonstrate the com- parative value of the method and certain other points of interest. Rabbits 140 and 141 were immunized in the following Iſla I) ſle I - I2/19. 5 cc. dense suspension dead pneumococcus cells, Subcutaneous. . I 2/22. 5 cc. moderately dense suspension dead pneumo- coccus cells, intravenous. 12/24. 5 cc. luxuriant broth culture heated to IOO" C., subcutaneous. - . - h 12/26. 3 cc. luxuriant broth culture heated to IOO" C., in- travenous. 12/31. .5 cc. luxuriant broth culture, fresh, subcutaneous. I/5 2 “ & K ( { ( & ( & ( & 1/8. 1.5 “ ( ( { { { { “ intravenous. 1/17. Rabbit 140 died from pneumococcus infection. Rabbit I 4 I lived, and January I4 was bled from the ear vein and the serum used for agglutination tests. 1 This was most marked when the serum and suspension of pneumococci were very fresh, and when stored in the ice-box a short time, absorbing considerable oxygen. By warming the fluids, driving off the oxygen, and by using deep, slender tubes, the diffi- culty was, in a measure, overcome. As will subsequently be shown, a precipitation may develop in the mixtures of sera with salt-solution suspensions. This is, however, a more or less specific phenomenon, and not to be confused with the ordinary precipita- tion which often occurs in sera and in some artificial media from pneumococcus growth. - THE AGGLUTINATION OF THE PNEUMOCOCCUS. 235 1/2 I. A serum broth culture of the same race of pneumo- coccus used in the process of immunization, but recently isolated from an animal, was mixed with serum of Rabbit 141 in the following experiment: Serum. Salt Sol. Culture. Dilution 370 C. 12 hours. .3 CC. O .3 CC. I—I +++ ... “ .2 CC. .3 “ I-6 –H–H .o; “ .25 “ .3 “ I-2O + slight Control .3 “ .3 “ ; O O 2/2. A dense, fine suspension in salt solution (.85 per cent) of centrifugalized glucose broth culture sediment, less than forty-eight hours old, was substituted for the serum broth culture. Pnc. Dilution Serum. Salt Sol. Suspen. 37°. 2 hours. 5 hours. 12 hours. 20 hours. 1/5 cc. qs. ad. .5 cc. I. cc. 1-7% ++++- Prac. Complete Complete I/IO 6& £ & 6 & .5 & 4 I. “ I—I 5 ++++ & & & & & & I/I5 6 & 6 & 6 & .5 66 I. “ I–22% +++ —H++- & 4 I/30 £& & 6 & 6 .5 & 6 I. “ I-45 ++ –H–H & 4 1/60 & & 6 & 66 .5 66 I. “ I-90 P + - V 6 & 1/45 “ “ “ .5 “ 1. “ 1–67% Complete I/90 & 6 , & 6 & 8 .5 66 I. “ I–I35 & & 1/18O & 6 & 66 .5 66 I. “’ I-270 6 & Control .5 “ 1. “ O Slight sedi- O ment only. 2/6. Experiment continued with fresh culture suspen- Sions. Pnc. Dilution Serum. Salt Sol. Suspen. 379. I hour. 4 hours. 20 hours. 1/180 cc. qS. ad. .5 cc. I. C.C. I-270 o o Prac. Complete I/3oo “ “ “ .5 “ I. “ I-450 O O ++- I/450 & & 66 & 8 •5 £& I. “ I-675 - O O —H P 1/900 “ “ “ .5 “ I. “ I–I,7Oo o O O Control .5 “ I. “ O O O O * 2/26. Experiment was repeated. Centrifugalized sedi- ment from plain broth culture was shaken in .85 salt solu- tion and divided into two portions, one of which, B, was heated 52°C., one-half hour. - 236 WADSWORTH. Serum. Salt Sol. Cult. Suspen. Dilution 370. 5 hours. 18 hours. * A .05 cc. qS. ad. .25 cc. .5 CC. I–I5 P +++ or “ “ “ .25 “ .5 “ I-75 P ++ .O I £ 6 & 4 & & .25 66 .5 & 4 I-75 P –H–H .OO2 ** “ “ .25 “ .5 “ I-375 O –H .OO2 “ “ “ .25 “ .5 “ I-375 O —H. .OOO4 “ “ “ .25 “ .5 “ I–I,875 O O B .OI “ “ “ .25 “ .5 tº I-75 O PO .OI “ “ “ .25 “ .5 “ I-75 O PO .OO2 ** “ “ .25 “ .5 “ I-375 O O .OO2 “ “ “ .25 “ 5 “ I-375 O O .OOO4 “ “ “ .25 “ 5 “ I–I,875 O O Controls: º Salt Sol. Cult. Suspen. Dilution, 379.5 hours. is hours. A. - O .25 CC. .5 cc. O O O Ascites Fld. .O5 cc. qS. ad. .25 “ .5 “ 1–15 O O o; “ “ “ .25 “ .5 “ I–I5 O O Careful study of these reactions under the microscope in hanging drop and stained preparations brought out many points of interest. The cells, morphologically and tinctorially normal, are usually single or in pairs, rarely in clumps; whereas the clumps are composed chiefly of poorly staining cells whose morphology is often made out with difficulty. It seems fair to assume that the agglutinin acts principally on the degenerating cells when present. The intact cells develop and, if sufficient nutriment is present, cloud the fluid as usual. On the other hand, normal fresh cells may be rapidly clumped in low dilutions of very active immune sera, and the reaction may be complete in a few minutes under these circumstances. It is evident from the above cited and other similar experi- ments that the agglutinative action of immune sera with heated pneumococcus may not be detected in as high dilu- tions as with fresh, unheated cells, and that the pneumococcuS agglutinin is fairly stable and does not disappear from the immune sera for a considerable time. In short, the pneumo- coccus agglutinin conforms to the general laws governing typhoid and other agglutinins, except as regards the action THE AGGLUTINATION OF THE PNEUMOCOCCU.S. 237 on motility. The greater delicacy sometimes seen in the re- actions on glucose growths is also suggested in the above cited and other series of experiments. , The agglutinative action of sera from normal animals on the pneumococcus. – The normal serum of the rabbit always failed to agglutinate the pneumococcus cells. The growth of the organisms with marked formation of precipitate gave a flocculent sediment in some instances, but this was not con- sidered a positive reaction. The normal blood sera of the cat and dog also failed to agglutinate. Normal bullocks' serum, however, gave some very marked reactions, even in dilution of one to fifty. Ascites fluids tested failed to agglutinate in dilutions of one to five and one to fifteen; but normal human Sera gave reactions in dilution up to one to ten in less than eighteen hours; one to thirty failed to agglutinate. The agglutinative action of the blood serum from pneu- monia patients on pneumococci. — Through the kindness of Dr. F. C. Wood, I obtained a specimen of blood from a case of lobar pneumonia. The serum gave a marked reaction in dilution of one to ten after five hours' incubation. The quantity of serum was small, and it was not possible to test the reaction further. Two specimens were obtained from two other cases of pneumonia through Dr. G. B. Rhoades. These yielded enough serum to retest my results. Dilution 370. 4 hours. 6 hours. 24 hours. I. Pneumonia Serum . . . . . . I-9 P ++ —H++ I–I8 P –H –H–H I–27 O +- ? —H-H Controls: Salt Sol. . . . . . . . . O O O O Ascites Fluid . . . . . . . I-3 O O O Dilution 370. 3 hours. 5 hours. 24 hours. II. Pneumonia Serum . . . . . . I–IO P —i-–H +++- - I–2O O –– ++--- I-5O O O —H- I—IOO O O ? slight Controls: Salt Sol. . . . . . . . . O O O O Normal blood sera, rabbit . . . I-8 O O O & 6 & & “ man . . . . I–IO O O + O O O I–50 238 WADSWORTH. The close relation that has always been thought to exist between the pneumococci and the streptococci suggested testing the action of pneumococcus serum on suspensions of streptococcus cells in Salt solution and, in reverse fashion, the action of streptococcus serum on pneumococci. The agglutinative action of pneumococcus serum on streptococci. — A strongly active serum from Rabbit 141, immunized against the pneumococcus, was added in various dilutions to a salt solution suspension of streptococcus growth. These mixtures settled rapidly, and the growth of the streptococcus in chains gave a granular or flocculent sediment. This appearance was more marked in the tests than in the controls with normal serum, but the difference was slight and not considered definitely significant. The agglutinative action of Streptococcus serum on pneumococci. —The serum from a rabbit immunized with centrifugalized cultures of streptococci was added to the salt solution suspensions of pneumococcus cells in the following dilutions: Streptococcus - Salt Sol. Susp. Serum. Salt Sol. Pneumococci. Dilution 370. Io hours. .25 CC. O .25 CC. I — I + + + ... I 66 qS. ad. .25 cc. .25 “ –5 —H + o; “ “ “ .25 “ .25 “ I—IO + .oi 25 “ “ “ .25 “ .25 “ I-40 O There was, apparently, considerable growth and some pre- cipitate formed even as early as ten hours, and as this was also found in the controls of normal rabbits' serum, the experi- ment was supplemented by tests on the pneumococcus cells heated to 52 degrees for one-half hour, thus preventing growth. In this experiment agglutination took place in dilutions up to one to ten after eighteen hours' incubation. Pneumococcus precipitin reaction. — Neufeld is, I believe, the only observer who has described a pneumococcus pre- cipitin reaction. He found that if normal rabbits' bile is added to a fresh pneumococcus culture in the proportion of THE AGGLUTINATION OF THE PNEUMOCOCCU.S. 239 one drop to one cubic centimeter of culture, the media be- comes clear and the cells are no longer demonstrable. If such cleared culture is added to the serum of highly immun- ized rabbits, a precipitation takes place. The greatest dilution at which the reaction developed was one to nine. This pre- cipitate Neufeld found was composed of highly refracting granules of variable size and easily observed under the micro- scope. These observations of Neufeld I was able to corrobo- rate with the serum from only one of my immunized animals; but by dissolving the pneumococcus cells in the dense salt solution suspensions with normal rabbits’ bile, I obtained a clear sterile fluid which formed voluminous precipitates with several of my pneumococcus sera in various dilutions. Pneumococcus Solu. (bile) Serum. Salt Sol. Pneumococcus. Dilution 379. I hour. 12 hours. .25 cc. O .25 CC. I — I + +++++ ... I 66 qS. ad. .25 cc. .25 “ I-5 + ++++ .O5 66 & 6 66 .25 66 .25 66 I—IO O +++ .O2 6 & £6 & 8 .25 6 & .25 64 I–25 O ++ * .OI 66 £ 6 & 8 25 £ 6 .25 “ I–50 O + .0075 “ “ “ .25 “ .25 “ I–66.6 O +? .005 “ “ “ .25 “ .25 “ I—IOO O O Controls .25 “ .25 “ O O O Normal Pneumococcus Serum. Salt Sol. Bile and Broth. Dilution 370. I hour. I2 hours. .25 CC. O .25 cc. I — I O O Normal Bile and Salt Sol. .25 CC. O .25 cc. I-I O O In order to simplify the procedure and remove all question of the part played by the bile, another method was adopted. The centrifugalized fresh pneumococcus cells were, in order to increase the plasmoptysis, first shaken with 17 per cent salt solution; then this was made isotonic (.85 per cent) by the addition of water, and filtered. The filtrate" thus obtained * It may be well to suggest that the precipitate formed from the media as a result of the growth may be partially redissolved by the different salt solutions, and some may be retained in the filtrate. How this, if true, might influence the significance of the results of these precipitin reactions is at present uncertain. 24O WADSWORTH. gave the precipitin reaction with immune sera in dilutions up to one to fifty, as did the bile solutions. The filtrates prepared in this way varied greatly, but even the weak- est showed some precipitate when mixed with the very active immune sera. Pneumococcus Salt Sol. Dilution - Immune Serum. Salt Sol. Ext. Pneumoc. 379. 34 hr. 1 hr. 5 hrs. 16 hrs. .25 CC. O .25 CC. I— I ++ +++ ---|--|--|- .O5 “ qS. ad. .25 cc. .25 “ I–IO –– ++ +++ — ..or 25 “ “ .25 “ .25 “ I–4O O O + + .OO5 “ “ .25 “ .25 “ I—IOO O O o +? OOI 25 “ “ .25 “ .25 “ I–4OO O O O Serum from Hu- man Prieumonia. I part O I part I — I + ++ ++ .O5 cc. qs. ad. .25 cc. .25 CC. I — IO O –– ++ .OI “ “ .25 “ .25 “ I–50 O P + Normal Human Serum. I part O I part I — I O P ++ .O5 cc. qs. ad. .25 cc. .25 C.C. I—IO O P + Normal Bullock Serum. .25 cc. O .25 cc. I — I O P —#- .o; “ qs. ad. .25 cc. .25 “ I—IO O O O Normal Rabbit Serum. .25 CC. O .25 cc. I-1 O O O In these and other similar experiments the precipitin reaction was obtained with pneumococcus immune rabbit serum in dilutions of one to forty or more; one to one hun- dred being doubtful. In tests of the serum from a case of lobar pneumonia precipitates were formed with dilutions of one to one in less than one-half hour incubation, with dilu- tions of one to fifty in sixteen hours, and with dilutions of one to one hundred very slight after twenty-four hours' incubation. Normal human serum, diluted one to ten, gave the reaction in sixteen hours; normal bullocks' serum, diluted one to five, in sixteen hours. With a weaker filtrate no reactions were THE AGGI,UTINATION OF THE PNEUMOCOCCU.S. 24. I noted with normal human serum, but the normal Sera of the bullock, cat, and dog each gave the reaction in dilutions of one to one. With very active immune serum, diluted one to ten, a precipitate was formed with this filtrate after three hours' incubation. Normal rabbit serum always failed to give any reaction with these filtrates. Once, however, a temporary clouding was noted as a ring at the junction of the two fluids in a one to one dilution. This disappeared on shaking the tube. Serum from a rabbit immunized against streptococcus gave a marked reaction with the strongest filtrate in dilutions of one to one after ten hours' incubation. This was confirmed by subsequent tests with another filtrate, but failed to develop with the very weak filtrates. Briefly summarized the results of my experiments show : I. That peculiar growth phenomena occurring in pneumo- coccus cultures in serum are not reliable tests of the presence of agglutinins or other adaptive substances. 2. That the presence of these substances in very active sera may be demonstrated, as was first shown by Neufeld, by the usual agglutination methods, but that as a delicate or practical routine procedure this method was a failure. 3. That by concentrating the pneumococcus cells in an isotonic solution, as by centrifugalizing broth cultures and Shaking the sediment with a small quantity of salt solution, a practical and accurate method is available for the precise study of the pneumococcus agglutination reaction. 4. That by this method an agglutinative action, heretofore not detected, has been demonstrated in the blood serum of various normal animals and in certain specifically immune Sera. 5. That it has been possible to obtain the pneumococcus precipitin reaction, not only with cultures cleared by normal bile, as did Neufeld, but with filtered salt solution extracts of the pneumococcus cells. 6. Finally, that precipitin reactions, heretofore not de- tected, have been demonstrated with the blood serum of Various normal animals and with certain immune sera. 242 WADSWORTH. BIBLIOGRAPHY. Behring and Nissen. Zeitschrift f. Hygiene, 1890, Vol. viii, s. 412. Bezancon and Griffon. Annales de l’Institut Pasteur, 1900, Vol. xiv, P. 449. Hiss. Centralblatt f. Bakt., 1902, xxxi, p. 302. Science, 1902, Mar. 7, p. 367. Huber. Centralblatt f. Innere Med., 1902, xxiii, 417–421. Kruse and Pansini. Zeitschrift. f. Hygiene, 1892, Vol. xi, p. 279. Metchnikoff. Annales de l’Institut Pasteur, 1891, Vol. v., p. 474. Neufeld. Zeitschrift f. Hygiene u. Infections-Krankheiten, 1902, Vol. x1, p. 54. \ Radziewsky. Zeitschrift f. Hygiene u. Infections-Krankheiten, 1901, Vol. xxxvii, p. 24–28. THE JOURNAL OF MEDICAL RESEARCH, October, 1903, Vol. X., No. 2. Extracted from The American Journal of the Medical Sciences, May, 1904 EXPERIMIENTAL STUDIES ON THE ETIOLOGY OF ACUTE PNEUMONITIS.; BY AUGUSTUS WADsworth, M.D., ALUMNI FELLOW IN PATHOLOGY; ASSISTANT IN B ACTERIOLOGY AND FIYGIENE, DEPART• MENT OF PATHOLOGY, COLLEGE OF PHYSICIA NS AND SURGEONS, COLUMBIA UNIVERSITY, N. Y. INTRODUCTION. ACUTE exudative inflammation of the lungs may develop as a metastatic lesion secondary to disease processes in other parts of the body, or it may accompany acute interstitial processes arising usually from the pleura, or it may arise as a primary infection of the lung forming patchy bronchopneumonic or diffuse lobar lesions. The bacterial incitants of acute primary infectious processes in the lung are now well known. Our knowledge of the essential condi- tions which determine the extent and nature of the disease processes in the lung and elsewhere, however, lacks the precision which can come only from exact experimental data. This may be attained only when it is possible to reproduce at will the various disease processes in animals. With pneumococcus pneumonia this has not heretofore been satisfactorily accomplished. As a result of a long series of experiments, I have finally determined a method by which the diffuse lesions of pneumococcus pneumonia may readily be incited in the rabbit. This method it is the purpose of this paper to record; but the failures and successes of others have in such large measure contributed to my final results that they may with advantage be briefly considered. The significance of this previous work, how- ever, may not be fully appreciated without summing up our knowl- edge of the distribution and disposal of bacteria in general in the respiratory apparatus, of the bacterial incitants of pneumonia, and of the paths of infection. BACTERIA OF THE NORMAL RESPIRATORY APPARATUS. Inspired foreign particles and bacteria, as shown by the researches of Arnold” and many others, are deposited chiefly on the moist surfaces of the upper respiratory passages, although they may occasionally be car- ried to the alveoli of the lung. For most bacteria the healthy mucous membrane is an unfavorable environment, yet this varies greatly in different parts of the tract and with different species of bacteria. In the mouth and nares bacteria are being constantly deposited on the mucous membranes. As indicated by the researches of Thomson and Hewlett,” Wright,” Nenninger,” Müller,” and Lascht- Schenko,” some of them disappear very rapidly; others may develop and exist as harmless parasites. These secretions thus contain * Read in abstract before the American Association of Pathologists and Bacteriologists, April, 1903. 2 WADSWOIRTEI: ETIOLOGY OF ACUTE PNEUMONITIS. many different species of micro-organisms, among which the bac- teria commonly associated with pneumonia—the pneumococcus, streptococcus, staphylococcus, and the bacillus of Friedländer— have often been found. Suffice it to note that the pneumococcus was present in a large proportion of Sternberg's cases,” 20 per cent. of Fraenkel’s cases,” 17 per cent. of Besser's cases,” 20 per cent. of Wolf's cases,” 30 per cent. of Bein's cases,” 55 per cent. of Miller's cases,” 4 per cent. of Neumann's normal cases,” and in 100 per cent. of the forty tonsils examined by Bezancon and Griffon.” Netter” divided his cases into those with and those without a history of pneu- monia, and found the pneumococcus in four out of five and in one out of five cases, respectively, and he also claims to have found the organisms virulent five, ten, or more years after a pneumonic infection. Furthermore, it is significant to note that it has been experimentally shown by Guarnieri,” by Bordonni-Uffredozzi,” by Grawitz and Steffen,” by Spolverini,” and by Ottolenghi" that the pneumococcus in sputum may remain alive and virulent for two to four months. In the trachea bacteria are also deposited on the mucous mem- branes,” but under normal conditions these are comparatively few in number and are rapidly disposed of by the cilia, and also by the action of the mucus, which, in the researches of Wurtz and Lemo- yez" with sputum, of Walthard,” of Stroganoff,” and of Mori- sani” with the uterine secretions, and more recently of Arloing” with pure mucus, has shown marked bactericidal qualities. Examina- tion of the human trachea at autopsy, as in the studies of Baum- garten," Hoffmann,” v. Besser,” Ritchie,” Claisse,” and Durck,” has revealed the presence of micro-organisms in a considerable pro- portion of the cases, but the trachea and lung at autopsy, although of normal appearance, may not be considered to be under normal con- ditions. Normal conditions, it is evident, are more closely approxi- mated in the researches conducted on animals, as in the studies of Hildebrandt.” Klipstein,” Gobell,” Rosenthal,” Beco,” Butter- sack.” Boni,” Barthel,” and Calamida and Bertarelli.” In these examinations of the trachea of animals, bacteria have not often been found, and the lower portions of the trachea and the bronchi have thus been considered to be practically sterile.i. * As suggested by the researches of Slawjansky107 and of Kuttner of with the injection of foreign substances, and of Pernici and Scagliosil48 with the injection of bacteria into the general circulation, the possibility of bacteria being carried to and through the mucous mem- brane from the circulating blood may also be considered. † In the experiments of Grawitz,77 Steffen, 77 171 and also of Hopkinsº this action of mucus was not detected in sputum sterilized by heat. † In the course of my experiments, examination of the lower portion of the trachea clearly indicated that in the rabbit, at least, this portion of the respiratory tract is normally free from micro-organisms; but that contaminations, attributed very largely to the death agony, readily occur; and that the bacteria of such contaminations may grow so that, unless the animal be examined immediately after death, considerable numbers of bacteria may be found. Further- more, animals dying of disease, when compared with those killed, frequently show có the presence of contaminating bacteria in this portion of the respiratory apparatus, although the entire tract was apparently normal. WADSWORTH: ETIOLOGY OF ACUTE PNEUMONITIS. 3 In the normal lung bacteria do not, so far as is known, exist as harmless parasites. The researches of Prudden and Northrup,” Tschistovitch,” “ Grammatschikoff,” Silfast,” and others show that pathogenic as well as non-pathogenic bacteria in the air spaces degenerate: Some are taken up by the phagocytes and carried into the lymph channels; others are destroyed by the body fluids. By these processes considerable numbers of bacteria may be disposed of in a comparatively short time. BACTERIAL INCITANTS OF PNEUMONIA. The infectious nature of pneumonia was first definitely established by the researches of Friedländer," Talamon,” and Fraenkel." The various studies of the pneumonic lung at autopsy in long series of cases by many sub- sequent observers, notably, Weichselbaum,” Wolf,” Finkler,” Prudden and Northrup,” Kreibach,” Netter,” Pearce,” Durck,” and Howard,” have served to indicate the etiological importance of the different bacterial incitants. The pneumococcus is now con- sidered as practically the sole incitant of the diffuse lobar lesions. Netter believes the disease to be specific; others concede a small percentage of the lesions to the bacillus of Friedländer. In broncho- pneumonia the more recent researches indicate that the pneumo- coccus rather than the streptococcus is more frequently the incitant, yet in the secondary lesions following certain of the exanthemata the streptococcus is still considered the more important. Although the inflammatory reactions induced by the pneumococ- cus, streptococcus, staphylococcus, and pneumobacillus are some- what similar, they differ in many respects, so that there are certain types of lesion more commonly associated with one species than with others. The pneumococcus lesions are in general more diffuse, Oedematous, fibrinous, hemorrhagic, and less necrotic than those induced by either the streptococcus or pneumobacillus. In pneumo- coccus lesions gangrene rarely develops, except as the result of Secondary infection; on the other hand, the pneumobacillus is not infrequently associated with such a condition. The pneumococcus and streptococcus are, on the whole, more similar in their pathogenic effects than any of the others; but in the lung the streptococcus rarely if ever gives rise to the lobar form of pneumonia. The staphy- lococcus processes are less variable, almost always circumscribed, rarely diffuse or fibrinous, as compared with those of the other three Species. These broad differences, to some extent dependent upon the viru- lence of the bacterial species, may be modified by many conditions, Such as the age or susceptibility of the individual, and more particu- larly by the susceptibility of the race or species of animal infected. In respect to infection with the pneumococcus, Behring and Nissen,” Fraenkel,” Gamaleia.” Kruse and Pansini,” and many others have determined that mice and rabbits are the most suscept- ible; guinea-pigs, rats, squirrels, cats, dogs, sheep, horses, and man 4 WADSWORTH ! ETIOI.OGY OF ACUTE PNEUMONITIS. are in varying degrees less susceptible; and fowls more or less 1Iſlſºl U1 Il 62. PATHS OF INFECTION IN PNEUMONIA. Infection through the air- passages is undoubtedly the rule in bronchopneumonia; but certain Secondary pneumonias arising by hamatogenous infection may so closely simulate bronchopneumonia that the rule must either be excepted for these cases or they must be classified with the metas- tatic pneumonias. Most observers believe that lobar pneumonia also develops from an infection through the air-passages; but regard- ing the universality of this there is some question, doubtless arising from the unsatisfactory results that have attended tracheal inocula- tions in animals and other similar experimental attempts to induce the lobar lesion. Two other paths of infection, the haematogenous and the lymphatic, have therefore been considered by some theoret- ically possible. As supporting haematogenous infection, the results of the blood examinations of pneumonia patients have been cited. Careful scrutiny of these, however,” reveals the fact that while bac- teriaemia develops early as well as late in the disease, there is no indication that this is other than secondary to the local lesion. The occurrence of pneumonia in the foetus offers the clearest example of haematogenous infection. Cases of such infection have been reported by Strachan,” Marchand,” Thorner,” Netter,” Viti,” Czmets- cha,” and Delestre.” These reports, however, are in many respects inadequate; in some histological examinations of the lesions were not made, in others infection by the air-passages after birth could not be excluded. They, therefore, fail to establish the occurrence of lobar pneumonia from placental infection. Experimental support of haematogenous infection in lobar pneumonia is certainly very meagre, uncorroborated, and relatively insignificant. Finally, suf- fice it to note that, clinically as well as experimentally, pneumonic lesions of the lobar type are uncommon in races or individuals very susceptible to systemic pneumococcus infection and more common in resistant individuals or species in which systemic infections are less usual. There are few facts to support a theory of lymphatic infection. Grober’s” interesting work on the tonsil and its lymphatic connec- tions is extremely significant as applied to tuberculous and pleural infections, but the plague bacillus is the only species which is defi- nitely known to reach the lung through the cervical lymphatics and give rise to an acute exudative pneumonitis. These plague pneu- monias, however, are of a peculiar metastatic lymphatic type, quite different from the lobar lesions. In view of these facts and in the absence of reliable positive data, * Witness, for example, the studies of Guarnieri,80 Fraenkel,0i Banti,8 Belfanti,10 Sittmann, 19° Sello, 64 James and Tuttle,” Prochaska, 45 White, 93 Baduel, Pane,” and W andel,189 many of which, and others not cited, have been reviewed recently by Ewing.00 WADSWORTH : ETIOLOGY OF ACUTE PNEUMONITIS. 5 the possibility of lobar pneumonia arising by hamatogenous or lymphatic infection must be considered as not yet established.* PREVIOUS RESEARCHES ON ExPERIMENTAL PNEUMONIA. The earliest attempts to induce pneumonia in animals, which were made when the disease was attributed to exposure or fatigue and be- fore its infectious nature was determined, are now chiefly of historic interest. The first significant experiments comprised the inoculation of animals with the exudates and autopsy material from pneumonia cases. After the infectious nature of the disease had thus been estab- lished and the several bacteria concerned isolated and differentiated, greater scope and precision of experimentation became possible, pure cultures replaced the exudates, etc.; or if the exudates were used the bacteria contained in the exudates were previously deter- mined, and special methods of inoculation were adopted. At first the lungs were inoculated through the chest-wall, then this rather crude method was discarded and the animals were forced by various contrivances to inhale the infectious material. Finally, tracheal injections of the infectious material were used. Although infection of the normal lung of various animal species were not infrequently secured by certain of these methods, typical diffuse pneumonic lesions rarely developed. Attention was therefore directed to the animal and attempts were made through injury to render the lung more susceptible to infection. The results of these experiments as re- gards the development of circumscribed bronchopneumonic types of lesions were more satisfactory and uniform; but as regards the development of diffuse lobar types of lesion were uncertain and far from satisfactory. The experiments of former observers may, therefore, be divided into two main classes: those on normal and those on artificially pre- disposed animals. PREVIOUS ATTEMPTS TO INDUCE PNEUMONIA BY INFECTION OF THE UNINJURED OR NORMAL LUNG. Intravenous, Subcutaneous, intraocular inoculations—in short, all methods giving rise to a pri- mary general infection, whether practised in the ordinary laboratory routine or in the hope of securing pneumonic lesions—have, with a few uncorroborated exceptions,f uniformly failed to give rise to the * This, however, does not exclude the possibility of a metastatic pneumonia, as has been exceptionally noted, simulating the diffuse lobar processes. In fact, between the more typical Processes of all forms of acute exudative pneumonitis intermediate lesions occur which histo- logically are distinguished with difficulty. f Schultzitº reports pneumonic lesions in eight out of fourteen rabbits inoculated intra- Venously with pneumococci. Apparently the collapsed lungs were examined, and certainly the lesions found were inaccurately described as typically exudative. Klipsteinio; reports one instance of a diffuse pneumonia being found after subcutaneous inoculation with a colon bacillus, and Fraenkel's noted the exceptional development of pneumonic lesions after sub- Cutaneous inoculation of the thorax with the pneumococcus. An early observer, Klebs, 100 6 WADSWORTEI: ETIOLOGY OF ACUTE PNEUMONITIS. definite lesions of either bronchopneumonia or lobar pneumonia. Intraperitoneal inoculation has similarly failed, the few exceptional lesions being of the pleuropneumonic type. With intrathoracic inoculation the needle may or may not pene- trate the lung; if the lung tissues are entered a local injury, with more or less hemorrhagic extravasation, results; if not, the pleural cavity only is inoculated and the infection of the lung is secondary, thus pleuropneumonia is incited. By intrathoracic inoculation of pneu- mococci, Talamon,” ineight out of twenty rabbits, Weichselbaum” rarely, Gamaleia,” in dogs and sheep always, obtained pneumonic lesions, some of which were described as diffuse and lobar in type. With the pneumobacillus, Friedländer,”f in thirty-two mice and in one out of four dogs, Weichselbaum,” occasionally in mice and rabbits, obtained pneumonic lesions. In other similar experiments of these and many other observers, notably Kruse and Pansini,” Fraenkel,” Salvioli,” “ Foa and Rattone," Klein,” Sternberg.” v. Besser,” and Silfast,” with pneumococci, streptococci, staphy- lococci, and the pneumobacillus, conducted on mice, rabbits, guinea- pigs, dogs, and horses, the inoculation failed to give rise to pneu- monic lesions, although in the very susceptible animals a general systemic infection often ensued. The inhalation of infectious material has been used chiefly as a test of the permeability of the normal lung to bacteria; but owing very largely to changes occurring in the virulence of the bacteria as a result of cultivation or the drying and pulverizing, and to the uncer- tainty as to how much if any of the infectious material ever reaches the air spaces, inhalation has proved a most uncertain method of infecting the lung. Pneumonic lesions usually circumscribed have, nevertheless, in exceptional instances, been noted, as in the researches of Buchner, Merkel, and Endelen,” “Eppinger,” and Grammat- schikoff,” with anthrax; of Friedländer,” Frnmerich,” and Ches- ter,” with the pneumobacillus; and of Silfast,” with streptococci. Weichselbaum” failed to secure pneumonic lesions in his experi- ments with inhalation. The inhalation of the plague bacillus, as in the experiments of Batzaroff" and others, doubtless owing to the extreme virulence of this organism and the susceptibility of the animals used, has been remarkably successful in giving rise to the different types of plague pneumonia. reports the development of pneumonic lesions after intraocular inoculations with exudates which were not obtained in Veraguth'slS3 similar experiments. In the course of my experi- ments, I have occasionally found at autopsy extensive consolidations of the lung, in which no bacteria or bacteria, other than those inoculated were present, and which could only be at- tributed to chance or an intercurrent infection. A similar undetected infection may account for some of these unusual findings, especially in the early researches. * Exudates were used for inoculation. & † Identity of cultures used is somewhat doubtful. - 1 Emmerich52 53 claims to have invariably secured typical diffuse lesions in animals which had inhaled wirulent pneumococci. His observations have never been corroborated ; in fact, such inhalations have usually proved harmless, wADswortri: ETIOLOGY GF ACUTE PNEUMONITIS. 7 By tracheal injection of the infectious material greater precision was given to the experiments and more definite results were thereby attained; but the virulence of the different bacteria in these experi- ments, as in those with inhalation, was a most important and variable factor. With pneumococci introduced in this way, Fraen- kel"* rarely, Salvioli” in one rabbit and in two of four guinea-pigs, Kruse and Pansini” in possibly one of twenty-eight dogs, Tschisto- vitch” in seven of nineteen cats, and exceptionally in rabbits, obtained diffuse pneumonic lesions. Similar experiments of these and other observers, notably, Gamaleia,” Klein,” Guarnieri,” Griffiini and Cambria.” Talamon,” Lipari,” Welch,” Arus- tamow," Müller,” Aufrecht,” Durck," Coco and Drago,” and Carnot and Fournier,” with pneumococci, conducted on rabbits, guinea- pigs, cats, dogs, and sheep, failed to give rise to typical lesions, the inoculation proving either harmless or, as in very susceptible animals, giving rise to systemic infection, usually without, but in some in- stances, with small circumscribed lung lesions. Patchy or circumscribed lesions have been occasionally induced by the tracheal injection of other bacteria, as in the experiments of Prudden and Northrup,” Orloff,” Fleck," Laehr,” Durck," Sil- fast,” and Beco,” with streptococci and staphylococci; of Lépine and Lyomet” with the typhoid bacillus; of Gamaleia” and of Tschisto- vitch” with the chicken cholera bacillus; and of Tschistovitch” with the bacillus of rouget du porc. Diffuse lesions following endo- tracheal inoculations have been reported by Gamaleia” with the bacilli of anthrax and chicken cholera, and by Bosc and Galavielle” with the micrococcus tetragenus. Finally, the various forms of plague pneumonia have been frequently induced by this method of inoculation. These are often hemorrhagic, usually bronchopneu- monic, sometimes confluent and diffuse. Experiments on the normal animal having given unsatisfactory results, attempts were made to render the lung more susceptible to infection by means of predisposants, which may be considered as acting chiefly upon the lung locally or upon the system at large. Many conditions, such as the inspiration of irritating substances, exposure to cold, trauma, fatigue, and disease, have long been recognized clinically as favoring the development of pneumonia in man. These and other conditions used as experimental substitutes for them have been extensively studied as to their effect per se on the animal and when combined with inoculation. PREVIOUS ATTEMPTS TO INDUCE PNEUMONIA IN THE LOCALLY PREDISPOSED ANIMAL. The early researches of Heidenhain,” Arnold,” and Massalongo,” and the more recent work of Klipstein and Beco,” in which the lungs of animals were exposed by inhala- tion or injection to various noxious substances in gaseous, liquid, or * More often with attenuated than with Virulent cultures. 09 102 - 8 WADSworTH: ETIOLOGY OF ACUTE PNEUMONITIS. solid state, suffice to show that a series of changes largely desoluama- tive, but often exudative, are effected in the tissues of the lung by these irritants.” That the changes thus induced in the lung tissue favor infection and the formation of lesions is well shown by the numerous experiments in which the injured lung has been directly inoculated through the trachea or the predisposed animal has been forced to inhale the infectious material. Such are the experiments of Buchner, Merkel and Enderlen,” and Muskatbluth” with anthrax; and of Prudden and Northrup,” Silfast,” and Beco" with streptococci and staphylococci; but the lesions in these experiments were usually circumscribed and bronchopneumonic in type. Typical diffuse lesions, however, Gamaleia” claims were always obtained when the injured lungs of resistant animals, such as the dog and sheep, were inoculated through the trachea with virulent pneu- mococci; in the normal lungs of these resistant animals the inocula- tion proved harmless, in the more susceptible animals this injury merely favored the development of fatal bacteriaemia without lung lesions. As a result of trauma the lung tissues may be injured and thus pre- disposed to infection. If the integrity of the thorax is destroyed and the lung thereby exposed to external contamination, a wound infec- tion accompanied by various disease processes of the lung, as well as of other tissues, may develop. If, on the other hand, the thorax remains intact, the trauma, acting purely as a predisposant, may give rise to changes in the lung, such as have been noted experi- mentally by Reineboth” “” and others,f which may favor the development of infectious material carried to the lung by the usual channels. In experimental pneumonia trauma is obviously a con- stant and possibly a determining factor. * Exposure to cold, long considered an all-important factor in the development of pneumonia, is now classed with trauma as a pre- disposant, which, only in certain cases, may be a determining factor.' Experimentally the effect of exposure to cold has been extensively studied. The changes brought about in the animal tissues have been attributed to the retention of deleterious products * Extensive lesions closely resembling those of acute exudative pneumonitis have been secured by such treatment without bacterial agency, but this is exceptional and the reports are confined to a few of the earlier observers. # Traumatic pneumonia has also been extensively studied by Litten, 13 Meunier,” Reynaud,” Souques, 108 Gauthier,78 and Schild.161 † Witness, for example, the table compiled by Comby:40 Bouil land attributed to cold . º e tº º . 75 per cent. Grisolle & 4 { { . g ſº $º * . 20 4 & ZiemSSen § { & & & • e tº tº . 9 { { Jurgenson * { § { e * 4. { % Griesinger i { { { & g o i.e te ... 2 { { Massalongo “ g & º sº * * g . () § { 3 Ruhemann, 156 however, considers the exposure to cold the exciting factor, and the presence of the bacteria the predisposing factor in the etiology of pneumonia. WADSWORTH ! ETIOLOGY OF ACUTE PNEUMONITIS. 9 of the body metabolism from diminished excretion (Eisenmann”), to reflex nervous action (Heymann”), and to the direct action of the cold on the tissues (Rosenthal”). These are very largely dependent upon the nature and extent of the exposure, upon the animal species used, and as shown by the researches of Nasaroff,” Durig," and Lode,” upon the nutritive condition of the individual exposed. As a result of exposure the body temperature may be greatly influenced, as in the experiments of Fve,” who found that the experimental extremes compatible with recovery approximate those clinically observed in man, viz., 45° C. (113°F) and 25 C. (77° F.). The marked fluctuations in blood pressure following exposure to either heat or cold, as in the experiments of Hegglin,” Grawitz," Hermann,” Hadl,” and others, may give rise to hyper- aemia, congestion, transudation, or even hemorrhage. Marked anaemia,” with reduction of both cells and haemoglobin and with haemoglobinuria, may develop, as in the studies of Chvostek,” Erhlich,” Grawitz,” Ebstein,” Fischl,” and Rheinboth and Kohl- hardt.” Other cells of the body may undergo parenchymatous degeneration, as was noted by Lassar” and many others since his time. That these changes in the tissues resulting from exposure to cold predispose to many infections has been established by a great variety of researches, notably the early observers, Pasteur,” Wagner,” Ernst,” and Sawtschenko,” and the more recent writers, Chelmonski,” Lode,” Dieudonne,” Silfast,” Sanarelli,” Kiss- kalt,” and Reineboth and Kohlhardt.” Owing to the anatomical structure of the lung these circulatory disturbances and the changes dependent upon them are especially well marked in this organ. Thus, in exposed animals, extensive lesions of the lungs were noted as early as 1862 by Walther,” and in 1870 by Wertheim,” and also recently by Durck.” The devel- opment of pneumonic infection in animals exposed to cold has been studied by means of the tracheal injection or the inhalation of infec- tious material, as in the experiments of Lode” and of Platania,” with the pneumobacillus and staphylococcus; of Silfast,” with streptococci; of Beco,” with staphylococci; of Carrière,” with attenuated pneumococci; and of Lipari,” with virulent pneumo- cocci. In these experiments infection developed, and lesions, usually circumscribed or bronchopneumonic, were obtained more frequently than in normal animals. Lipari” alone reports definite results. Diffuse lobar pneumonia was obtained in all the exposed, but in none of the normal, rabbits of his experiments. Circulatory changes, whether purely reflex and evanescent or associated with definite lesions and damage to the tissues, have been considered determining factors of local predisposition in the lung * The stimulating effect of mild exposure to cold has long been recognized, and the blood as Well as other tissues may be greatly benefited by such exposure, as in the studies of V. Breiten- stein,20 Thayer,177 Winternitz,103 and of Speck.169 >k 10 WADSWORTEI: ETIOILOGY OF ACUTE PNEUMONITIS. as in other organs and tissues. Attempts to demonstrate this experi- mentally have no been satisfactory: thus, Buchner” concluded that arterial hyperaemia was salutary;Kisskalt,” that arterial hyperaemia favored infection, but that stasis hyperaemia, as in tuberculosis, may act differently; Frenkel,” that sensory paralysis increased the susceptibility, while vasomotor paralysis increased the im- munity, but neither had any effect in determining the general infec- tion. Other observers, notably, Hermann,” Kasparek,” Meltzer,” Nekam,” Charrin and Ruffer,” and Roger,” have obtained results which suggested that innervation of a part favored the development of infection. These uncertain results are obviously to be attributed to many variable and indeterminate conditions, relating chiefly to the susceptibility of the animal, on the one hand, and to the nature and virulence of the bacteria on the other.” The normal lung, however, as compared with the other tissues and fluids of the body, is such an unfavorable environment for bac- terial growth that a hyperaemia, with transudation by creating a more favorable soil, may be theoretically considered to favor infec- tion in this organ. For experimental support of this, suffice it to refer to the above-noted predisposing action of irritation and trauma and of exposure to cold, the effects of which in the lung are largely due to circulatory changes. The so-called vagus pneumonias have been studied by many of the early investigators, by Schou,” and recently by Müller” and by Esser.” Section of the vagus nerve, on one or both sides, has been frequently followed by the development of pneumonia, which has usually been of the patchy, bronchopneumonic type, though occa- sionally diffuse. The paralysis of the larynx, which facilitates infec- tion of the lung by bacteria contained in the secretions of the upper respiratory passages and the circulatory changes set up in the lung, are the main factors in these experimental pneumonias. The results of these experiments prove that the method is uncertain, and, as compared with the direct inoculation of the lung with pure cultures of accurately determined bacterial species, such procedures are obviously crude. Clinically, previous disease processes, especially those of the respiratory apparatus, are said to precede the development of pneu- monia, and experience has shown that such inflammatory processes not only impair the protective resources of the parts, but afford favorable conditions for an increase of virulence in the bacteria present in the secretions. In Esser's" experiments the permanent changes left by healed inflammatory processes also favored the development of infection, and Aufrecht" cites statistics to show that the primary pneumonia predisposes to subsequent pneumonic infec- * If the circulating blood is a favorable environment for the growth of the bacteria, a hyper- aemia obviously favors infection ; if unfavorable, retards infection. The experiments in this field have thus as yet added but little to our general theoretical knowledge. WADSW ORTEI: ETIOLOGY OF ACUTE PNEU MONITIS. 11 tions. The predisposition effected by disease processes in other parts of the body, except possibly the heart, is obviously the result of an increased general susceptibility of the individual. PREVIOUS ATTEMPTS TO INDUCE PNEUMONIA IN THE SYSTEMI- CALLY PREDISPOSED ANIMAL. Many conditions influencing the sys- tem at large are well-known predisposants to infection. Some of these are clinically known to favor the development of pneumonia, and certain of the substances or conditions used as local predisposants un- doubtedly act to some extent on the general system. Special methods of Securing a systemic predisposition, however, have not been used in the experimental study of the etiology of pneumonia, except in- directly by Kaminer,” in his work on fibrin formation in the lung. Kaminer obtained pathological changes in the lungs of rabbits poisoned by phenylhydrazin in small repeated subcutaneous doses. In one animal a diffuse fibrinous lesion was found in the lung at autopsy. Bacterial examination failed to reveal micro-organisms. SUMMARY OF THE RESULTS OF PREVIOUS RESEARCHIES ON Ex- PERIMENTAL PNEUMONIA. General systemic infection of the animal by intravenous or by subcutaneous inoculation, whether practised on the normal or on the predisposed animal, has uniformly failed to give rise to pneumonic processes. Local infection of the lung from the inhalation of infectious material or by inoculation with virulent bacteria through the chest-wall, or, preferably, through the trachea, has, in some instances, been followed by the development of pneumonia. Practically all the various forms of primary acute exudative pneumonitis have thus been occasionally induced in pre- disposed and even in normal animals by many different kinds of bacteria. Of the different forms of pneumonia the bronchopneu- monic or patchy has been most commonly incited; the typical diffuse lobar lesions have been secured with great difficulty and exceptionally. It is evident and was early recognized that the nature and viru- lence of the bacteria inoculated were extremely important factors in each experiment; but it does not alone suffice to secure infection of the lung by means of highly virulent cultures, for this may, and does, as a rule, in susceptible animals, lead to a general fatal bac- teriaema, which in nowise contributes to the development of local processes. A precise experimental control of subtle relationships between the virulence of the micro-organism, on the one hand, and the susceptibility of the animal, on the other, should therefore be sought. The virulence of the micro-organisms may be, and has been in many previous experiments, controlled by passing the cul- tures through animals of the species to be experimented upon. The susceptibility of the animal, however, has not been so definitely kept in hand. The various grades or conditions of susceptibility were not accurately determined and cannot be precisely compared, however suggestive in theory they may be. - 12 WADSWORTH . ETIOLOGY OF ACUTE PNEUMONITIS. THE WRITER’s RESEARCHES ON ExPERIMENTAL PNEUMONIA. My studies were undertaken with a view of determining accurately the effect of increasing or diminishing the virulence of the incitant upon the development of pneumonic infection in normal animals and in animals whose susceptibility, local and systemic, was defi- nitely increased. By using one micro-organism, the pneumococcus, and carefully controlling its virulence, and one animal species, the rabbit, whose susceptibility to pneumococcus infection was well known, it was hoped that comparison of the results obtained in a series of systematic researches would lead to more precise methods, and thus to more uniform and certain results. - TECHNIQUE. The media used for growing the pneumococci were made from meat infusion and were brought to a final reaction of less than 1 per cent. acid,” phenolphthalein being the indicator. Although glucose proved a valuable addition to the media, it was rarely used, as these cultures were very short-lived, probably on account of the acid developed. Ascites fluid was, in some instances, added to the broth. In tubes of pleuritic fluid, heated until semi- solid, although still transparent, the pneumococcus stock cultures were kept at a temperature of about 10° in an accessible form, alive and virulent for several (eight or more) months. By this means the changes in the biological state of the micro-organism that usually attend repeated transfers were largely eliminated. It was, therefore, possible to work with the same unmodified races of pneumococci in all the experiments.f Two races of pneumococci were used, one of low, the other of high virulence for rabbits. The first (A) was isolated from a fatal case of lobar pneumonia. The virulence of this organism may be approximately suggested by the fact that intravenously inoculated, 1 c.c. of a twenty-four-hour broth culture proved fatal; injected through the trachea this dose was harmless; subcutaneously inocu- lated, 0.5 c.c. of a twenty-four-hour broth culture induced an abscess at site of inoculation in two rabbits, and fatal septicaemia in two others inoculated from the same culture tube. The second (B) was obtained, through the kindness of Dr. Alexander Lambert, from the laboratory of the New York Health Department, where, by repeated serial inoculation in rabbits, it had been brought to a high virulence, 0.00001 c.c. of a twenty-four-hour serum broth culture at one time proving fatal to the average rabbit. This culture had, however, lost some of its virulence, for I found that although 0.1 c.c. of a twenty-four-hour broth culture intravenously inoculated excited severe reactions in rabbits, these were not always fatal. Owing to * A reaction of 0.5 per cent. acid, which Dr. T. C. Janeway in some unpublished experiments has found to be an optimum, was in most instances used. # For a more complete account of this method and media, see the Proceedings of the New York Pathological Society, April, 1903. WADSWORTH . BTIOLOGY OF ACUTE PNEUMONITIS. 13 the previous adaptation of this race to rabbits, the potential strength of the culture was perhaps greater than was represented by my figures. The inoculations were made with a Koch syringe into the ear vein or through the soft parts between the tracheal rings into the lumen of the trachea, which had previously been exposed by an incision through the skin and fasciae. By careful technique in operating infection of the wound was avoided. The culture fluid was directed to the left lung by elevating and tilting the animal, but forced inspira- tions of the fluid carried it, in some instances, to the right as well as to the left lung. At the autopsies cultures and smears were usually made from the heart blood, trachea, and liver; in addition, the pleurae were often, the spleen and kidney sometimes, examined. The smears were stained by the Welch” method for staining the capsules and also by the newer methods recommended by Hiss.” The lungs were distended with alcohol and all the specimens were hardened in this fluid. Sections of the lung, liver, spleen, and kidney were stained and examined for bacteria by the Gram-Weigert method. Other methods were also used in special instances, and by some of these, as yet not carefully determined, capsules could be demonstrated on the pneumococci in the sections of the lesions. SCOPE OF THE ExPERIMENTS ON NoFMAL AND PREDISPOSED ANIMALS. Intravenous, subcutaneous, and intrathoracic inocula- tions with pneumococci of varying grades of virulence were made in normal and in predisposed animals, with a view of retesting and corroborating the results similar methods had given in other hands. A number of tracheal injections with pneumococci of varying degrees of virulence were made in normal animals with the same object in view. The series of researches on predisposed animals with tracheal injections were thus accurately controlled. The predisposition was Secured in three ways: the systemic susceptibility was increased by a chronic phenylhydrazin poisoning; the lung was locally injured by exposure to severe cold; and finally, by combining the two methods, a predisposition both local and systemic was obtained. In sets* of parallel experiments the effect of these forms of predis- position upon the development of pneumonic infection was studied º the tracheal injection of pneumococci of low and of high viru- €In Ce. ATTEMPTS TO INDUCE PNEUMONIA IN THE UNINJURED OR NORMAL RABBIT. Subcutaneous, intravenous, and intrathoracict inoculation of normal animals with pneumococci of varying grades of virulence failed to give rise to exudative pneumonia. Tracheal injection of * Each set comprised three rabbits, not including the controls. - f Of fifteen normal rabbits, four cats, and one guinea-pig, inoculated through the right chest- Wall, one cat developed a typical but small area of diffuse pneumonitis. The lungs of the other animals were comparatively free from lesions, although all the rabbits and the guinea-pig died, and local processes were found in the pleura and pericardium. 14 WADSWORTEI: ETIOLOGY OF ACUTE PNEUMONITIS. these organisms, however, yielded variable results. Pneumococci of low virulence when injected through the trachea had practically no effect on the lung and were rapidly disposed of. More virulent organisms incited varying, but for the most part Small, central areas of exudative pneumonitis: of seven animals inoculated by the trachea, four died, and in two of these typical and extensive lesions of diffuse pneumonitis were found at autopsy. Pneumococci of still greater virulence injected through the trachea of five normal rabbits incited fatal bacteriaemic infections, with small areas of central broncho- pneumonitis in but two. Thus in my experiments on normal animals, as in those of other observers, the diffuse lesions of exuda- tive pneumonitis were rarely incited, and only when pneumococci of just the requisite grade of virulence were injected through the trachea; higher grades inducing fatal bacteriaemias, lower grades proving harmless. ATTEMPTS TO INDUCE PNEUMONIA IN THE POISONED OR SYSTEM- ICALLY PREDISPOSED RABBIT. In order to determine the effect of increasing the rabbit's systemic susceptibility upon the development of pneumonic infection, studies were made with intravenous inocula- tion and with tracheal injection of pneumococci of high and of low virulence on animals previously poisoned with phenylhydrazin hydrochlorate. As brought out by many researches, notably those of Heinz,” Kaminer," Mya and Sanarelli,” and Lubarsch,” this drug in rabbits, besides many interesting but for the present pur- poses unimportant structural changes in the blood elements, gives rise to severe grades of anaemia of a secondary pernicious type with haemoglobinaemia, haemoglobinuria, and albuminuria. Various stages of parenchymatous degeneration and an oedematous condition of the tissues are also to be found in such animals. - Six rabbits previously poisoned by the repeated subcutaneous administration of phenylhydrazin in small doses were divided into two sets and injected through the trachea with pneumococci of low and of high virulence, respectively. All died in from twelve to fifty- six hours of a general bacteriaemic infection, and the only pneumonic lesion found was a small patch of exudative pneumonitis in the rabbit which lived fifty-six hours. The development of rapidly fatal general infections with little or no lung lesion in these poisoned animals, whether inoculated in the ear vein or through the trachea, and whether cultures of high virulence or of low virulence were used, shows that such rabbits are even less favorable subjects for the experimental incitement of diffuse lesions than the normal animal. The studies were then directed to the effect of increasing the local susceptibility of the lung. ATTEMPTS TO INDUCE PNEUMONIA IN THE INJURED LUNG OR Local LY PREDISPOSED RABBIT. An increase in the susceptibility of the lung without much disturbance of the system at large is obviously best accomplished by means of local irritation. Such WADSWORTEI: ETIOLOGY OF ACUTE PNEUMONITIS. 15 methods have been so exhaustively studied by others and have proved, in the rabbit at least, so consistently unsatisfactory as a means of obtaining diffuse lesions that exposure to cold—another method of rendering the lung more susceptible—was adopted. The animals were tied to trays and placed in ice baths at tem- peratures of 7°C. to 12°C. for five minutes. Immediately after the bath the animals were taken from the trays and wiped with towels. After these exposures various changes were noted. In the lungs large and Small areas of hyperaemia and congestion, occasionally small hemorrhagic extravasations were found. The rectal tempera- ture fell for a period of five, ten, or even twenty minutes after the bath; the lowest.observed in animals which recovered was 26.5°C. A general bacteriaemic infection without lung lesions followed the intravenous inoculation of these animals with pneumococci of vary- ing degrees of virulence, and the exposure of normal rabbits to cold when combined with the tracheal injection of pneumococci also proved an inadequate means of securing definite pneumonic lesions. A study of the effect of combining these methods was therefore made. ATTEMPTS TO INDUCE PNEUMONIA IN THE SYSTEMICALLY AND LOCALLY PREDISPOSED RABBIT. Intravenous inoculation as here- tofore failed to give rise to local processes in the lung and similarly the tracheal injection of the more virulent pneumococci incited but little reaction” in the lungs of these animals, all three of which died of a general infection. With pneumococci of comparatively slight virulence the tracheal injection of these emaciated and exposed rabbits gave rise to extensive and diffuse lung reactions, which were considered typical and comparable to the lobar pneumonias of man in two of the three animals of the experiment. Of these three rabbits: one died in twenty hours and developed slight bronchial reactions similar to those found in the animals injected with the more virulent organisms; another died in thirty hours, and diffuse lesions in an early stage involved the upper por- tion of the lower and the lower portion of the upper left lobes; the third lived four and one-half days, and at autopsy typical red and gray hepatization, macroscopically and microscopically, were found in the upper left lobe. SUMMARY OF THE RESEARCHES ON NORMAL AND PREDISPOSED ANIMALs. It was thus finally possible to approximate at least the balance of conditions as to the virulence of the bacteria on the one hand and on the other the special degrees of local and systemic susceptibility of the animal necessary for the development of typical diffuse lesions of the lung. With an increased susceptibility of the animal pneumococci of low virulence only could be used. More- * One rabbit of the set of three lived sixty hours and showed more marked lesions than the others which died in twenty-four and thirty-six hours. None of the lesions were typical. 16 WADSWORTH ! ETIOLOGY OF ACUTE PNEUMONITIS. over, both the local and systemic susceptibility of the animal in varying degrees must be increased. Such a fine balance of condi- tions, it is obvious, might not always be easily obtained, and the practical solution of the problem had to be sought along other lines. Heretofore, in the experiments of others and in my own described above, the predisposition was secured by methods which increased the susceptibility of the animal to infection. In the lungs of such animals organisms of sufficient virulence to incite extensive lesions gave rise to a rapidly fatal bacteriaemia, and the reaction of the body tissues was thereby generalized or inhibited, so that the local processes failed to develop. The attempts of others to obviate this early development of a bacteriaemic infection by selecting less sus- ceptible animals for experimentation having proved uncertain and as yet impracticable, I sought a method of rendering the rabbit less susceptible to such infection. As this could easily be accom- plished by immunization, my studies were directed to this field of experimentation. ATTEMPTS TO INDUCE PNEUMONIA IN THE IMMUNIZED RABBIT. Preliminary Ea:periments. Two animals were immunized by the inoculation of non-virulent cultures of pneumococci. Their blood serum failed to agglutinate pneumococcus cultures in dilution of 1:1, and they died from the virulent tracheal injection in twelve and eighty hours. Two other rabbits were immunized by the sub- cutaneous inoculation of virulent pneumococci, which induced serious but not fatal local disease processes. Their blood sera agglutinated the pneumococcus cultures in dilution of 1:1, and they survived the tracheal injection. In the rabbit which died twelve hours after the tracheal injection, an early stage of an exuda- tive pneumonitis involving the whole left lung was found at autopsy. Similarly extensive and diffuse lesions of a later stage were present in the animal which lived eighty hours. The other two rabbits recovered from the effects of the tracheal injection and were killed on the third day. Lesions of a resolving diffuse exudative pneumonia involving large areas in one animal and small, more circumscribed areas in the other were found in the lungs. - Additional researches were divided into three sets of experiments, respectively, comprising: five immunized and two normal rabbits, four immunized and three normal rabbits, four immunized and two normal rabbits. Of the thirteen immunized rabbits injected through the trachea, but one died of the infection, and of the seven normal animals three recovered from the inoculation. Diffuse and typical lesions were incited in a considerable proportion of the animals, and the processes were more marked and more frequently found in the immunized than in the normal control animals; but the difference was not sufficiently definite, and it was evident that there were certain WADSWORTH: ETIOI,OGY OF ACUTE PNEUMONITIS. 17 errors in the methods used. Obviously either the virulence of the pneumococci inoculated or the degree of immunization attained by the animals was at fault. At first the immunization of the animals was thought to have been inadequate, but the fallacy in this was apparent on recalling the fact that many animals had died in the process of inmunization, and the survivors used for the experiment, though often greatly emaciated, withstood the virulent tracheal injection. By means of a specially devised technique insuring accurate results” the blood sera of these animals was tested as to its agglutinative action on pneumococcus cells, and the results were compared with those obtained with other blood sera of better-known strength. Thus it was determined that the rabbits had attained an extremely high degree of immunization. The cultures used for inoculation were of pneumococci grown in exudates taken from animals dying of pneumococcus infection and diluted with broth in about equal proportions. Inoculation with such cultures often failed to kill the normal rabbit, whereas cultures of this pneumococcus in broth or in normal sera by trachealinjection or by intravenous inoculation in much smaller doses (Tºm c.c.) invariably proved fatal. A similar loss of virulence in organisms grown in immune sera was noted by Metschnikoff,” and that this loss in virulence is only apparent, not real, was determined by the researches of Metschnikoff's pupils, Isaeft” and Arkharow,' and also by the recent studies of Walker.” The lack of virulence in the tracheal injections of my experiments was therefore attributed to the presence, in the exudates used for culture media, of protective sub- stances, which, in the lungs of the immunized or even of the fresh normal rabbit, became active. In this way an effect was obtained comparable to a lack of virulence not actually existing in the pneu- II].OCOCC1. . It thus became evident that in these researches, as compared with the first preliminary experiments, the conditions essential to the devel- opment of pneumonic processes in the rabbit had unwittingly been altered by the above-noted errors in technique; so much so in fact that the necessary balance for the formation of typical diffuse lesions was present in a majority but not in all of the immunized animals. Final Ea:periments. By avoiding what in the previous experi- ments seemed to be the chief faults, namely, a too high degree of immunization in the animal, and a neutralization of the virulent inoculation, more complete results were expected and obtained in a final experiment on eleven immunizedi and five normal rabbits. * Fully described in a paper published in the Journal of MedicaliResearch, 1903, vol. x., No. 2. f Eight normal rabbits in the process of immunization received : March 6, 1903. 1.5 c.c. of a Salt solution suspension of pneumococcus cells dissolved by normal rabbit bile; subcutaneous. 10th, 5 c.c. sterilized broth culture of pneumococci: intravenous. 12th, 2 c.c. of the bile solu- 18 WADSWORTH ! ETIOLOGY OF ACUTE PNEUMONITIS. March 6, 1903. The eleven immunized and five normal control rabbits were tracheally injected with 1 c.c. of a culture of the ex- tremely virulent pneumococci grown in normal rabbits’ serum di- luted with broth. - Of the five control animals, three died in forty-eight hours without lung lesions; a fourth lived four days, and a small area of exudative pneumonitis was found at autopsy. The fifth was dying on the fifth day, when it was killed, and a similar lesion of the lung was found in this animal. Of the eleven immunized animals none died; but a few were seriously ill for twenty-four to thirty-six hours. The first set of eight were killed on the third day; the second set of three on the fourth day after the tracheal injection. The various lesions of acute dif- fuse pneumonitis involving considerable areas of the left lung were present at autopsy. The smallest lesion found in any of these rabbits comprised at least a quarter of the lower lobe. In one animal, however, the entire left lung, with the exception of the lowermost tip, was solid. Cross-section showed the granular appear- ance seen more markedly in the human lung. SUMMARY OF THE ExPERIMENTS ON IMMUNIZED ANIMALs. It is evident from these experiments that in the lungs of immunized rabbits virulent pneumococci do not give rise to an early general infection, but are confined to the lung, where, if the animal be not too highly immunized, they develop, inducing disease processes comparable to the lobar pneumonia of man. In the experimental control of these conditions under which diffuse lesions may be obtained, there are but two chief variable factors to be considered— the virulence of the micro-organism on the one hand, and the degree of immunization on the other—both easily and accurately determined by the ordinary routine methods of the modern labora- tory. GENERAL CONSIDERATIONS. THE HISTOLOGY OF THE ExPERIMENTAL PNEUMONIAs. The type of inflammation in all the experimental lesions obtained with the pneumococcus was purely exudative and diffuse, usually with abun- dant formation of fibrin. All stages in the development of the reac- tions were present in the different lesions, and it was impossible to draw any sharp lines of distinction between the simple bronchitis and the bronchopneumonic reactions, or between the bronchopneu- monias and the diffuse lobar lesions. The one merged gradually into that of the other, although the types were easily distinguished. The leukocytes were usually numerous and particularly prominent tion of pneumococcus cells : subcutaneous. 26th. 1.5 c.c. of clear salt solution extract of pneu- mocOccus Cells : intravenous. Three normal rabbits in the process of immunization received : March 2, 1903. 3 c c. of a salt solution suspension of sediment from pneumococcus growth in glucose broth ; intravenous, 3d, 3 C.C. Subcutaneous. - WADSWORTH ! ETIOLOGY OF ACUTE PNEUMONITIS. 19 in the early stages of the lesion. The exfoliation of epithelium also occurred early in the process, but the desquamated cells were rela- tively more numerous in the later stages. Few fibrin fibrils could be found in the very early stages, but granules were often abundant. As has been noted by Kohn,” Hauser,” Bezzola,” and others, the fibrin developed and was more abundant in the early stages along the walls of the air spaces, and in places the fibrils could be traced from one alveolus to another. In the more advanced lesions the fibrin formation was most marked in the centre of the exudate, a clearer zone lying next the alveolar wall.” The pneumococci were present, often in large numbers, in all the early stages of the lesions. The cocci were for the most part outside the cells. In resolving lesions the bacteria were often absent, even though large portions of the lung were involved. This was par- ticularly noticeable in the lesions which were found in the immunized animals. Areas of irregular development gave to the lobar lesions during the early stages a patchy appearance. These areas may or may not correspond with the lobule of the lung; certainly they were not as distinct as the lobular patches present in human lobar pneu- monia, nor could the intralobular structure described by Bezzola” and Ribbert” be definitely made out. The lung of the rabbit doubtless offers certain differences in the anatomical structure or arrangement of its lobules which may possibly explain these slight discrepancies. Furthermore, the posture of the animal doubtless influences the absorption and accumulation of exudates in the lung and thus, to some extent, the character of the lesion. THE DEVELOPMENT OF ExPERIMENTAL PNEUMONIA. It is obvi- ous that the development of localized disease processes in the lung, as elsewhere, is determined by conditions referable either to the infecting micro-organism or to the animal infected. The infecting micro-organisms are extremely susceptible to environmental influ- ences and thus exhibit considerable biological variation of both species and race. The biological character of the bacterial species as regards the peculiarities of growth and the elaboration of toxins and the special adaptation of the individual or race to the animal infected, commonly though inadequately expressed by the term virulence, are thus essential determining, but extremely variable factors in the development of all infections. On the other hand, the animals infected, within certain variable limits, afford favorable or unfavorable environments for the incitants of infection; in short, the animals are in varying degrees susceptible of infection.f This * In distended or inflated lungs an artefact is often produced which should not be confused With these appearances. f An important and co-operating factor intimately associated with the susceptibility of the animal is the chemotaxis. In animals very susceptible to certain infections it has been shown experimentally that, with extremely virulent cultures of the incitants, the chemotaxis is nega- tive. In less susceptible animals or in immunized animals or with less virulent incitants the chemotaxis is in varying degrees positive, 20 WADSWORTEI: ETIOI,OGY OF ACUTE PNEUMONITIS. Susceptibility may be local as affecting certain tissues or parts of the animal or it may be systemic. When the bacterial incitant is extremely virulent or the animal very susceptible a systemic in- fection is induced, but when the bacterial incitant is less virulent or the animal only partially susceptible, a local infection is in- duced. The morphological character of the disease processes depends in part upon the grades of susceptibility in the animal tissues, but chiefly upon the nature and virulence of the bacterial incitants, which, as noted above, are extremely variable. Different races of the same bacterial species thus often give rise to very different types of lesion in different or even in the same species of animal. Simi- larly, according to the degree of specialization necessary in the bac- terial incitant, the different types of lesion may each be induced by a large or a small number of bacterial species or even by only one species, as in the so-called specific lesions. For example, in the lung, many different kinds of bacteria give rise to broncho- pneumonia, but few to diffuse lobar pneumonia, and only one, the tubercle bacillus, to a specific productive necrotic process called tuberculosis. Owing to the special anatomical arrangement of the epithelial, vascular, and lymphatic tissues of the lung, and to the physiological or pathological adaptation of these tissues, the lung surface acts, to some extent, similarly to the skin and mucous membrane as a barrier to infection, yet is peculiarly fitted not only for the rapid absorption and elimination of toxic material, but for the rapid diffusion of such material and the accumulation of exudates. The problem of infec- tion in the lung is thus obviously complicated, and the relationship between the above-cited essential factors of infection more subtle than in other parts of the body. For most bacteria the normal lung is unfavorable, and similarly for the bacteria commonly found in pneumonic lesions the lung, as compared with other tissues, is, in most animals, a less favorable environment; thus for infection of the lung special adaptation or virulence of the bacterial excitant is essential. If this virulence be too great, the micro-organisms gain early access to and develop more rapidly in the other more favorable tissues of the susceptible animals. If this virulence be moderate, localized processes are incited in the lung, the surface of which, acting as a barrier to infection, prevents the invasion of the more sus- ceptible tissues until the animal becomes, in a measure, immunized or the organisms acquire the requisite virulence for systemic infec- tion. For the development of extensive diffuse lesions of the lung a relatively extreme adaptation or virulence of the incitant is obvi- ously essential; thus these lesions rarely develop in animals very susceptible to the infection. Furthermore, bacteriaemias, whether temporary or terminal, are apt to complicate such disease processes. A relative systemic insusceptibility, whether natural to the animal WADSWORTH ! ETIOLOGY OF ACUTE PNEUMONITIS. 21 species or acquired by the individual,” is, therefore, an extremely important factor in the development of lobar pneumonia. For the formation of circumscribed or patchy lesions of the lungs less special- ization or virulence of the incitant is required; thus the protective mechanism of the lung is usually an efficient barrier shielding the more susceptible tissues from infection, and bacteriaemias are com- paratively rare. Occasionally, however, as in susceptible animals infected with incitants which possess or quickly attain a high viru- lence, a systemic infection supersedes the localized pneumonic lesion. Plague pneumonia is an apt illustration of the development of bronchopneumonia under these conditions, which, except for the lack of systemic insusceptibility, are the same as give rise to lobar pneumonia. Thus it is that in bronchopneumonia and in lobar pneumonia similar etiological conditions in slightly varying relationship give rise to slightly varying lesions differing only in type. THE PNEUMONIC PROCESS IN MAN IN THE LIGHT OF THESE EXPERIMENTAL RESEARCHES ON ANIMALs. All grades of reac- tion from the idiopathic lung congestions (Maladie de Woillez) to the diffuse lobar pneumonic lesions may be induced by pneumo- cocci in the lung of man. In these the different races of pneu- mococci exhibit wide variations in virulence. Carrière” found attenuated pneumococci in the lung congestions, whereas it is well known that those isolated from lobar lesions are usually extremely virulent. Eyre and Washbourn" tested four stocks of pneumococci: one from the sputum of a healthy person, three from pneumonia cases. The pneumonia cultures became excessively virulent by passing them through from eight to eleven rabbits, 0.00001 loop proving fatal in rabbits. The sputum culture was passed through forty-three rabbits before 0.001 loop was fatal, and fifty-three before 0.00001 loop proved fatal. The subsequent observations of these authors confirm the above-noted results. It is well known that, relative to other species, man is more resist- anti to pneumococcus infection. However, the resistance offered by different races of men varies greatly. Billings,” from the census report, found that the negro is markedly susceptible to pneumonic infection. Marchoux” reports an epidemic of pneumococcus infec- tion of a bacteriaemic type occurring in a body of French colonial negro troops. Although disseminated lesions were present, well- * This systemic insusceptibility may be acquired by individuals normally susceptible, in the early stages of the infection which may for a time be confined to the air spaces where the exu- dates collect; thus lobar pneumonia is occasionally induced in very susceptible animals. Such a fine balance of conditions is for the present at least beyond experimental control. # The presence of the pneumococcus in the sputum of health and in many of the disease processes of the nose, throat, and ear suggests that more or less specific resistance may be developed at times in some if not in many individuals. By means of a special technique I have obtained agglutinin and precipitin reactions with the blood serum from certain animal Species not usually susceptible to pneumococcus infection, and with the normal human sera tests these reactions were more marked and developed in higher dilutions. (Journal of Medical Research, 1903, No. 2, vol. x.) 22 WADSWORTH: ETIOLOGY OF ACUTE PNEUMONITIS. developed pneumonias were comparatively rare. Kolle” noted a similar epidemic among the negroes of South Africa, but lesions were found in the lungs of some of these cases. Brodie, Rogers, and Hamilton” studied two epidemics of severe general pneumococcus infection, characterized by rhinitis and gastrointestinal disturbances, occurring in the Kaffirs. Pneumoniclesions were rare. Tostivant and Remlinger” describe the marked susceptibility of certain Arab tribes to pneumococcus infection and the difference in the lesions induced. Aside from racial susceptibility, the individual at different ages and under different conditions offers striking analogies, in some instances to the susceptible, in others to the resistant, animal. In general, the newborn and the infant are remarkably susceptible to infection. At this age lobar pneumonias are rare; infection of the lungs with pathogenic bacteria gives rise to the bronchopneumonic type of lesion. Here, then, is a direct analogy to the susceptible animal. In the adult, as in the resistant or immunized animal, lobar lesions are more commonly found in the primary pneumonias; but, on the other hand, the adult may be rendered in varying degrees susceptible by other disease processes, and exudative pneumonitis developing under these conditions, the so-called secondary pneu- monia, is more apt to be of the bronchopneumonic type. It has long been recognized clinically that in the lobar pneumonias of man there is no definite relation between the extent of the lesion and the gravity of the infection upon which a prognosis may be safely based. This is even more pronounced in the experimental pneumonias in animals, and, as shown by the researches on animals, is due to the fact that extensive lesions may be incited by virulent infections when confined to the lung without seriously endangering the life of the individual; whereas less extensive lesions are often followed by general infections which terminate fatally. Thus the lesions in the fatal lobar pneumonias of old people are often small, sometimes unrecognized; whereas the lungs of a robust individual with good protective resources may be extensively involved, and yet recovery take place. As suggested by the researches on animals, the generalization of the infection is the real danger. Upon the determination of this only can a prognosis be safely based. It is therefore evident that the various phenomena of pneumo- coccus infection observed in animal inoculation also occur in man. The rule of one is, however, the exception of the other; thus, general infections, except as transitory or terminal conditions complicating the local processes are comparatively rare in man, but common in the ordinary laboratory animal. SUMMARY AND CONCLUSIONS. As a result of the numerous and varied researches of other ob- servers it has been determined that the bronchi and lung under WADSworTH: ETIOLOGY OF ACUTE PNEUMONITIS. 23 normal conditions are practically free from micro-organisms; that the secretions of the upper respiratory tract in both health and disease often harbor the bacteria commonly found in pneumonia, namely, the pneumococcus, the streptococcus, the staphylococcus, and the pneumobacillus; that the incitants of pneumonia may be carried to the lung by the lymph channels, inducing an interstitial pleuropneumonitis; or by the bloodvessels, giving rise to secondary metastatic processes; or, as is usual, by the air-passages, inciting the various lesions of exudative pneumonitis, of which two main types, the bronchopneumonic and the diffuse lobar, are recognized; and, finally, that the bronchopneumonic lesions, whether arising in man or induced experimentally in animals, may be incited by a great variety of bacteria, most frequently by the pneumococcus or the streptococcus; whereas, comparatively few species of bacteria give rise to diffuse lesions in animals, and in man practically but one species, the pneumococcus, ever attains the necessary specialization for the incitement of lobar pneumonia. In the researches of others patchy or circumscribed lesions com- parable to the bronchopneumonia of man, have been induced experimentally with comparatively little difficulty. Previous at- tempts to induce diffuse lobar lesions in animals, though successful in a few exceptional instances, have failed to determine the exact conditions under which the lesion develops and offer no reliable method of securing typical pneumonic processes. By means of accurately controlled series of experiments which allowed of precise comparison, it was possible in my researches to determine the effect of increasing or diminishing the virulence of the incitant, the pneumococcus, in systemically predisposed, in locally predisposed, and in both systemically and locally predisposed animals. These experiments show that the incitement of diffuse lesions in the normal rabbit is extremely uncertain, and only possible in the predisposed rabbit when both the general and local suscepti- bility are increased and when organisms of comparatively low viru- lence are used. Thus, and owing chiefly to the fact that the lung surface acts as a barrier to infection, the development of acute exudative pneumonitis offers an especially clear example of the nice balancing of the essential conditions determining infection. These conditions are, on the one hand, the specialization or virulence of the incitant, and, on the other hand, the animal susceptibility, both local and systemic. Organisms of low virulence induce evanescent bronchial reactions; more virulent organisms by a local infection, give rise to the more typical bronchopneumonic lesions; while organisms of still greater virulence, if confined to the lung, incite diffuse processes of the lobar type, but if not so confined and bac- teriaemic infection occurs, the lung lesions are less marked and of the bronchopneumonic type. The extremely fine balance of these conditions essential to the formation of lobar lesions in normal as A 24 WADSWORTEI: ETIOLOGY OF ACUTE PNEUMONITIS. well as in predisposed animals is as yet for practical purposes be- yond experimental control. Finally, as a result of these systematic researches on normal and predisposed animals, an entirely new procedure was adopted: the preliminary immunization of the rabbits, so that extremely virulent cultures of pneumococci can be used without giving rise to bac- teriaemic infection. The experiments with this procedure showed that diffuse exudative lesions comparable to the lobar pneumonia of man may be incited experimentally in the immunized rabbit; and that in securing this result there are but two chief variable factors, the virulence of the incitant and the immunization of the animal; both easily and accurately controlled by the routine tech- nique of the modern laboratory. I am deeply indebted to Professor T. Mitchell Prudden and to Professor Philip Hanson Hiss, Jr., for their suggestions and criticism. BIBLIOGRAPHY. . Arkharow. Archiv. de méd. exp., 1892, p. 498. - . Arloing. Journ. de physiol. et de path. gén., 1902, p. 291. Bibliog. , Arnold. Untersuchungen über Staubinhalation u. Staubmetastase, Leipzig, 1885. , Arustamow. Woenno-Medicinskij, 1889, Hft. 2–4. Baumgarten Jahrbrt., 1889, p. 61. . Aufrecht. Nothnagel specielle Path. u. Therapie, 1899, part ii. vol. xiv. pp. 1–63. Bib. . Aufrecht. Ibid., pp. 55, 56. - . Baduel. Riforma Medica, 1899, No. 15, p. 170. . Banti. Lo Sperimentale, 1890, vols. lxv.-lxvi. pp. 349, 461, 573. 9. Barthel. Centralblatt f. Bakt., 1898, vol. xxiv. pp. 401, 433, 574–577. 10. Baumgarten. Mycologie, 1889, vol. ii. p. 602. 11. Batzaroff. Annales de l’Institut Pasteur, 1899, p. 385. 12. Beco. Archiv. de médecine expér., 1899, p. 317. 13. Beco. Ibid., 1901, p. 51. Bibliog. 14. Behring and Nissen. Zeitschrift f. Hygiene, 1890, vol. viii. p. 412 15. Bein. Charité Annalen, 1895, vol. xx. Bibliog. 16. Belfanti. La Riforma Med., 1890, vol. i. p. 338. 17. v. Besser. Ziegler's Beiträge, 1889, vol. vi. p. 333. 18. v. Besser. Ibid., p. 339. 19. W. Besser. Ibid., pp. 351–362. 20. Bezancon and Griffon. Soc. méd. des hôpitaux, April 15, 1898. 21. Bezzola, Virchow's Archiv, 1894, vol. cxxxvi. p. 345. 22. Billings. Census Report, 1890. 23. Boni. Deutsches Arch. f. klin. Med., 1901, vol. lxix. p. 542. Bibliog. 24. Bordonni-Uffredozzi. Centralblatt f. Bakt., 1891, p. 305. 25. Bosc aud Galavielle. Soc. de biol., 1898, p. 981. . 26. v. Breitenstein. Archiv f. exper. Path, u. Pharm., 1896, vol. xxxvii. p. 253. Bibliog. 27. Brodie, Rogers, and Hamilton. Lancet, 1898, p. 1045. 28. Buchner. Münchener med. Wochenschrift, 1899, pp. 1261 and 1301. 29. Buchner, Merkel, and Enderlen. Archiv f. Hygiene, 1888, vol. viii. 30. Buttersack. Zeitschr. f. klin. Med... vol. xix. p. 411. Bibliog. 31. Calamida, and Bertarelli. Centralblatt f. Bakt., 1902, Vol. xxxii. p., 428. 32. Carrière. Revue de méd., 1898, pp. 765 and 951. Ibid., 1899, p. 54. Presse méd., Jan. 27, 33, Carnot and Fournier. Archiv. de méd. exper., 1900, vol. xii. pp. 367–369. 34. Charrin and Ruffer. Compt.-rend. de la Soc. de la biol., 1889, p. 208. 35. Chelmonski. Deutsches Archiv f. klin. Med., 1897, vol. Iix. p. 140. 36. Chester. Report of Delaware Agricultural Experimental Station, 1899–1900. 37. Chwostek. Ueber das Wesen der Hemoglobinurie, Leipzig u. Wien, 1894. Bibliog. 38. Claisse. Les infectiones bronchique, Paris, 1893. La semaine méd., 1893, No. 38, pp. 297-300. - WADSworTH: ETIOLOGY OF ACUTE PNEUMONITIS. 25 39. Coco and Drago. Gazzetta degli ospitali, 1899, p. 104. 40. Comby. Traité de mal. de l'enfance, 1898, vol. iv. p. 56. 41. Czmetscha. Prager med. Wochenschr., 1894, vol. xix. p. 233. 42. Delestre. Compt.-rend, de la Soc. de biol., 1898, p. 150. 43. Dieudonne Arbeit, a. d. kaiserl. Gesundheitsamte, 1894, vol. ix. p. 492. 44. Durck. Deutsches Archiv f. klin. Med., 1897, vol. lviii. pp. 368–415. 45. Durck. Ibid., pp. 418–429 46. Durck. Ibid., pp. 435–444, Bibliog. 47. Durig and Lode. Archiv f. Hygiene, 1901, vol. xxxix. p. 46. Bibliog. 48. Ehrlich. Zeitschrift f. klin. Med., 1881, Vol. iii. p. 383. 49. Ebstein. Arch. f. exper. Path., vol. ii. 50. Eisenmann. Archiv f. Wissensch. Heilk., 1862, vol. vi. p. 325. 51. Emmerich. Fortschritte der Med., vol. ii. p. 153. 52. Emmerich. Seventh International Congress of Hygiene, London, 1891. British Medical Journal, 1891, pp. 379, 380. - 53, Emmerich. Münchener med. Wochenschr., 1891, p. 554. 54. Enderlen. Deutsch. Zeitsch. f. klin. Med., vol. xv. p. 50. 55. Eppinger. Die Hadenkrankheit., Lubarsch and Ostertag, 1896. 56. Ernst. Ziegler's Beiträge, 1890, vol. viii. p. 203. 57. Esser. Centralblatt f. d. gesammte Med., January 26, 1901. 58. Esser. Arch. f. exper. Path. u. Pharm., 1903, vol. xlix. p. 200. Bibliog. 59. Eve. Journal of Physiology, 1900, Vol. xxvi. p. 119. 60. Ewing. Clinical Pathology of the Blood, 1903, p. 289. 61. Eyre and Washbourn. Lancet, 1899, vol. i., p. 19. 62. Finkler. Die Acuten Lungenentzundungen, 1891. 63. Finkler. Thesis, 1895, Bonn, Baumgarten Jahrbt., 1895. 64. Fischl. Zeitschrift f. Heilk., vol. xviii. p. 321. Prager med. Wochenschr., 1897, p. 49. 65. Fleck. Diss., Bonn, 1886. 66. Foa and Rattone. Gazzetta degli ospitali, 1885, No. 12, p. 94. 67. Fraenkel. Zeitschrift f. klin. Med., vol. x. p. 401; vol. xi. p. 437. Bibliog. 68. Fraenkel. Ibid., pp. 417, 418. 69 Fraenkel. Ibid., p. 423. 70. Frenkel. Achiv. de méd. exper., 1892, p. 638. 71, Friedländer. Virchow's Archiv, 1882, vol. lxxxvii. p. 319. Fortschr, der Med., 1883, vol. vii. p. 715. - 72. Gamaleia. Annales de l’Institut Pasteur, 1888, p. 440. 73. Gauthier. Lyon méd., 1900, vol. xcv. pp. 329, 372. 74. Gobell. Diss., 1897, Marburg. - 75. Grammatschikoff. Arb. a. d. Geb, de path. anat., Inst. zu Tübingen, 1891, p. 450. 76. Grawatz. Zeitschrift f. klin. Med., 1892, Vol. xxi. p. 459. Centralbl. f. inn. Med., 1894, Vol. xv. p. 33. - 77. Grawitz and Steffen. Berl. klin. Wochenschr., 1894. 78. Griffini and Cambria. Giornales internazionale de sc, méd., 1882, vol. iv. p. 353. Archiv. ital. de biol., 1883, vol. iii. p. 189. 79. Grober. Deutsches Archiv f. klin. Med., Vol. lxviii. p. 296. 80. Guarnieri. Attid. e. accad. med, di Roma, 1888, Ser. 2, Vol. iv, p. 97. 81. Hadl. Inaug. Diss., Würzburg, 1896. 82. Hauser. Münch. med, Wochenschr., 1893, p. 155. 83. Hegglin. Zeitschr. f. klin. Med., 1894, vol. xxvi. 84. Heidenhain. Virchow's Archiv, 1877, Vol. lxx. p. 441. 85. Heinz, Ziegler's Beiträge, 1901, vol. xxix. p. 299. 86. Hermann. Annales de l’Institut Pasteur, 1891, p. 243. 87. Hermann, L. Pflüger's Archiv, 1870, vol. iii. p. 8. 88. Heymann. Berliner klin. Wochenschr., 1872, p. 447. 89. Hildebrandt. Ziegler's Beiträge, 1888, vol. ii. p. 411. 90. Hiss, Centralblatt f. Bact , vol. xxxi. p., 302. 91. Hoffmann. Nothnagel’s Specielle Pathologie, 1896, vol. xiii. (Th. iii. abth. i.). 92. Hopkins. International Dental Journal, August, 1903. 93. Howard. Cleveland Journal of Medicine, 1900, vol. v. p. 450. 94. Isaeff. Annales de l’Institut Pasteur, vol. vii. p. 260. 95. James and Tuttle. Presbyterian Hospital Reports (New York), vol. iii. 96. Kaminer. Zeitschr. f. klin. Med., 1900, vol. xli. p. 91. 97. Kasparek. Wiener klin. Wochenschr., 1895, pp. 570, 593. 98. Kisskalt. Münchener med, Wochenschr., 1900, p. 110. 26 WADSworTH: ETIOLOGY OF ACUTE PNEUMONITIS. 99. 100. 101. 102. 103, 104. 105. 106. 107. 108. 109. 110, 111. 1899. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. Paris, 122. 123. 124. 125. p. 80. 126. 127. 128. 129, 130. 131. 132. Kisskalt. Archiv f. Hygiene, 1900, vol. xxxix. p. 142, Bibliog. Klebs. Archiv f. exper. Path., vol. iv. p. 420. Klein. Centralblatt f. d. med. Wiss., 1884, No. 30. Klipstein. Zeitschr. f. klin. Med., 1898, vol. xxxiv. p. 191. Kohn. Münch, med. Wochenschr., 1893, p. 42. Kolle. Deutsches med. Wochcmschr., 1898, p. 425. Kreibach. Breiträge zur klin. Med. u. Chir., 1896, Heft 13, p. 1. Kruse and Pansini. Zeitschr. f. Hygiene, 1892, vol. xi. pp. 337–355. Kuttner. Centralblatt f. die med, Wissensch., 1875. Laehr. Diss., Bonn, 1887. Laschtschenko. Zeitschrift f. Hygiene, vol. xxx. p. 125. Lassar, Virchow's Archiv, Vol. lxxix. p. 168, - Lépine and Lyomet. Soc. de biol., 1899, p. 583. Compt.-rend. de la Acad. Sc., Feb. 13, Lipari. Il Morgagni, 1888, pp. 523, 575, 651. Litten. Zeitschr. f. klin. Med., 1882, vol. V. p. 26. Lode, Alois. Archiv f. Hygiene, 1897, vol. xxviii. p. 344. Bibliog. Lode, A., and Durig, A. Münchener med. Wochenschr., 1900, p. 109. Lubarsch. Geschwulsten u. Infectionskrankheiten, 1899, p. 144. Marchand. Arch. f. path. Anat., 1887, vol. 109, p. 86. Marchoux, Annales de l'Institut Pasteur, 1899, p. 198. Massalongo. Archiv. de Physiol., 1885, 3d ser., T. vi. Meltzer and C. Meltzer. Journal of Medical Research, 1903, No. 1, vol. x. Meunier. Du rôle du systeme nerveux dans l’infection de l'appareilbroncho-pulmonaire, 1896. . Metschnikoff. Annales de l’Institut Pasteur, 1887, Vol. i. p. 42. Ibid., 1892, vol. vi. p. 297. Miller. Dental Cosmos, 1891, Vol. xxxiii. Morisani. Arch. di Ost. e Ginec., 1897, vol. iv, p. 133. Müller, W. Deutsches Archiv f. klin. Med., 1901, vol. lxxi, p. 513; and 1902, vol. lxxii. Bibliog. Müller, F. Münch. med. Wocenschr., 1897, No. 49, p. 1382. Bibliog. Muskatbluth. Centralblatt f. Bakt., 1887, vol. i. p. 321. Mya, and Sanatelli. Vide Lubarsch. Nassaroff. Virchow's Archiv, Vol. x.c. p. 482. Nekam. Centralblatt f. Bakt., 1894, vol. xvi. p. 932. Nenninger. Zeitschr. f. Hygiene, 1901, vol. xxxviii. - Netter. Soc. de biol., November 29, 1887; Traité de médecine (Bouchard-Brissaud), 1901, Vol. Wi. pp. 485, 486. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147, 148, 149. 150. 151. 152. 153. 154, 155. 156. 157. 158. Netter. Ibid., pp. 475, 476. Netter. Bulletin de la Soc. anat., April 9, 1886. Netter. Traité de médecine (Bouchard-Brissaud), 1901, vol. vi. p. 506. Neumann, R. Zeitschrift f. Hygiene, 1902, vol. xl. p. 33. Orloff. Wratsch, 1887, pp. 385 and 401, * Ottolenghi. Arch. f. le Sc. Med., vol. xxii. p. 425. Pane. Riforma medica, 1899, vol. iii. p. 376. Pasteur. Bulletin de l'Acad. méd., 1878, vol. vii. p. 253. Pearce. Boston Medical and Surgical Journal, No. 23, Vol. cxxxvii. Bibliog. Platania. Giornale internaz. delle sc. med., 1889, p. 344. Bibliog. Pernici and Scagliosi. Riforma medica, 1892, pp. 97, 98. Prudden and Northrup. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCEs, 1889, Prochaska, Archiv f. klin. Med. Vol. lxx. p. 559. - Reineboth. Münch, med. Wochenschr., 1898, p. 1170. Reineboth. Deutsches Archiv f. klin. Med., 1899, vol. lxii, p. 63. Reineboth. Ibid., 1900, vol. lxix. p. 144. Reineboth and Kohlhardt, Ibid., vol. lxv. p. 192. Reynaud Bulletin médicale, July 1, 1899. Bibliog. - Ribbert. Fortschritte der Med., 1894, p. 371; Virchow's Archiv, 1894, Vol. cxxxvi. p. 359. Ritchie. Journal of Pathology and Bacteriology, 1901, p. 13. Bibliog. . Roger. Compt. rend, de la Soc. de biol., 1896. Rosenthal. Inaug. Diss., Erlangen, 1893. Rosenthal, J. Berliner klin. Wochenschr, 1872, p. 453. Ruhemann. These, Leipzig, 1898. Zeitschrift f. Krankenpflege, 1898. Salvioli. Arch. per le sc. méd., 1884, No. 7, vol. 8. Salvioli and Zaslein. Centralblatt d. med. Wiss., 1883, p. 721. WADSWORTH ! ETIOLOGY OF ACUTE PNEUMONITIS. 27 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. Sanarelli. Centralblatt f. Bact., 1891, vol. x. p. 817. Sawtschenko. Ibid., vol. ix. p. 473. Schild. München. med. Wochenschr., 1902, Vol. xlix. p. 1569. Schou. Fortschritte der Med., 1885, No. 15. Schultz. Arch. de Sc. de biologie, St. Petersburg, 1901, Vol. viii. p. 1. Sello. Zeitschrift f. klin. Med., vol. xxxvi. p. 112. Silfast. Ziegler's Beiträge, 1899, vol. xxv. Sittmann. Deutsches Arch. f. klin. Med., 1894, Vol. liii. p. 323. Slawjanski. Virchow's Archiv, 1869, vol. xlviii. Souques. Presse médicale, 1900, p. 109. Bibliog. Speck. Zeitschr, f. klin. Med., 1901, wol. xliii. p. 377. Spolverini. Annali d'igiene sperim , 1899, p. 202. Steffen. Centralblatt f. Bakt., 1895, vol. xviii. p. 464. Sternberg. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCEs, 1885, vol xc. p. 117. Sternberg, Text-Book of Bacteriology, 1893, p. 298. Strachan. British Medical Journal, 1886. Stroganoff. Centralblatt f. Gynäk., 1893, No. 40. Talamon. Progrès méd., 1883, pp. 281, 301, and 1030. gº Thayer. Johns Hopkins Hospital Bulletin, 1893, p. 37. Thomson and Hewlett. Medical and Surgical Transactions, 1895, p. 239. British Medi- cal Journal, 1896, wol. i. p. 137. 179. 180. 181. 182. . Veraguth. Virchow's Archiv, 1880, vol. lxxxiii. p. 238. Bibliog. . Viti. Arch. ital di pediatris, 1890 . Wagner. Annales de l’Institut Pasteur, 1890, vol. iv. p. 570. . Walker. Journal Pathology and Bacteriology, 1902, vol, viii. p. 34. Thorner. These, Munich, 1888. Cited by Netter. Tschistovitch. Annales de l’Institut Pasteur, 1889, p. 337. Tschistovitch. Ibid., 1890, vol. iv. p. 285. Tostivant and Remlinger. Archiv. de méd. et pharm, mil., 1901, vol. xxxviii. p. 107. 187. Walthard. Centralblatt f. Bakt., 1895, vol. xvii. p. 311; Arch. f. Gynāk, vol. xlviii. pp. 233–244. 188. Walther. Virchow's Archiv, 1862, Vol. xxv. 189. Wandel. Deutsches Archiv f. klin. Med., 1903, vol. lxxviii. p. 1. 190. Weichselbaum. Wiener med. Wochenschr., 1886, Nos. 39, 40, and 41. 191. Welch. Johns Hopkins Hospital Bulletin, 1892, p. 125. 192. Wertheim. Wiener med, Wochenschr., 1870. 193. 194. 195. 196 White. Journal Experimental Medicine, vol. iv. p. 425. Widal and Bezancon Revue trimens. suisse d'odontolog., 1894, p. 185. Winternitz. Centralblatt f. klin. Med., 1893, vol. xiv. p. 1017. 197. 198. Wolf. Wiener med. Blatter, 1887, Nos. 10, 11, 12, 13, and 14. Wright. New York Medical Journal, July 27, 1889. Wurtz and Lemoyez. Compt.-rend. de la Soc. de biol., 1894. A SIMPLE METHOD OF OBTAINING BLOOD FROM THE RAB,BIT. AUGUSTUS WADSWORTH, M. D. Dr. Wadsworth demonstrated a simple method which he used for obtaining blood from the rabbit. He said that although he had usually had little difficulty in bleed- ing rabbits, there were animals, especially those anaemic or emaciated, which bled with great difficulty from the ear veins, yielding only a few drops at the most. His experience in bleeding rabbits had brought out the fact that there was less difficulty in summer than in winter, and on warm than on cold days, and when the animal was tilted head downwards. He had therefore adopted the method of warming and tilting his animals, which brought the blood into the peripheral circulation and distended the ear veins. Anaemic and emaciated animals which had proved refractory yielded under these condi- tions sufficient blood (five c.c. or more) for most pur- poses. After having tried various methods of warming his animals, Dr. Wadsworth had finally adopted the plan of putting them on a hot water bed. An ordinary com- mercial rubber bag, sixteen inches square, was partially filled with hot water and placed cornerwise on an animal tray. The animal, placed on the bag, was tied by its legs to the angles of the tray, and a blanket, wrapped around the bag and animal, increased the surface in con- tact with the water bag and kept in the heat. The rabbit was then ready to be tilted. A board longer than the tray, with a support behind by means of which the angle might be varied, and in front two right angle hooks screwed about six inches from the bottom of the board to hold the tray, served this purpose very well. The ear, previously shaved and sterilized with carbolic solu- tion and alcohol, hung down below the tray and a longi- tudinal incision was easily made with a sharp knife in the distended marginal vein. A transverse incision was to be avoided as it often allowed the vein to retract in its sheath and might sever the artery which accompanies the vein. A dull knife or a clumsy incision was apt to injure the walls of the vessels, so that clotting took place in the lumen of the vessels much more quickly than if a clean incision were made. With this method twenty, or even more, c.c. of blood might easily be drawn at one t11 me. A CONVENIENT METHOD FOR GROW ING AND STORING V IR U LENT PNEU MOCOCCI. AUGUSTUS WADSWORTH, M. D. Dr. Wadsworth called attention to the difficulty of keeping pneumococci alive and virulent for any length of time by ordinary culture methods. Any method which would do this, obviously would be of great service to any one conducting a long series of experiments with this organism. Other observers had suggested various spe- cial media with which this might be accomplished. Washbourn and Eyre had used agar smeared with rabbit's blood and had found their organism alive after some months. Although Dr. Wadsworth had not made an extended study as to how long the pneumococcus in exceptional instances might remain alive and virulent under these conditions, he had frequently found such tubes sterile after a few weeks and therefore could not rely on this method. Bezancon and Griffon had sug- gested the serum of young rabbits as a favorable medium for growing and raising the virulence of the pneumo- coccus, for the preservation of the living cultures they had recommended the serum of older rabbits. Guarnieri had suggested a nutrient peptone meat infusion medium containing four per cent. gelatin and .3% to .4% agar; but the virulence was maintained for only five to fourteen days and the cultures were not viable for much longer periods. Dr. Wadsworth then described a medium which in his experience had given more certain and better results than any he had as yet tried. He had chanced, while steriiized ascitic or pleuritic sera at low temperatures (60° to 70° C.) to obtain in one instance a soft jelly-like consistency of the serum, which still re- tained its original transparency. Being unable to obtain this a second time with other sera, he then tried adding Small quantities of agar (.3%) to the ordinary fluid serum, and found this made a very good substitute for the transparent semi-fluid medium. The serum to be used for this medium was drawn by sterile apparatus into sterile flasks and tubed as soon as possible. This was done to avoid the usual intermittent sterilization at low temperatures which often clouded the serum and could not be relied upon to destroy the spores of some contaminating organisms, the spores often failing to develop into vegatative forms in the undiluted serum. A dilution of the tubed serum with ordinary nutrient peptone agar in the proportion of about one part of agar to five of serum gave a semi-fluid medium which served all practical purposes. Dr. Wadsworth said it was well to avoid using media containing glucose or other sugar which might be fermented by the pneumococcus, as rapid death and degeneration of the cells took place in such media. He had found cultures of pneumococcus in glucose broth dead after three days' incubation. Compared with the ordinary agar or serum agar tubes the advantages were obvious and numerous:– there was practically no environmental change from that found in the body fluids of the animal; the growth was less compact; the products diffused more rapidly; and finally, subcultures were easily made with the platinum loop or with capillary pipettes by means of which colo- nies may be transferred from any part of the tube to fresh media. t In the semi-fluid serum the pneumococcus always grew luxuriantly along the puncture, forming colonies of considerable size, if allowed to reach its maximum growth. Microscopical preparations from such cultures showed less degeneration of the cells than in ordinary media. Furthermore, capsules were formed and were readily demonstrated by the use of gentian violet and washing with 20% copper sulphate, as recommended by Hiss. (Centra/b/att f. Bakt., 1902, Bal. XXXV., p. 302.) Cultures incubated for twenty-four hours or less and placed in the ice box at from 8° to 10° C., had been found to be alive and virulent for eight or more months in the pure serum medium, and three or more months in the serum agar substitute medium. The virulence of the pneumococcus had not been accurately determined but the results of animal inoculations had been compared at an interval of three months in the case of the growth in the serum agar tubes and had failed to show any ap- preciable loss of virulence in 1 c.c. doses. Peptone broth subcultures from an agar serum tube stored for three months in the ice box, in doses of 1 c.c. had killed by intrathoracic inoculations a medium sized rabbit in about eighteen hours with the usual local and general lesions of pneumococcus infection. In view of the recent researches on bacteriolysis, and more particularly those of Walker on the increase in virulence of micro-organisms when grown in the sera of specificially immunized animals, such a medium as he had shown, Dr. Wadsworth thought, might become of considerable practical service if only the blood serum of the immunized animal be substituted for the ascitic or pleuritic fluids he had used. The organisms might thus be kept in the sera of any animal for which it had be- come adapted for a number of weeks with but little bio- logical change. ZJascussion. DR. W. H. PARK asked Dr. Wadsworth whether he thought that the bacteria would live longer in a medium so prepared than they would if it contained more agar. He had himself used .7% agar. DR. WADsworth said that he thought that the ad- vantage of having the medium as fluid as possible were that one could get at the colonies more easily and that it allowed growth to spread out into the medium. On pure theory he thought that if the products of the or- ganism were allowed to diffuse through the media, the organisms might perhaps retain their viability and viru- lence longer than on solid media; but he did not know what the results of experiments would show in regard to this. He thought the pure serum was preferable to the agar serum. In other tubes of pure unheated fluid serum which were kept some time and had evaporated to a semi-fluid consistency, the pneumococcus apparently re- mained alive and virulent for a considerable time as in the other tubes, though accurate observations had not been recorded. He did not know just how long the virulence might be sustained in these tubes. He sug- gested that the less the serum was diluted the better. Extracted from The American Journal of the Medical Sciences, March, 1904. - - - -- - ------------------ - --- HAEMOLYMPH NODES.” By HUGHES DAYTON, M.D., ALUMNI FELLOW IN PATHOLOGY, COLLEGE OF PHYSICIANS AND SURGEONS, COLUMBIA. UNIVERSITY. THE earliest mention of the occurrence of nodes macroscopically resembling the spleen was that of Leydig, in 1857, who described their occurrence along the abdominal aorta. Since 1875 the presence of lymph nodes of a red color or containing red blood cells has been noted by a number of observers, and various explanations have been offered to interpret these findings. To Warthin, of the Uni- versity of Michigan, we are indebted for the most systematic and extensive study of these structures, and in his recent articles are contained the most convincing proof of the existence of the so-called haemolymph node as an organ Sui generis. It is not the purpose of the writer to review exhaustively the work of others in this field, or to describe in detail the histology of these nodes, but rather to present the results of a personal study of their individuality and functions. A summary of the work of others in this field had been prepared, but the literature has since been so satisfactorily abstracted by Warthin in his most recent publication that this and the complete bibliography are omitted. The histology is well presented in his other articles. These so-called haemolymph nodes have been found by various observers in man, monkey, OX, sheep, goat, horse, pig, dog, cat, weasel, ferret, stoat, rabbit, squirrel, rat, mouse, bat, water-vole, mole, fowl, and turkey. The structures denominated haemolymph nodes vary in size from that of a pinhead to that of a lymph node, though averaging smaller than the latter. They are dark red, brown, or mottled red and white, usually smooth, soft, and elastic, but easily ruptured. Few or many bloodvessels surround or enter the modes, in some cases at a hilum. Lymphatics are apparently absent in some cases. When ruptured the nodes resemble a blood clot. Each node possesses a delicate capsule of connective tissue with some non- striated muscle fibre. Just within the capsule is a peripheral blood sinus from which irregular prolongations run throughout the mode. Between the sinuses are cords of lymphoid tissue, and usually cell collections resembling lymph follicles toward the periphery of the * A study from the Department of Pathology of the College of Physicians and Surgeons, Columbia University, New York. 2 DAYTON : EIAEMOLYMPEI NODES. mode. In some cases the division into cortical and medullary portions is very indistinct, and the relative amount of lymphoid tissue is generally smaller than in a lymph node. The sinuses are traversed by an irregular reticulum, usually finer than that of a lymph node, this and the sinus walls apparently being lined with endo- thelium. Proliferation of the endothelium appears to furnish the large phagocytes lying in the sinuses and in some cases nearly filling them. An artery enters at and a vein emerges from the hilum. The artery divides into branches which run in trabeculae extending from the capsule through the sinuses. The finer branches sometimes appear to terminate in the sinuses; others break up into capillaries in the lymph cords, some of these capillaries appar- ently emptying into the sinuses. The sinuses terminate in veins which unite into one, emerging at the hilum. In some nodes small arteries enter at various points in the capsule, and in the transition forms afferent lymphatics may be seen entering the peripheral sinus. In some transition cases the blood and lymph sinuses seem to be separate systems; in others bloodvessels and lymphatics appear to mingle their contents in common sinuses. The sinuses contain varying numbers of red cells, either free or in various stages of disintegration within the phagocytic cells already mentioned, and leukocytes. The characteristic feature is said to be the occurrence normally, in the sinuses of a node, of destruction of red cells which have been brought to the node by bloodvessels. The various interpretations which have been placed upon the presence of red blood cells within the sinuses of structures re- sembling lymph nodes, or the occurrence of bodies macroscopically suggesting accessory spleens or red or mottled lymph nodes, are that they are: (1) neoplasms; (2) masses of newly formed splenic tissue; (3) lymph nodes engaged in the formation of new red cells; (4) lymph nodes whose sinuses contain blood as the result of con- gestion, hemorrhage, diapedesis, absorption of extravasated blood, etc.; (5) haemolymph nodes. The interpretation as neoplasms was advanced by but one observer, Mosler, who described the dark-red modules found in the greater and lesser omenta of a splenectomized dog and termed them “hemorrhagic telangiectatic lymphomata.” Their micro- scopic structure was the same as that described by others as haemolymph nodes. Tizzoni and others have interpreted them as newly formed splenic tissue, basing their view upon their occurrence in animals the spleens of which they had removed. The investiga- tions of Morandi and Sisto, and of Warthin appear to have conclu- sively disproved the splenic nature of these modules. The idea that the lymphoid nodules with sinuses containing red cells were structures engaged in the formation of red blood corpuscles has been supported by a number of the earlier observers, but recently by Retterer only. Some were led to this belief by finding in the sinuses large cells IDAYTON : EIAEMOLYMPEI NODES. 3 containing what appeared to be red blood cells, some of which they described as being extruded. These are now generally regarded as endothelial cells displaying phagocytic activity and ingesting the red cells. Retterer believes that lymph nodes produce blood plasma, leukocytes, and red cells, the last derived from lymphocytes by a haemoglobin change of the nucleus. The evidence of all other investigators appears to disprove this construction of the presence of red cells. The view that the appearance of red cells in the sinuses is due to hyperaemia, hemorrhage, diapedesis, or the absorp- tion of extravasated blood by normal lymph nodes has been advo- cated by several writers, of whom Saltykow is the most decided representative at present. His study of the subject has been careful and systematic. Examination of sixty autopsy cases has convinced him that the red nodes are merely lymph nodes whose sinuses contain blood brought by lymph vessels from regions in which hemorrhage has occurred, or entering them by diapedesis from capillaries in the node, or by hemorrhage from them into the lymphoid tissue or sinuses, the presence of red cells in the sinuses exciting the endothelial cells of the latter to intense phagocytic activity. His reasons for assuming that the nodes are merely lymph nodes are: in a large proportion of the cases an extravasation of blood is found in the node itself or the surrounding cellular tissue; in many nodes the vasa efferentia and subcapsular sinus are the only or chief containers of red cells, giving the impression that they are filled from without; the red nodes are found in the situations of normal nodes, and all transition forms to the ordinary type exist. The interpretation as organs sui generis, variously denominated haemolymph, haemal, and haemolytic nodes, dates from the time of Gibbes' preliminary report in 1884. Since the present writer began his investigations the ground has been covered with remarkable unanimity by Morandi and Sisto, Weidenreich, Lewis, and Warthin, to whose articles he would refer those desiring the bibliography, review of the literature, and full description of the histology. They agree that the haemolymph nodes are a distinct set of organs whose functions are the production of leukocytes, and especially the de- struction of red cells and to a less degree of leukocytes. Morandi and Sisto give as their reasons for ascribing to these organs a haemo- lytic function the constant occurrence in their sinuses of cells con- taining red cells in various stages of disintegration, and the increase of those phagocytic cells after splenectomy, and still more after subsequent administration of haemolytic substances. The writers mentioned agree also that there is a complete series of transition forms from the structure of the spleen to that of an ordinary lymph node. Warthin regards them as a set of structures subject to Variation, one form passing into another as the needs of the body require. He describes the formation of new haemolymph nodes as beginning with angiectatic dilatation of the capillaries of a fat lobule, 4 DAYTON : HAEMOLYMPIEI NODES. the fat cells becoming enlarged, the capsule of the lobule thickened. Lymphocytes are infiltrated along the walls of the capillaries; fat cells are absorbed, and some are converted into reticular cells; pro- liferation of endothelium in the dilated capillaries divides them into blood sinuses. Continued lymphoid formation, development of sinuses, and absorption of fat complete the transition into a haemo- lymph node. Progressive hyperplasia of lymphoid tissue encroach- ing upon the sinuses may transform this or a previously existing haemolymph into a lymph node. In the writer's own investigations haemolymph nodes have been observed and studied in the adult, infant, dog, cat, red deer, rabbit, and guinea-pig. Failure to discover them in the gray squirrel was probably due to incomplete examination of a single subject. In all experimental cases the nodes were removed immediately after death and fixed in Orth’s or Zenker's fluid, alcohol, or for- maldehyde solution. Celloidin was usually employed for em- bedding, occasionally paraffin. Haematoxylin and eosin were ordinarily used in staining. The experimental work was confined to the dog and rabbit, the normal condition of these animals being verified by autopsy and microscopic examination. I. Evidence in Favor of Other Interpretations than Organs Sui Generis. In support of the view that red cells in the sinuses are brough by afferent lymphatics from regions in which there is extravasation of blood is the fact that in a number of cases in which the sinuses of modes contained red cells large numbers were seen in surrounding lymph spaces. This is not at all conclusive, as similar collections of red cells were found in the lymph spaces surrounding nodes whose sinuses showed no characteristics of haemolymph nodes. In several cases nodes containing red cells were observed in which afferent lymphatics were filled more or less completely with red cells. In these the appearances of a haemolymph node were present, yet there was distinct proof that red cells were being brought by lymph vessels. An example is shown in the accompanying figure. This is reproduced from a photomicrograph of an axillary node removed during an amputation of the breast for carcinoma. In this case there was obviously extravasation of blood in the region whose lymphatics were tributary to the node. Several other such instances were found. The fact that in some instances red cells are brought to a node by afferent lymphatics does not prove that this is the invariable explanation of their presence. In a red deer ex- amined a few minutes after death, immediately following a wound with comminution of the pelvic bones and extensive laceration of the adjacent soft parts, large numbers of small dark-red nodes were found in the prevertebral region of the abdomen and thorax and DAYTON : H.EMOLYMPEI NODES. o along the renal vessels. The peripheral and central sinuses of these nodes contained many red cells but no evidence of recent or former phagocytosis, while in places where the sinuses were completely filled with red cells the adjacent lymphoid tissue was also so densely packed with red cells as to suggest a local hemorrhage. The blood- vessels of the nodes were greatly congested. There were no hemor- rhages in the neighboring tissues. While these nodes were so ex- tensively distributed as to suggest that local extravasation of blood could not have been responsible for their appearance, the complete absence of phagocytosis in all examined would suggest that the Lymph node, showing red blood cells in sinuses and in afferent lymphatic. red cells had entered the sinuses only shortly before fixation of the tissues. The presence of hemorrhages in the lymphoid tissue would also favor the view that these were ordinary lymph nodes with recent hemorrhages into the lymphoid tissue and thence into the sinuses, caused probably by great disturbance of the circulatory system incident to the traumatism and shock. With a view to determining the possible influence of circulatory disturbances in producing appearances which might be described as haemolymph nodes, the following experiment was carried out: A rabbit was anaesthetized with ether and the inferior vena cavaligated :: 6 L)AYTON : HAEMOLYMPEI NODES. in two places, just above and below the left renal vein. The animal was found dead in the morning, about twenty hours later. Viscera normal, except great congestion of the left kidney and spleen. In contrast to the negative results obtained in previous examinations for nodes containing red cells in normal rabbits, red corpuscles were discovered in six of the nodes studied. That circulatory disturbances will not, however, always cause the presence of red cells in lymph nodes was demonstrated by two other experiments upon rabbits. In one, ether anaesthesia, ligation of aorta, inferior vena cava, and right ureter en masse. Rabbit found dead next morning. Viscera normal, except congestion of kidney and desoluamation of cells of its tubules. In spite of the intense circulatory disturbances, red cells were found in the sinuses of only three of eleven nodes, and of the positive specimens one showed hemorrhages into the lymphoid tissue and blood in adjacent lymph spaces. In another rabbit: ligation of the right femoral vessels after bleeding from the artery, under ether anaesthesia, repeating this eighteen days later with the left femoral vessels, and killing a day later by severing the medulla. General anaemia; cheesy area surrounded by consolidation in lung; other viscera practically normal. Nodes from the immediate vicinity of the former and recent operations showed no red cells in their sinuses. In one mesenteric node they were found enclosed in phagocytes; in two others pigment masses were observed in such cells. Haemolymph nodes were found in cases in which there was no hyperaemia. These observations would seem to show that circulatory disturb- ances may cause the pressure of red cells in sinuses of nodes, but are not sufficient to account for their appearance in most cases. That the presence of red cells in the sinuses may be caused by toxic substances within the body was shown by subcutaneous injection of ricin into a normal dog. In all of twenty-one nodes from the cervical, axillary, bronchial, retrosternal, gastric, mesen- teric, retroperitoneal, and pelvic regions many red cells were found in the sinuses and in adjacent lymph spaces, with marked con- gestion of bloodvessels. Similar results, with marked endothelial hyperplasia in the sinuses, were obtained by ricin injection into a rabbit. While these experiments demonstrate that a poison in the circulation may induce the presence of red cells in the sinuses of lymph nodes the toxic action was too severe to draw strict inferences in regard to the behavior of nodes under comparatively normal circumstances. In three dogs examined, into the peritoneal cavities of which 1 per cent. solution of zinc chloride had been injected for experimental pur- poses, the abdominal cavities were found post-mortem to contain bloody exudate. In one animal red cells were found in the sinuses of eleven out of fifteen mesenteric and prevertebral nodes; in a second DAYTON : HAEMOLYMPH NODES. 7 dog, red cells in sinuses of six out of eight nodes from the anterior mediastinum; in the third, red cells or their remains free or in phagocytes in all but two of twenty-seven nodes from the pre- vertebral, renal, mesenteric and anterior mediastinal regions. In Some cases the adjacent lymph spaces contained red cells. The red corpuscles in these nodes were evidently brought by afferent lymphatics, although none of the sections studied showed their entrance, since at various points in their circumference the periph- eral sinuses were closely packed with red cells, and communicating portions of the sinuses contained phagocytes filled with red cells and some red cells free, while toward the centre of the node phago- cytosis was just beginning, and since adjacent lymph spaces were filled with red cells. It is unlikely that all of the nodes containing red cells were haemolymph nodes. Probably many were lymph nodes whose endothelium was exerting phagocytic powers in response to stimulation by the toxic substance, zinc chloride, or by some substance produced by its action, perhaps by damaged red cells from the peritoneal cavity. The occurrence of red cells in many of the nodes in these cases may, therefore, have been due to absorption of extravasated blood or to the presence of a toxic substance in the blood. That the presence of toxins does not necessarily cause phagocytic destruction of red cells in the sinuses was shown by negative findings in two rabbits. One of these had been injected with the toxin of the bacillus of rabbit septicaemia (Lartigau); the other had been subjected to the influence of the toxin of tubercle bacilli. To summarize: absorption of extravasated blood, circulatory dis- turbances, and toxic substances in the circulation may cause the appearance of red cells in the sinuses of lymph nodes, but there are many nodes in which this explanation is unsatisfactory. II. Evidence in Favor of Interpretation of “Harmolymph Nodes” as Organs Sui Generis. 1. Constant Occurrence. The constant occurrence of haemolymph nodes in the dog was shown by the following experiments: Four dogs were instantly killed by dividing the medulla, two of these Subsequently being found to have microscopically normal viscera, two chronic diffuse nephritis and associated cardiac lesions; one by bleeding while chloroformed, viscera microscopically normal; one by chloroform alone; three by intravenous injections of staining fluid or pigment in suspension, viscera normal, except congestion. In each of three cases a number of haemolymph nodes were found. The occurrence of all stages of phagocytic destruction of red cells in the sinuses in normal animals killed without unnecessary trauma- tism, and the nodes of which were hardened within a few minutes after death, shows that this presence of red cells in the sinuses 8 DAYTON : HAEMOLYMPH NODES. and this blood destruction were not the result of the manipu- lations or a pathological condition but evidence of the performance of a normal function of the nodes. In one guinea-pig killed by chloroform a haemolymph node was found, showing red cell destruc- tion by phagocytes. In two of three cats examined after killing with chloroform nodes were obtained, the sinuses of which contained red cells free and in phagocytes. Although red cells were found in lymph spaces in fat surrounding the nodes, there was no reason for attributing the presence of red cells in the sinuses to patho- logical conditions caused by killing with chloroform, or by rupture of bloodvessels while struggling, since some of these cells were already within phagocytes. In three normal rabbits, two with cholecystitis from bacterial injection, two injected with bacterial toxines, and one splenectomized and subsequently immunized to bullock's blood, red cells were not found in the sinuses, though others have detected them in rabbits, but pigment masses were observed in some phagocytes in sinuses. In human subjects nodes containing red cells were discovered in such varied cases as infants dying from asphyxia neonatorum, gastroenteritis, status lymph- aticus, and cerebral hemorrhage after forceps delivery; adults, from pneumonia, pulmonary and laryngeal tuberculosis and splenic anaemia. In a number of autopsy cases in which but few nodes were studied haemolymph nodes were not found. 2. Intimate Association with Ordinary Lymph Nodes. The nodes whose sinuses contain red cells are often closely connected with ordinary lymph nodes. In one normal dog examined two nodes were found so closely approximated as to be separated by only a few connective-tissue fibres, except at the point where blood and lymph vessels were situated. The circulatory relations could not be exactly determined. One was a typical lymph node with fine reticulum and no red cells in its sinuses; the other possessed a coarser reticulum and its sinuses contained many red cells. The two nodes were so closely connected that if the presence in the blood of some substance stimulating lymph nodes to accomplish the destruction of red blood cells was the cause of the one assuming the appearance described as that of a haemolymph node, it would seem certain that the same influence would have acted upon the other node. On the contrary, the general appearance of the two nodes was so different and the distribution of red cells in one so distinctly marked that it would appear there was a fundamental difference in their structure and circulatory connections. 3. Circulatory Relations. The most direct evidence of the exist- ence of haemolyph nodes as organs sui generis would seem to be the demonstration of bloodvessels directly entering the sinuses, though it is possible that diapedesis may be responsible in some cases for the presence of red cells in them. In both dog and man I have been able to discover what appeared to be direct termination DAYTON : HAEMOLYMPEI NODES. 9. of a capillary in a sinus, but the possibility of artefacts in so delicate a tissue as a capillary wall or the lining of a sinus makes such ocular evidence unreliable. Many hamolymph nodes appear to be more vascular than lymph nodes. The interstitial injections which have been employed by others seem unreliable, as the needle may pass into either a lymph or a blood sinus, or both, so that both systems may be injected, rendering differentiation uncertain. Two attempts to make a physiological injection of dogs through the femoral vein failed, one because the insoluble substance suspended caused embolism, the other because of the diffusibility of the material employed. In a third case insoluble Berlin blue suspended in normal salt solution warmed to body temperature was slowly injected into the proximal stump of a divided femoral vein of an anaesthetized dog, while bleeding was permitted from the distal stump in order to prevent undue increase of blood pressure. After ten or fifteen minutes of such injection the animal died. The pericardial, thoracic, and abdominal cavities were at once injected with 5 per cent. formaldehyde solution to fix the tissue cells, and the body placed in cold storage for thirty-six hours. All organs normal. In twelve of twenty-five nodes examined red cells were discovered in the sinuses. In only three was the Berlin blue found in the sinuses with the red cells, but in these three at least the coincident presence in sinuses of the recently injected pigment, together with fresh red cells not in phagocytes, would strongly point to a direct communication between the circulatory system and the sinuses of the nodes. 4. Histological Differentiation. The greater coarseness of the reticulum of haemolymph nodes as compared with that of lymph nodes has been confirmed by the present study, as has the lack of differentiation between cortical and medullary regions in typical haemolymph nodes. The frequent presence of mast cells and also of eosinophiles has been noted. The eosinophiles seem most frequent in haemolymph nodes and in lymph nodes in cases with extreme haemolysis such as splenic anaemia. This suggests an association between the eosinophilic granules and the destroyed red cells. Warthin's statement that phagocytes containing red and white cells are found in all sinuses of haemolymph nodes, while in lymph nodes they are usually confined to the medullary portion, appears to be borne out. While having no positive evidence to present in favor of the belief that transformation of fat lobules into haemolymph or lymph nodes, or of the latter into each other, occurs, the writer has observed, in several instances, appearances which favor this view. 5. Functions. No evidence of red cell formation could be found. The chief function is obviously destruction of red blood cells; others are formation of leukocytes, their destruction by phagocytes, and probably formation of blood plasma. They may also be concerned 10 DAYTON : EIAEMOLYMPEI NODES. in the production of eosinophiles and of phagocytes for the general circulation. That destruction of red blood cells is an intermittently exercised function and often a purely local process, not due to haemolytic substances in the general circulation, is shown by the occurrence simultaneously in different nodes or in different parts of the same node of its various stages. These stages are: red cells free in sinuses; apparently adherent to the periphery of the phagocytes; unchanged but contained in enlarged phagocytes; red cells disintegrating within phagocytes which are now stained more deeply red by eosin; red cells not seen, phagocytes deep reddish-yellow; phagocytes Smaller, staining normally, containing granules or globules of brownish-yellow pigment; phagocytes of normal size, pigment free IIl SII] UISéS. In a node physiologically injected with Berlin blue, for example, one portion of the sinuses contained red cells which had entered long enough before to have been nearly destroyed by phagocytes, while in another portion were fresh red cells which had entered with the injected pigment immediately before death. In explanation of this intermittent action it seems most logical to believe that in some cases the bloodvessels bringing the red cells become blocked temporarily by the mass of red cells and enlarged and proliferated phagocytes in the sinuses. Blood pressure in the node is probably an important factor. In other cases the intermittent action is difficult to explain, but it may be due in some to the presence of haemolytic substances in the blood plasma. These might cause changes in either blood cells or vessel walls which would lead to diapedesis into the sinuses. CONCLUSIONS. Both histological and experimental evidence is strongly indicative of the existence of the haemolymph node as an organ sui generis. In the light of our present knowledge the chief practical point, however, is to recognize the capability for phago- cytic destruction of red blood cells which is possessed to a high degree by certain lymphoid structures, rather than to dwell upon the individuality of the haemolymph node. The occurrence of transition forms from the node containing blood sinuses only to that with sinuses containing lymph alone renders a strict classification impossible. For practical purposes Warthin's grouping of all varieties under the heading of haemolymph nodes appears eminently satisfactory. To Dr. George C. Freeborn I am indebted for many valuable suggestions in connection with this investigation; to Dr. A. J. Lartigau for aid in the operative work, and to Dr. Edward Leaming for the accompanying photomicrograph. I)AYTON : HAEMOLYMPIFI NODES. 11 BIBLIOGRAPHY. Lewis. Journal of Physiology, 1902, vol. xxviii., Nos. 1 and 2. Morandi and Sisto. Archiv. per le Sci. Med., 1901, vol. xxv., No. 13. Retterer, Comptes Rendus de la Soc. de Biol., 1902, T. liv., p. 33. Saltykow. Zeit. für Heilkunde. Abtheil für Path. Anat., 1900, S. 301. Warthin. Journal of the Boston Society of the Medical Sciences, Vol. v. p. 415. Warthin. Journal of Medical Research, 1902, vol. ii. p. 435. Warthin. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCEs, October, 1902. Warthin. Transactions of the Chicago Pathological Society, November, 1902, vol. v., No. 8. Weidenreich. Anatom. Anzeiger, 1902, Ergänzungsheft, Bd. xxi., S. 47. A BACTERIOLOGICAL STUDY OF TRACHOMA, WITH REMARKS ON THE OCCURRENCE OF THE INFLUENZA GROUP OF BACTERIA IN CONJUNCTIVITIS." By ARNOLD KNAPP, M.D. FROM THE PATHOLOGICAL LABORATORY, THE COLLEGE OF PHYSICIANS AND SURGEONS, COLUMBIA UNIVERSITY, N. Y. N 1902–4 I examined bacteriologically I2O cases of tra- I choma. I was led to do this, thinking that owing to the systematic school inspection the cases which applied for treatment were unusually early ones, and that with the method of forceps expression a larger quantity of material could be obtained for examination than had been possible for previous investigators. The study of the bacteriology of trachoma, moreover, has recently received a stimulus by the publications of L. Müller of Vienna. Müller in an examination of trachoma in Egypt and Graz found an influenza-like bacillus in 59 out of I55 cases— more frequently in recent cases. He has never found this bacillus in non-trachomatous cases, and is inclined to regard it as the inciting factor in trachoma. Culturally and morphologically this bacillus cannot be distinguished from the influenza bacillus. Müller does not regard it, how- ever, as identical with the influenza bacillus, as no other symptoms of influenza were present. Müller's bacillus has the following characteristics–Morphology: Very small, short bacilli, generally twice as long as they are broad; * Read in part before the Section on Opththalmology, N. Y. Academy of Medicine, May 17, 1904. 463 464 * Arnold Knapp. their outline is not sharp ; rounded ends; they stain with difficulty, best with carbol-fuchsin ; Gram negative. Cul- ture: Exclusive growth on hemoglobin media, on which, after 24 to 48 hours, small, clear, dew-drop colonies appear, seen only with a magnifying glass. In the proximity of contaminating colonies, like those of the staphylococcus pyogenes albus, the colonies of the Müller bacillus are much larger. (This increase in development only occurs in the neighborhood of colonies of marked alkaline reaction.) This trait is also common to the influenza bacillus, and is spoken of, for short, as a form of symbiosis, though not with absolute correctness. In subsequent cultures threads are present. Animal experimentation was negative, while experimentation with the true influenza bacillus is variable (Lindenthal). t In the cases of trachoma which I examined, the symptoms had generally not been sufficient to draw attention to the lids. On examination the upper and lower retrotarsal folds were found thickly studded or diffusely infiltrated with follicles, which in some cases encroached upon the tarsus. There was very little secretion, though generally a history of the lids being closed in the morning was elicited. The conjunctiva seemed but slightly thickened and moderately red. I should like to add here that clinically these cases have been regarded by many ophthalmologists to be relatively benign and the measures used for their compulsory treat- ment unnecessarily severe. Though many of these cases were mild and amenable to treatment, I was able to follow two cases which notwithstanding treatment proceeded to the infiltration of the tarsal conjunctiva, presenting the rough thickened appearance which we know precedes the pannus and the other deleterious after-effects of trachoma. I think that we must recognize a variation in the intensity of the trachomatous process, possibly due to a variable toxicity and to other—especially hygienic—factors. During the first winter, the follicle contents obtained by expression from 80 cases were examined. Though this has been done before by careful investigators with negative re- A Bacteriological Study of Trachoma. 465 sults, it seemed to me of promise from the above-mentioned reasons, and from the fact that with the method of expres- sion—as the conjunctiva was not cleansed in any way, the infecting agent might be detected, even if only present in the conjunctival discharge or in the epithelium. BACTERIOLOGICAL EXAMINATION OF THE FOLLICLE CONTENTS. The follicle contents were thinly spread on cover glasses and stained in many ways for bacteria. Clear pictures were very difficult to obtain on account of the great quantity of cellular elements present. The best results were given by: (1) overstaining with Loeffler's methylene blue and decolor- izing with dilute acetic acid ; and (2), by the Weigert-Gram method. In general this examination was negative. In the first series of investigations, the follicle contents were directly streaked with the roller forceps on agar, glucose agar, ascitic agar (ascitic fluid #, agar #), and hemoglobin agar (human blood from the finger). Colonies of the S. p. a. and xerosis bacillus developed (excluding manifest contaminations). In the second series, the follicle contents were first trans- ferred to broth tubes and placed in the incubator; after 3–24 hours, they had become clouded. Streaks and pour plates were then made, but were all quickly overgrown by the S. p. a. In the next series, the follicle contents were carefully triturated, either with a blunt platinum wire on the side of the test-tube or in a sterile watch-glass or mortar, and the emulsion was streaked on the above media and pour plates were made. Besides the s. p. a. and xerosis bacillus, in four cases very small dew-drop colonies appeared on about the third day, which did not enlarge in size. Transplantation was only successful on hemoglobin media. Two series of the organisms have been cultivated to the sixth generation by frequent transplantations on hemoglobin agar. Control plants on serum agar were always negative. Morphology: short, thin bacilli. The pour plates showed S. p. a. colonies 466 Arnold Knapp. and particles of tissue which did not change. A more fluid medium was tried, consisting of # 7, agar and 5 % gelatine with pepton, glucose, and meat extract of neutral reaction. These plates were rapidly spread over by the s. p. a. In brief, the follicle contents were examined by many staining methods and cultivated on a large variety of nu- trient media, with the only noteworthy result that in four cases an influenza-like bacillus was found. BACTERIOLOGICAL EXAMINATION OF THE CONJUNCTIVAL DISCHARGE. During the past winter I have continued the investigation especially with a view of testing Müller's results. Only the conjunctival discharge was examined in forty more or less recent cases of trachoma (no scars), and with the use of a hemoglobin-containing medium. In these I found the in- fluenza-like bacillus in four. Two of the series have been kept alive for eight months by frequent transplantations on hemoglobin agar. Control transplants on serum agar were often made. This influenza-like bacillus, which was therefore found in eight cases, presented the following characteristics: small size, rounded ends, difficulty of stain—best with diluted carbol- fuchsin. Grows only on hemoglobin media as small dew- drop colonies scarcely visible to the naked eye. The culture can be continued indefinitely on proper media. At no time does a growth take place on serum agar. The bacilli were found to be non-pathogenic for rabbits, and subconjunctival injections in rabbits have produced no lesion whatever. It is evident that this bacillus agrees in every way with the one described by Müller, and which he has called the tra- choma bacillus. In the rest of this paper I shall for brevity also make use of this name to signify an influenza-like bacillus found in some cases of trachoma. Among the cases of trachoma examined, some were complicated with acute epidemic conjunctivitis. The Koch- Weeks bacillus was present, but I never found the Koch- Weeks and Müller bacillus together. The cases in which the Müller bacillus was found presented no clinical peculi- A Bacteriological Study of Trachoma. 467 arities. One of the most interesting presented the typical clinical picture of “acute trachoma.” In this case I found the greatest number of Müller organisms. ON THE OCCURRENCE OF INFLUENZA-LIKE BACTERIA IN THE CONJUNCTIVA. I wish now to go over to the second part of my paper which deals with influenza-like bacilli encountered in con- junctivitis. There are a number of bacilli which resemble the influenza bacillus, so that some authors speak of an in- fluenza group. This group, as far as the conjunctiva is con- cerned, may be said to be composed of the influenza bacillus, the pseudo-influenza bacillus, and the Müller bacillus, if we accept Müller's statement in regard to this last organism. The Koch-Weeks bacillus is regarded by some to be so similar to the influenza bacillus as to be identical with it— Kamen, Rymovitsch, Jundell; the opposite view is held by Zur Nedden, Axenfeld, Müller, Morax. I shall endeavor to show what we know of these organisms in conjunction with their appearance in conjunctival lesions. The Koch-Weeks Bacă//us : The cultural peculiarities of the Koch-Weeks bacillus, the well-recognized infectious agent of acute epidemic conjunctivitis, have recently been studied, especially by Morax, Weichselbaum and Müller, and Kamen. I have examined more than thirty cases and can but confirm the description given by these authors. The bacilli in cover-glass specimens present a characteristic pic- ture, especially if they are stained with diluted carbol- fuchsin for not less than ten minutes. The extremely sharp outline, the length as compared to the breadth, is sufficient for identification. Morax even believes that culture is un- necessary and misleading, from the possible superabundant growth of a saprophyte masking the real inciting organism. I have found that the organism grows best on serum agar (ascitic fluid 4, agar #). The colonies are small, clear, watery, opalescent, and visible to the naked eye. No ten- dency to confluence. “Symbiosis” well-marked. They grow on hemoglobin agar—the colonies are then smaller, clearer, and more watery—but the growth is not so profuse 468 Arnold Knapp. as on Serum agar. It is a most susceptible organism and extremely difficult to keep alive. I did not succeed in maintaining the growth beyond the third or fourth gen- eration. Morphologically it is easily differentiated from the rerosis bacillus by the Gram stain, which decolorizes it, and in cul- tures the colonies can be at present readily distinguished, as the xerosis colonies after a few days present a character- istic corrugated, leaf-like appearance. A word as to the xerosis bacillus. This non-pathogenic bacillus of the diph- theria group is a constant inhabitant of the conjunctiva and has long ceased to be regarded as having any pathogenic importance. It also grows best on serum agar. Its growth is retarded on hemoglobin agar and is scanty on plain agar, thus resembling the Koch-Weeks bacillus in the choice of media. The Influenza Bacillus : The influenza bacillus has been found in some cases of conjunctivitis (Zur Nedden, Morax, Jundell), frequently associated with influenzal infection of the respiratory passages. Jundell published a series of con- junctivitis cases in patients suffering from influenza of the upper respiratory passages. Zur Nedden has also observed cases of conjunctivitis where the influenza bacillus was ob- tained on bacteriological examination. In eight out of thir- teen cases other signs of influenza were present. The conjunctivitis in all of these cases was mild and without the presence of follicles. I recently examined bacteriologically a Severe case of pseudo-membranous conjunctivitis in an infant where both corneae were involved, and found a typical influenza bacillus. There was no evidence of trachoma. No other manifestations of influenza were present. This organism was non-pathogenic for the rabbit. The Pseudo-Influenza Bacillus. Pfeiffer, the discoverer of the influenza bacillus, also described a pseudo-influenza bacillus which he found in the pneumonias of children not of influenzal origin. The pseudo-influenza bacillus has also been found by Kossel and Hartmann in the otitis media of sucklings. The characteristics which are supposed to differ- entiate it from the influenza bacillus are that it is somewhat A Bacteriological Study of Trachoma. 469 larger, and in subsequent generations on artificial media is more apt to produce degenerative forms (threads). These involution forms have been shown (Zur Nedden) to depend on accidental factors and are influenced by modification of the culture media. It is non-pathogenic for animals. As the pathogenicity of the true influenza bacillus for animals is very uncertain, we must conclude that in the present state of our knowledge we cannot distinguish be- tween these organisms, namely, the influenza bacillus, the pseudo-influenza bacillus, and the Müller “trachoma bacillus.” & I have compared a true influenza series from a clinically well-marked case of influenzal pneumonia with the appa- rently genuine influenza bacillus obtained from the above- mentioned case of conjunctivitis without other influenzal manifestations, and with the Müller organism obtained from cases of trachoma. The morphology was the same, the cultural conditions were the same, and animal experimenta- tion was negative in all. If the absence of other clinical signs of influenza excludes a true-influenza infection, my severe case of pseudo-mem- branous conjunctivitis must be regarded as a pseudo-influ- enzal infection. On the other hand, do we not in the conjunctiva find pneumococcic and diphtheritic infections without systemic involvement P Zur Nedden (l. c.) de- scribed five cases of primary influenzal conjunctivitis. Müller, as Zur Nedden has correctly pointed out, has overlooked the fact that the influenza bacillus can cause conjunctivitis, and he has not been able to show that his bacillus is not identical with the influenza bacillus. It is therefore most probable that in my 8 cases out of I2O, and in the many cases of Müller where this influenza-like bacil- lus was found, the presence of this micro-organism was acci- dental and independent of the inciting agent. This I think is especially shown by the case of clinically “acute tra- choma' previously mentioned, where these bacilli were present in such abundance, as at the present day acute trachoma is always regarded as a mixed infection. The Koch-Weeks bacillus is surely not identical with the 47O Arnold Knapp. influenza organism. Morphologically it is larger and more sharply defined. Its cultivation succeeds best on serum agar, but it grows on human hemoglobin agar, while the influenza bacillus requires hemoglobin-containing media. The colonies in the cultures show distinct differences: the Koch-Weeks colonies are opalescent and bluish, while the influenza colonies remain transparent and like dew- drops. The Koch-Weeks bacillus is kept alive with great difficulty, while the influenza organism can be easily culti- vated for an indefinite time. My conclusions are as follows: I. An influenza-like bacillus identical with the Müller “trachoma bacillus’’ was found present in 8 out of I2O fresh cases of trachoma which I have examined. It was present in the greatest numbers in a case of clinically “acute trachoma.” 2. This organism could not be differentiated from the true- or pseudo-influenza bacillus—morphologically, cultur- ally, or by animal experimentations. 3. It seems probable that its presence in these cases of trachoma was accidental. - 4. There is an influenzal conjunctivitis without other manifestations of influenza. 5. The Koch-Weeks bacillus is not identical with the influenza organism. BIBLIOGRAPHY. KAMEN. C. f. Baż., 1899, vol. xxv., p. 40I. RYMovitsch. Wratsch, No. 20, p. 683, 1900–190I, quoted from Axen- feld in Kolle and Wassermann. JUNDELL. Influenza Conj. bei Säuglingen. Mittheilungen aus der Augen- Älinik Stockholm, vol. iii., IQO2, p. II. ZUR NEDDEN. A lin. Momatschr. A zºgen/élèzzik, January, 1900, vol. xxxviii., p. I73; Vol. xli., 1903. MORAx and BEACH. ARCHIVES OF OPHTH., vol. xxv., p. 54. L. MüLLER. Arch. f. A ugen/t/A., vol. xl., 1900, p. 13; Graefe's Arch., vol. lvii., Igo3, p. I 38. - AxENFELD. Ergebnisse, Lubarsch and Ostertag, 1895–1899. Handbuch path. Mikroorg. Kolle and Wassermann, vol. iii., p. 512. KNAPP, ARNOLD. Severe pseudo-membranous conjunctivitis in an infant, presumably due to the influenza bacillus. ARCHIVES OF OPHTHALMOLOGY vol. xxxiii., p. 386. Reprinted from THE JOURNAL OF INFECTIOUS DISEASEs, vol. I, No. 3, May 30, Igo4, pp. 463 515 • THE BACTERIAL PRECIPITINS.* CHARLEs No RRIs, M.D., Instructor in Bacteriology and Hygiene, College of Physicians and Surgeons, Columbia University; Former Fellow of the Alumni Association of the College of Physicians and Surgeons, New York. CONTENTS. I. INTRODUCTION. II. THE PRECIPITATING ACTION OF NoFMAL AND ANTIBACTERIAL SERA on BROTHS OF WARIOUS REACTIONS AND UPON VARIOUs BACTERIAL BRoth FILTRATEs. III. THE PRECIPITATING ACTION OF THE SERA of RABBITs ADAPTED To VARIOUS SPECIES OF THE Colon-TYPHOID GROUP IN THEIR RESPECT- IVE CULTURE FILTRATES. GENERAL PLAN of STUDY. Series I. a) Anti-B. No. 1 rabbit serum. b) Antityphoid rabbit serum. c) Anticoli rabbit serum. Summary of Series I. Are precipitins developed in the serum of rabbits which have adapted themselves to injections of peptone? Series II. a) Antityphoid rabbit serum. b) Anti-B. No. 1 rabbit serum. c) Anticoli rabbit serum. Summary of Series II. Series III. a) Anti-B. No. 1 rabbit serum. b) Anti-B. typhoid rabbit serum. RECORD OF FURTHER TESTs MADE witH THE ANTISERA of SERIES II AND III, witH THEIR CORRESPONDING AND witH ADDITIONAL FIL- TRATES. A STUDY OF THE PRECIPITATING VALUE OF SERA ADAPTED TO SPECIES OF THE CoION-TYPHOID GROUP FOR THEIR HOMOLOGOUS AND HETER- OLOGOUs CULTURE BROTH FILTRATEs. A DETERMINATION OF THE MAXIMAL LIMITS OF PRECIPITATION OF ANTISERA FOR THEIR COR- RESPONDING AND HETEROLOGOUS FILTRATES. IV. THE PRECIPITATING ACTION OF THE SERA OF RABBITS ADAPTED TO VARIOUs RACES OF THE B. DYSENTERIAE SHIGA, AND ALSO TO OTHER SPECIES OF THE So-CALLED DYSENTERY GROUP, WHICH FERMENT MANNITE, IN THEIR RESPECTIVE CULTURE FILTRATES. * Received for publication February 23, 1904. 463 464 CHARLES NORRIS V. THE PRECIPITINS DEVELOPED IN RABBITS ADAPTED TO VARIOUS BAC- TERIAL SPECIES WHICH DO NOT BELONG TO THE COLON-TYPHOID GROUP. 1. The precipitims developed by adaptation of rabbits to B. proteus. 2. The precipitins developed by adaptation of rabbits to B. prodi- giosus. 3. The precipitins developed by adaptation of rabbits to the Spiril- lum, cholerae and S. Metch/mikovi. 4. The precipitims developed by adaptation of rabbits to B. diph- theriae. 5. The precipitims developed by adaptation of rabbits to B. pyo- Cyanews. 6. The precipitins developed by adaptation of rabbits to human and bovine tubercle bacilli. 7. The precipitims developed by adaptation of rabbits to the Staphy- lococcus pyogenes awrews, to the pneumococcus, and to the strepto- COCG2,S. a) The precipitins developed by adaptation of rabbits to the Staphylococcus pyogenes aureus. b) The precipitins developed by adaptation of rabbits to the pneu- mococcus and to the streptococcus. Résumé of previous studies upon the precipitins developed by adaptation of animals to the pneumococcus and to the strepto- COCCUIS. VI. THE PRECIPITATING ACTION OF THE ANTI-B. No 1, ANTITYPHOID, ANTICOLI, AND ANTIDYSENTERY SERA UPON THEIR HOMOLOGOUs AND HETEROLOGOUs SALINE Solution ExTRACTS OF YOUNG AGAR CUL- TURES OF SPECIES OF THE COLON-TYPHOID GROUP. VII. RELATIONSHIPS ExISTING BETweBN THE BACTERIAL AGGLUTININS AND PRECIPITINs. VIII. CONCLUSIONS. INTRODUCTORY. A LONG series of complicated studies upon the body fluids made during the past few years has led to the discovery of a number of hitherto unknown or little understood substances which seem to be of great significance in the processes of immunization of man and the lower animals to bacteria or to their toxins, and in the adaptation of the body to a great variety of alien materials. Many of these substances are present in the blood serum under normal conditions, and they may be increased in amount, or entirely new substances, it seems, may be formed when the body adapts itself to foreign material artificially introduced into the veins, subcutaneous tissue, or the peritoneal cavity, or after adap- THE BACTERIAL PRECIPITINs 465 tation following upon the course of a bacterial infection or disease. The substances, whether normally present in serum or developed as a result of adaptation, are commonly designated “antibodies,” and the serum containing them is known as an “antiserum.” For example, the serum of a normal rabbit may contain a substance which agglutinates the typhoid bacillus—an “agglutinin” specific within certain limits for this bacillus. The normal agglutinating substances are, however, greatly increased in amount or in their activity by the adaptation of the rabbit to the typhoid bacillus or to its metabolic products artificially introduced into the body of the animal; or, if the agglutinin be not normally present, it may be formed de movo in response to the process of artificial adapta- tion. The agglutinin is an example of an antibody, and the serum in which it is developed is called an antityphoid serum. Besides bacteria and their products, various animal or vege- table substances, generally albuminous in nature, such as alien blood sera, serous fluids, or their isolated albumins and globulins, as well as milk, peptone (?), egg albumin, various ferments, and the cells of various organs, or their extracts, derived from differ- ent animals, induce, after introduction into the animal Organism, the formation of antibodies. These are of several kinds; thus, specific antitoxic, bactericidal, cellucidal, agglutinating, and pre- cipitating properties have been demonstrated in sera. As a result of the adaptation of animals to serum containing these antibodies, a second series of substances may in many instances be developed which neutralize or render inactive these antibodies. In this manner, antiagglutinins, antiprecipitins, anti- coagulins, antiamboceptors (antilysins), etc., are said to be formed. Rudolf Kraus, April 30, 1897, described a class of hitherto undiscovered substances in the Sera of animals immunized or adapted to certain species of bacteria. Kraus found that cholera, plague, and typhoid antisera, when added to their corresponding bacterial filtrates, cause precipitates. The substance of the serum which causes the precipitation was called a precipitin. On add- ing an antiserum, say of typhoid, to the heterologous filtrates of plague or cholera, for example, no such reaction, however, took place. The reaction between these antisera and their correspond- 466 CHARLES NORRIS ing filtrates is thus, according to him, strictly specific, and hence as diagnostic as the Gruber-Durham reaction of agglutination was at that time considered to be. This view for the bacterial pre- cipitins has never been seriously questioned,” although A. Was- sermann” observes that, as in the case of the agglutinins, bacteria belonging to species closely related possess similar precipitin receptors, indicating that agglutination and precipitation are not strictly specific phenomena. Recently Tupnix,” in an abstract, has indicated that bacterial precipitins are by no means as specific as one might infer from the articles which had up till then appeared. A few other references of similar import are found in the literature, but they need not be cited. On the other hand, the researches on non-bacterial precipitins have clearly shown that the precipitins which are developed in an animal of one species after adaptation to the serum of another species are not specific. These hemo- or sero-precipitins, as they are called, thus precipitate, not only the serum of the species of animal by the use of which they have been developed, but also the sera of species more or less closely related. They are, how- ever, markedly special, for a precipitin invariably gives a quicker and more copious reaction in the homologous than in a heterolo- gous serum. The general analogy existing between the action of the bac- terial precipitins and sero-precipitins led the writer to believe that a more careful study than had hitherto been attempted might reveal relationships between the bacterial precipitins similar to those existing between the sero-precipitins, as mentioned above. Furthermore, the constant association of the bacterial aggluti- nins and precipitins in antisera, and the striking points of simi- larity in the action of these two antibodies, seemed to indicate that between the bacterial precipitins relationships would be demonstrated similar to those which had been conclusively shown to exist between the bacterial agglutinins at the time these studies were begun—in the fall of 1901. Before undertaking the main part of our task upon the inter- * The term “bacterial precipitin” is used by the writer to indicate the substance in the serum of normal or adapted animals which induces the precipitation of bacterial broths, filtrates, or other culture extracts. THE BACTERIAL PRECIPITINS 467 relations of the bacterial precipitins, a series of observations were made to determine the precipitating action of normal rabbit serum upon various bacterial filtrates. Although bacterial precipitins have not hitherto been demon- strated in normal sera, the fact that they have been found in association with the agglutinins in the serum of animals adapted to different bacterial species—so constantly indeed that some observers consider these antibodies identical substances—seems to indicate that their presence in normal sera may either have escaped attention, or that, if present, they are in quantities so small that their detection is not possible.* On the other hand, the possibility of precipitates forming in mixtures of such complex fluids as nutrient broth and serum, by some action other than that induced by normal precipitins, cannot be summarily dismissed. These reasons led the writer to investigate the precipitating action of normal and various antibacterial sera upon plain nutrient broth and upon various bacterial filtrates. The tests which are now to be described are preliminary, and serve as controls for the subsequent series described later on. *The precipitins, however, unlike the agglutinins, are infrequently found in appreciable quantities in normal serum. According to Linossier and Lemoine,4 the hemo- or sero-pre- cipitins of one species for the serum of other species of warm-blooded animals do not exist in the serum of warm-blooded animals. Noguchi 5 has, however, found normal precipitins in horse serum, for the serum of some species of cold-blooded animals, and also that “the sera of some cold-blooded animals contain precipitins.” Noguchi states: “It might be said that the existence of normal precipitins for a given serum is more likely to occur in a widely than a closely related species.” He assumes that the relationship between the natu- ral precipitins and those produced by adaptation will be found analogous to that which exists between the corresponding natural and artificial agglutinins and hemolysins. Many observers have noted the fact that the development of adaptation precipitins for the serum of all species is more readily induced in the organism of a species not too closely related. Rabbit serum, according to Lamb,6 contains normally a precipitin for cobra venom, and also, according to Obermayer and Pick,7 for the dysglobulin of the whites of eggs. iſ The culture filtrates which were employed during the course of our investigations, when not otherwise so stated, were obtained from broth made up with Liebig’s beef extract, 5 g., 1 per cent. peptone, and 0.5 per cent. sodium chloride, the reaction being approximately 0.5–1 per cent. acid to phenophthalein. The broth cultures were passed through Berkefeld filters, after an incubation of several months at 37° C. The reaction of our filtrates was alkaline to litmus paper, except that of the diphtheria filtrate, which was acid. The pre- cipitation and agglutination tests were made in sterilized tubes of approximately the same size, and of narrow caliber. The macroscopic agglutination tests described in this paper were made with suspen- sions of eighteen-hour agar cultures in 10 c.c. of sterile normal saline solution. After Settling for half an hour, 0.9 c.c. of the emulsion, 0.1 c.c. of the serum, or of its dilutions in normal saline solution, was introduced into each tube. Controls were always made. 468 CHARLES NORRIS II. THE PRECIPITATING ACTION OF NORMAL AND ANTIBAC- TERIAL SERA. ON BROTHIS OF WARIOUS REACTIONS AND UPON VARIOUS BACTERIAL BROTH FILTFATES. Normal rabbit and ox serum, as well as antibacterial rabbit sera, were added to broths of various composition and reactions. These broths had a basic composition of 5 g. Liebig’s meat extract, 1 per cent. Witte's peptone, 0.5 per cent. NaCl; but varied in their reaction, as follows: 1 per cent. and 0.5 per cent. alkaline, neutral; 0.5 per cent., 1 per cent., and 1.5 per cent. acid, phenolphthalein being the indicator, or were made from the usual meat infusion of different reactions, or of broth made up without additions of pepton. The tests were made with the following proportions: to 0.5 c.c. of broth was added 0.1 c.c., 0.25 c.c., and 0.5 c.c. of serum. The tests showed that when normal rabbit serum is added to the meat extract broth of 1.5 per cent. acid reaction, small flocculi may develop after twenty-four hours at 37° C. In most cases, however, there is merely a fine granular deposit, which develops on the sides of the tube and is clearly seen only with a low power lens, the fluid remaining clear. This precipitation was never seen in broth of other reactions. In the usual meat infusion, pepton salt broth, normal rabbit serum produces no granular precipitate or flocculi. With normal ox serum no precipitation occurs in broths of different reactions. Normal rabbit and ox serum give rise to opalescence and later to cloudiness, with or without the formation of granules and flocculent precipita, in broth made up without the usual addition of pepton or salt. The most marked reactions are obtained in broth of 1 per cent. or 1.5 per cent. acid reaction. In 4 per cent. acid broth a marked turbidity is constantly developed. The reactions obtained in alkaline broth are slight. The nature of the slight precipitates was not determined; in Some cases they were possibly small coagula, due to late or uncompleted clotting of the Sera, or to the deposition of the cellular elements contained in the sera. Since the reaction of our filtrates was usually alkaline, the cloudiness and precipitates that are devel- oped in broth of 0.4 or 1.5 per cent. acid reaction can be disregarded in our subsequent tests. The insignificant and late precipitates which are occasionally formed may be disregarded, since they cannot be mistaken for the precipitates developed by the bacterial precipitins upon broth filtrates. The numerous tests made show conclusively that normal rabbit and ox serum, as well as various anti- bacterial rabbit sera, exert no precipitating action upon the usual peptone salt broth of slight alkaline or acid reaction. Of especial importance was the observation that normal rabbit serum does not precipitate the various bacterial broth filtrates,” which for the sake * The broth filtrates of the following species of the colon-typhoid group were tested: B. typhosus, B. coli “Escherich,” B. psittacosis, B. typhi murium, B. enteritidis, B, icteroi- des, B. Schottmüller (Seemann and Müller), B. Gwyn, B. Cushing, and B. No. 1, a species of the intermediate group of Durham, The species of the Dysentery group tested were the B. dysenteria e Kruse, B. dysenteriae “New Haven " (Duval), B. dysenteriae “Y” (Hiss). The filtrates of the following bacteria were also tested: B. proteus, B. prodigiosus, B. pyocyaneus, B. tuberculosis, human and bovine, B. diphtheriae, Sp. cholera Asiaticae, Sp. Metchnikovi, Staphylococcus pyogenes aureus, Pneumococcus, and Streptococcus pyogenes. THE BACTERIAL PRECIPITINs 469 of brevity need not be enumerated here. We are therefore warranted in drawing the conclusion that normal rabbit serum contains no determinable amount of normal bacterial precipitins. The few specimens of ox serum examined likewise indicate the absence of bacterial precipitins in this serum. The question of the presence or absence of normal bacterial precipitins in the serum of other warm- or cold-blooded animals was not determined. Having demonstrated the absence” of determinable amounts of precipitating substances in the serum of normal rabbits for the filtrates of the different bacterial species which were employed during the course of these studies—an observation in complete accord with those of previous investigators—the writer now pur- poses to describe a series of observations on the precipitating action of a number of antibacterial sera upon the filtrates of their corresponding and of heterologous species of more or less closely related bacteria: in other words, to test the specificity of the precipitins. III. THE PRECIPITATING ACTION OF THE SERA OF RABBITS ADAPTED TO WARIOUS SPECIES OF THE COLON-TVPHOID GROUP IN THEIR RESPECTIVE CULTURE FILTRATES. GENERAL PLAN OF STUDY. The plan followed in these researches was, first, to study the precipitating action of the serum of rabbits adapted to several distinct species of the so-called colon-typhoid group, in order to determine whether the precipitative relationships exist between various species of this group; and, secondly, to search for more remote relationships, as indicated by slight and late reactions between the bacilli of this and those belonging to other groups of bacilli—namely, B. proteus, B. prodigiosus—as well as those * In a strict sense the absence of precipitin haptins for bacteria in normal serum can- not be determined positively, for they may be present in quantities so slight as to escape detection. The writer's observations convince him that the precipitins and agglutinins in antibacterial serum, as a rule, bear the relative proportion, roughly speaking, of 1–100 and that, as a rule, the agglutinin must react in a dilution of 1-100 “macroscopic test '' before the presence of a precipitin can be certainly detected. The fact that normal rab- bit serum rarely agglutinates bacterial species above 1–100 accounts sufficiently, if the above argument is correct, for the non-detection or absence of the normal precipitins. The writer believes, as will be described later, that this fact cannot be regarded as proof of the identity of those two substances, for an antiserum may precipitate the filtrate of a given species of bacterium without agglutinating the bacterium ; and he believes, furthermore, that the precipitins may be formed for a given species without giving rise. to any, or only agglutinins capable of reacting in low dilutions. 470 CHARLES NORRIS bacteria belonging to other classes of bacteria, the spirilla and the cocci. These introductory remarks suffice to indicate the scope of the writer's studies. The first part of our investigations will now be described: - The species used in these series of tests were, first, a bacillus belonging to the hog cholera or intermediate group of Durham, marked B. No. 1; second, a collection culture of B. typhosus, designated B. typh. “Coll.; ” and, third, a colon bacillus which does not ferment saccharose.* Sets of rabbits were adapted to each of these species, and the antisera thus obtained were tested, as far as practicable, at the same time, so that their precipitating action upon the same filtrates could be carefully compared. For convenience the results obtained will be described in series. SERIES I. ANTI-B. No. 1 RABBIT SERUM. Rabbit I received 14 c.c. of heated and unheated broth cultures at inter- vals of several days (November 9, 1901, to January 1, 1902). Bled five days later. The serum agglutinated B. No. 1 up to 1–10,000 (microscopic test); the B. typh. “Coll.” to 1–2,000 (microscopic test). w The serum was tested upon the filtrates of the three species which had been grown in broth at 37° C. for twenty-nine days. The filtrates were markedly alkaline to litmus. TABLE I. THE PRECIPITATING ACTION OF THE ANTI-B, No. 1 SERUM UPON ITS CORRESPONDING FIL- TRATE AND UPON THE FILTRATES OF B. TYPHOSUs AND B. Coll. B, No. 1 filtrate: Anti-B. No. 1 serum: |Tubes kept at 37° C. 2–3 hrs., Cloudiness and 1 c.c. 0.5 c.c. flocculent precipitates which soon settle. B. typh, filtrate: Anti-B, No. 1 serum : Cloudiness is slightly delayed when com- 1 c.c. 0.5 c.c. pared with the homologous filtrate, but at 5 hrs. the reaction is only somewhat less marked. Same phases of reaction. B. coli filtrate: Anti-B, No. 1 serum : | Same, phenomenon, only onset of reaction C, C 5 c.c. is slower, and the precipitum, formed is less than in case of typhoid filtrate. Tests were then made with similar quantities of the filtrates, but with 0.1 c.c. of serum. With B. No. 1 filtrate the various phases of reaction occurred as rapidly, the precipitum being, however, less abundant; with the typhoid filtrate the reaction was slightly delayed when compared with that * B. No. 1 and the colon bacillus were isolated from the organs of a patient dying from an infection of five days’ duration, clinically a case of severe gastro-enteritis with toxemia. - - THE BACTERIAL PRECIPITINs 471 in the homologous filtrate; a similar, but less marked, reaction was obtained with the colon filtrate.* - - To avoid the criticism that we were possibly not dealing with a true typhoid bacillus, the B. No. 1 serum was tested, with identical results, against a filtrate of the well-known typhoid bacillus isolated by Pfeiffer, and known as the “Pfeiffer” culture. On account of the saprophytic condition of the “Pfeiffer” culture, which was only slightly motile, and the rapid pseudo- clumping of its broth cultures and agar Suspensions in controls, the agglu- tinative value of the serum for this bacillus could not be determined. We thus see from our experiments that the anti-B. No. 1 serum precipitated in dilutions of 1–3 and 1–10, not only the corresponding, but the heterologous, typhoid and colon filtrates; the onset of the reaction cloudiness, however, developing earliest in the homologous filtrate. The amount of precipitum ultimately formed in the B. No. 1 and typhoid filtrates was apparently similar. ANTITYPHOID RABBIT SERUM. At intervals of three to six days Rabbit II received subcutaneously ten inoculations of broth cultures, at first heated and then unheated, in all about 25 c.c. Agglutination value, 1–10,000, complete in three hours by macro- scopic test; 1–5,000, complete in one hour. The serum did not agglutinate B. No. 1 or the B. coli. -- - TABLE II. THE PRECIPITATING ACTION OF THE ANTITYPHOID SERUM UPON ITS CoRRESPONDING FIL- TRATE AND UPON THE FILTRATES OF B. No. 1 AND B. COLI, B. typh, filtrate (Coll.): Antityph. serum: At 37° C.: 1 hr., cloudiness; 2 hrs., flocculi; 1 c.c. 0.5 c.c. and later, copious precipitate. B. No. 1 filtrate: Antityph. serum : At 37° C., 2 hrs., cloudiness, flocculi; and - 1 c.c. 0.5 c.c. later, copious precipitate. B. coli filtrate: Antityph. Serum : No reaction at 2 or 4 hrs. ; at 18 hrs., small .C. 0.5 c.c. precipitate. In a filtrate of a 1 per cent, mannit-broth typhoid culture grown for two weeks at 37° C., having a terminal acid reaction, the onset of the reaction was delayed to twelve hours, and the precipitum formed was less copious than that formed in the B. No. 1 filtrate. With the anti-B. No. 1 serum, this filtrate likewise gave a delayed reaction. These experiments show, therefore, that the antityphoid serum precipitates its corresponding filtrate and the B. No. 1 filtrate in *In judging the presence or absence of slight precipitates, to secure accuracy it is imperative to examine the deposit for bacteria; for non-cloudiness in serum mixtures does not signify absence of bacterial growth. When in doubt, tests were made by culture and by examination of coverslips. Usually there is no difficulty in distinguishing between precipi- tates and bacterial bottom growths, especially if the development of the reaction is watched. 472 CHARLES NORRIs the dilution of 1–3 and 1–5; the reaction being quicker, however, but not more copious, in the homologous filtrate obtained from a broth culture of the same age and terminal reaction. The reac- tion in the colon filtrate was considerably delayed. That the age and perhaps the terminal reaction of the filtrate exert considerable influence upon the time of onset of the precipitation and the amount of precipitum is shown by the observation that the anti-B. No. 1 serum gives rise to a quicker and more copious reaction in the alkaline typhoid filtrate (four weeks' broth culture at 37° C.) than does the antityphoid serum in the acid typhoid filtrate obtained from a two weeks’ mannit-broth culture. ANTICOLT RABBIT SERUM. Rabbit III, at intervals of three to five days, received subcutaneous inoculations of heated and unheated broth cultures, in all 24 c.c., one to four days’ growth at 37° C. (January 3 to February 10, 1902). Bled several days after last injection. No agglutination of the homologous colon bacillus either by the macroscopic method or by the hanging drop. A slight tendency to thread formation was observed, which ceased in a dilution of 1–1,000. With a serum dilution of 1–10, complete settling occurred only after twenty-four hours at 37° C. TABLE III. - THE PRECIPITATING ACTION OF THE ANTICOLI SERUM UPON ITS CORRESPONDING FIL- TRATES AND UPON THE FILTRATES OF B. N.O. 1 AND B. TYPHOSUS. * Coli filtrate: Anticoli serum : Cloudiness in all tubes in 1 hr. at 37° C., 0.5 c.c. 0.05 c.c. which increases rapidly. 6, hrs., well- 0.5 0.10 marked precipitum, most copious in tube 0.5 0.15 containing 0.25 c.c. serum. The fluid in 0.5 0.25 all the tubes becomes clear. B, No. 1 filtrate : Anticoli serum : At 6 hrs., a slight precipitate in the form of 0.5 c.c. 0.10 c.c. fine granules has appeared on the sides of 0.5 0.25 - the tubes. At 24 hrs. in the tube contain- ing 0.25 c.c. of serum the precipitum com- ares in amount to that formed in the omologous filtrate 0.05 c.c. of serum. B. typh. Coll. filtrate: Anticoli serum : Similar reactions to that obtained in B. No. 0.5 c.c. 0.10 c.c. 1 filtrate. 0.5 0.25 4 B. typh. Pfeiffer fil. : Anticoli serum : Similar delay in reaction to that obtained }; C.C. §§ C.C. with Typh. Coll. filtrate. 0. .25 Thus the anticoli serum in dilutions of 1–6 precipitated the correspond- ing coli filtrate, and also the typhoid and B. No. 1 filtrates. The onset of the reaction in the heterologous filtrates was, however, delayed, and the amount of precipitum formed was considerably smaller than in the colon filtrate. SUMMARY or SERIES I. The results of the first series of tests may be briefly summa- rized. Precipitative relationships exist between three species of THE BACTERIAL PRECIPITINS 473 the colon typhoid group; namely, the intermediate B. No. 1, the B. typhosus, and the B. coli, micro-organisms distinguished from each other by constant biochemical characters of differential value. ARE PRECIPITINS DEVELOPED IN THE SERUMS OF RABBITs which HAVE ADAPTED THEMSELVES TO INJECTIONS OF PEPTONE 2 The rabbits of Series I described above were adapted by means of the usual peptone-broth cultures. Myers,” however, had noted the fact that Witte's peptone gave rise to the formation of substances in the serum of the animals undergoing adaptation which precipitated peptone solutions. This observation led us to suspect that the precipitative relationships noted above between filtrates of various bacterial species, distinct at least as to their bio- chemical characters, were possibly due alone to the development of a peptone precipitin in the sera of the rabbits of Series I which had been inoculated, as mentioned above, with peptone-broth cultures. To remove all doubt upon this score, the writer was led to repeat Myers's observations, in order to test his statement that the inoculation of peptone in the animal organism is followed by the formation of substances which precipitate peptone solutions. Our experiments and the literature bearing upon this topic may be briefly reviewed. Two rabbits were treated with 108 c.c. of a (Witte's) 4 per cent. peptone Solution, six intraperitoneal inoculations being made at intervals during four weeks. The sera obtained did not cause even an opalescence in fresh peptone solutions, when added either in equal volumes or in the proportion of 1–5, after twenty-four hours at 37° C. A rabbit treated with large quantities of broth likewise failed to develop precipitating agents in its serum. Intra- venous inoculations (five to six inoculations of 0.5 g. of Witte's and Merck's peptone) have likewise proved negative in the hands of my colleagues, Drs. Wood and Jessup, to whom I am indebted for the information. Myers's observations, although widely quoted and accepted, have likewise not been confirmed by several other investigators. He claims to have obtained precipitins in the serum of rabbits that had adapted themselves to solutions of Witte's peptone. In an article published previous to Myers's, Tchistowitch" states that he had failed to obtain peptone precipitins in the serum of rabbits adapted to peptone. Buchner and Geret” cast grave doubts upon the nature . of the peptone precipitum obtained by Myers, since it gave no biuret reaction, and, furthermore, because the antipeptone serum after inactivation by heat was reactivated by normal serum. They endeavored to obtain peptone precipi- tins by adapting rabbits to peptone, prepared according to Kuhne's method. The rabbits yielded a serum which, when added to a 2 per cent. peptOne normal saline solution, gave rise to cloudiness and a fine precipitate composed of so-called globulites, which were found to be crystals of barium sulphate of unusual form. Rabbits treated with a single injection of 5 c.c. of ox serum after twenty- four hours likewise yielded a serum which developed these crystals. These observers found that the peptone of commerce invariably con- tained barium salts. The solutions of peptone freed from the barium salts 474 - CHARLES NORRIS were, however, not precipitated by the sera of the rabbits which had been adapted to peptone or to serum. - Michaelis also failed to obtain precipitins in the serum of rabbits which had been treated for six weeks with subcutaneous inoculations of a 10 per cent. Solution of Merck's egg peptone (albumose); the receptors which in the whole albumin molecule induce precipitin formation being, Michaelis believes, destroyed by the pepsin digestion to which the egg albumin has been Sub- jected. Similarly with “Peptonum siccum Riedel,” which is a mixture of the proteoses derived from the fibrin of ox serum, no precipitating serum was obtained. Although Kowarski obtained precipitins for the albumoses of seeds, Michaelis considers that these albumoses are not comparable to the pepsin albumoses of animal origin, and that, even if in truth albumose and peptone precipitins can be formed, the receptors which would induce the for- mation of the egg-peptone antibodies cannot be identical to those of egg albumin, for an egg albumin precipitin does not precipitate egg-peptone solu- tions.” Obermeyer and Pick” have likewise failed to obtain peptone precipitins, although they noted occasionally slight clouding of the peptone solutions by antipeptone serum. That the question is not finally settled is indicated by the recent investigation of Rostoski and Sacconaghi,” who report that after several failures they succeeded in obtaining antibodies or precipitins in the serum of rabbits which had been inoculated with the peptic and tryptic cleavage products, of horse serum albumin. The sera precipitated the solu- tions of albumin and those of all the cleavage products, considerable quanti- tative differences being observed in the reaction, the weakest being obtained in peptOne solutions. Rostoski assumes that the immunizing body, whether it be a special side molecule of the giant albumin molecule, or a body which is merely intimately attached to the albumin molecule, is variously resistant to digestion, accord- ing to its association with the different cleavage products. In some cases it is more resistant than the albumin body itself, especially to the action of trypsin. - Ascoli has observed that antisera precipitate the cleavage products of albumin. Thus antihorse and antiox serum of the rabbit precipitate in a slight degree Witte's peptone.* He assumes that digestion destroys some of the groups of the albumin molecule (side chains or receptors of Ehrlich), whereas other groups, those genetically related to the peptone molecule, are left unaffected. P. A. Levene" claims to have shown that a precipitin active for a given proteid is capable of forming precipitates with the primary digestion products of these proteids. - - A further proof that the precipitative relationships which had been shown to exist were not due to the formation of precipitins for peptone was furnished by the fact that the antisera of Series I do not precipitate our con- trols of plain peptone broth, or even 1 per cent. peptone solutions, Never- theless, to remove all possibility of error, all subsequent rabbits “in this series of studies” were inoculated with emulsions of agar cultures, and not with peptone-broth cultures. * Possibly this is accounted for by the globulite reaction of Buchner and Geret. THE BACTERIAL PRECIPITINs 475 SERIES II. The tests with the antisera of animals of the first series adapted to the three species above mentioned were preliminary. Those made with the sera of the rabbits of Series II, adapted to the same species, are now given, so far as possible in tabulated form, for brevity of presentation. The rabbits of this series, as well as all subsequent animals, were adapted by inoculations of salt emul- sions of agar cultures, grown for twenty-four hours at 37°C. The early inoculations were made with heated and the later with living cultures. - ANTITYPHOID RABBIT SERUM. Rabbit IV received several inoculations.” Agglutination limit of its serum was found to be 1–1,000 for B. typh. “Coll.” No action on B. No. 1 or B. coli. TABLE IV. THE PRECIPITATING ACTION OF THE ANTITYPHOID SERUM UPON ITS CORRESPONDING FIL- TRATE AND UPON THE FILTRATES OF B. No. 1 AND B. COLI AND OF WARIOUS OTHER SPECIES OF THE COLON-TYPHOID GROUP. Typh. fil. (Coll.): i Antityp. serum 0.5 c.c. 0.1 c.c. At 2 hrs. at 37C. *: At 7 hrs. : pre- || At 19 hrs. : scanty cloudiness. cipitation has precipitum. settled out. 0.5 0.25 Fine flocculent At 7 hrs. ; pre- Moderate amount precipitate. cipitation has of precipitate. settled out. B. No. 1 fil. : e 0.5 c.c. 0.1 Negative. Negative. Negative; 0.5 0.25 Negative. Negative. Distinct flocculent - precipitate. P. coli fil. : 0.5 c.c. 0.1 Negative. Negative. Small flocculi. , , 0.5 - 0.25 Negative. Negative. Precipitum. B. icteroides fil. : : - - 0.5 c.c. 0.25 A. 19 hrs. : nega- At 43 hrs. : slight cloudiness? IWe, B. psittacosis fil. : 0.5 c.c. 0.1 Slight precipi- tum. 0.5 0.25 Abundant pre- cipitum. B. typhimurium fil. : - 0.5 c.c. 0.1 At 2 hrs. : nega- At 19 hrs. : small flocculi. tive. - 0.5 0.25 Slight cloudi- || At 19 hrs. : small precipitate. IlêSS, B. dysen. Kruse fil.$ B. dysen. Flexner fil.: - * - 0.5 c.c. 0.25 Small precipitate. B. enteritidis : - • O C.C. 0.1 At 2 hrs. : nega- || At 7 hrs. : nega- At 19 hrs. : negative. tive. tive. - 0.5 0.25 At 2 hrs.: nega- At 7 hrs. : nega- |Small flocculi. tive. tive. * Rabbit IV developed snuffles during adaptation. t The typhoid, colon, and B. No. 1 filtrates in these and the subsequent tests were obtained from meat-extract broth cultures grown at 37° C. for ninety-seven days. f The following broth filtrates of same length of incubation at 37° C. were also tested. $ We are indebted to Professor Flexner for these cultures, which do not ferment man- nite, and therefore belong to the true Shiga type, incorrectly called “alkaline” type of B. dysenteriae. 476 CHARLES NORRIS A staphylococcus filtrate gave no precipitate after forty-three hours. It has been shown that, although the agglutinating and precipitating action of the antityphoid serum was considerably weaker than that of the typhoid antiserum of the first series, nevertheless the serum precipitated the filtrates of all the colon-typhoid group tested, the negative result obtained with the filtrate B. icteroides being most probably due to the scanty growth of the bacillus in the broth from which the filtrate was obtained. ANTI-B. No. 1 SERUM. Rabbit V between March 19 and April 2, 1902, received five agar cultures; bled April 18. The following microscopic agglutination tests were made with the serum upon B. No. 1: 1–20,000: nearly complete in 94 hr.; complete in 3 hrs. at 37° C.; and 1 40,000, complete at 6 hrs. The serum tested upon the typhoid bacillus gave the following reaction: 1–100: at 9% hr. reaction begins, large clumps, but fluid still cloudy. 1–1,000: at 3% hrs., large clumps; tube is nearly clear. 1–10,000: negative at 3% hrs.; at 5 hrs., small clumps. At 5 hrs. none of the tubes are, however, perfectly clear. TABLE W. THE PRECIPITATING ACTION OF THE ANTI-B, No. 1 SERUM UPON ITS CORRESPONDING FIL- TRATE AND UPON THE FILTRATES OF B. TYPHOSUS, B. COLI, AND OF WARIOUS OTHER SPECIES OF THE COLON-TYPHOID GROUP. B. No. 1 filtrate: Anti-B. No. 1 serum : 0.5 c.c. 0.10 c.c. Instant cloudiness. 0.5 0.25 Instant cloudiness; precipitum at room temperature forms in less than % hr. B. typh, filtrate: g 0.5 c.c. 0,10 Instant cloudiness; at 1% hrs., fine pre- cipitum. 0.5 0.25 Instant cloudiness; at 1% hrs., fine precipi- tum ; at 6 hrs., flocculent precipitum. B. colon filtrate: 0.5 c.c. } O After several hours, cloudiness. 0.5 5 At 1% hrs., cloudiness; at 6 hrs., a finely g flocculent precipitate. B. psittacosis: 0.5 C.C. 0.10 Immediate cloudiness; reaction in both 0.5 0.25 tubes complete at 6 hrs. B. typhi murium : 0.5 c.c. 0.10 As above. 0.5 0.25 B. enteriiidis: 0.5 c.c. 0.2; 0.5 0 5 Cloudiness at 1% hrs. : precipitate at 6 hrs. .10 Cloudiness at 1% hrs. ; none at 6 hrs., but * slight precipitate at 18 hrs. B. icteroiºsº ..) C. 0.5 ; Both tubes show slight precipitate at 19 hrs. tº .25 B. dysenteriae— Flexner and Kruse: 0.5 c.c. 0.25 No cloudiness at 6 hrs. 0.5 0.1 At 19 hrs., slight flocculi in both tubes. Staphylococcus filtrate: No reaction after 72 hrs. THE BACTERIAL PRECIPITINS 477 In the hanging drop, the following reactions were obtained with the typhoid bacillus: 1–20: complete clumping at 1 hr.; also at 24 hrs. 1–200; complete reaction; clumps are small, but slightly motile. No free bacilli. w 1–20,000; at 94 hr. small motile clumps; free motile bacilli. These agglutinative relationships are of considerable interest, especially in connection with those we have demonstrated to exist in the case of the precipitins.” The onset of reaction and amount of precipitum developed in the filtrates of B. psittacosis, and typhi murium and the homolo- gous B. No. 1 filtrate were similar, the typhoid filtrate perhaps giving a less copious precipitum. With B. enteritidis the reaction was less marked, although the addition of 0.25 c.c. of serum to 0.5 c.c. of filtrate gave rise to a copious precipitum at six hours. The filtrates of the dysentery bacilli gave a flocculent precipitate at nineteen hours with 0.1 c.c. and 0.25 c.c. of serum. It has thus been demonstrated with anti-B. No. 1 that a serum adapted to one species of the colon-typhoid group precipitates, in dilutions of 1–6, the filtrates of many different species of this group. *Castellani and Durham, Theobald Smith, H. Bruns and H. Kayser, 15 and many others, have recorded numerous examples of agglutinative affinities between the members of the colon-typhoid group. Carl Sternberg has shown that antityphoid serum may agglutinate, in dilutions of 1–5,000, several examples of so-called paratyphoid bacilli obtained from potable waters. The interreactions described above between the typhoid bacillus and the anti-B. No. 1 serum do not always occur, as shown later on. No antityphoid rabbit serum has been obtained which agglutinated the B. No. 1, except in 1–10 dilutions, although such sera frequently gave rise to more copious precipitates in the B. No. 1 filtrates than did some of the weaker anti-B. No. 1 sera, especially when tested in dilutions of 1–10 or 1–5. This statement does not apply to the reactions obtained with higher dilutions of serum, the homologous serum always precipitating its corresponding filtrate in higher dilution than does the heterologous serum. fThe B. psittacosis and B. typhi murium, both obtained from Král, cannot be differ- entiated from our B. No. 1 by cultural tests. In fact, the B. psittacosis is agglutinated and its filtrate precipitated by the anti-B. No. 1 serum quite as effectively as the homologous bacillus, in some tests even more so. The B. typhi murium, however, does not agglutinate in such high dilutions or furnish quite as copious precipitates as does the B. No. 1. Never- theless, as detailed later on, this was determined only by a careful estimate of the maximal serum dilutions which determine precipitation. Whether an animal immunized against B. No. 1 is also protected against these two bacilli of the intermediate group was not deter- mined, although it seems most probable. The B. enteritidis is not agglutinated by the anti-B, No. 1 serum at all; by the antityphoid sera often in dilutions of 1–100 and 1-200. (Durham has made similar observations upon the action of antityphoid serum on B. enteritidis.) 478 CHARLES NORRIS ANTICOLI SERUM. Rabbit VI, between March 1 and April 2, 1902, received seven agar cul- tures; bled April 7. TABLE W. THE PRECIPITATING ACTION OF THE ANTICOLI SERUM UPON ITS CORRESPONDING FILTRATE AND UPON THE FILTRATES OF B. No. 1 AND B. TYPHOSUS, AND OF WARIOUS OTHER SPECIES OF THE COLON-TYPHOID GROUP. Colon fil. : Anticoli ser. : * 0.5 c.c. 0.1 c.c. Immediate At 5% hrs., At 19 hrs., complete. cloudiness; flocculi still at 1 hr., floc- floating. culi... 0.5 0.25 Immediate Reaction com- cloudiness; plete. at 1 hr., floc- culi. - Bºphi murium - 0.5 c.c. 0.10 At 1 hr., nega- | A 5% hrs., At 19 hrs., moderate 0.5 0.25 tive. cloudiness? flocculent precipi- tum in both tubes. B. No. 1 fil. : 0.5 c.c. 0.10 Cloudiness? Slight cloudi- || In 0.1 c.c., flocculent 0.5 0.25 ness in 0.25c.c. precipitum less abundant than ho- * mologous tube. B. psittacosis fil. : - 0.5 c.c. 0.10 Cloudiness? Marked cloudi- || In 0.1 c.c., cloudiness; 0.5 0.25 In GSS, 0.25 c.c., m a r ked * cloudiness. B. typhoid fil. : - 0.5 c.c. 0.10 Cloudiness? Marked cloudi- | Both tub e s show 0.5 0.25 DeSS, moderate amount of precipitum. B. enteritidis fil. : g g - 0.5 c.c. 0.10 Negative. Slight cloudi- Both tubes show 0.5 0.25 ness in both. moderate amount of precipitum. B. icteroides fil. : 0.5 c.c. 0.10 At 1 hr., nega- || At 5% hrs., At 19 hrs., negative. 0.5 0.25 tive. negative. - B. dysenteriae fil., Kruse and Flex- - In OI! : 0.5 c.c. 0.10 At 1 hr., nega- || At 5% hrs., Very, slight, reaction 0.5 0.25 tive. negative. in both tubes. Staphylococcus fil. : - 0.5 c.c. 0.10 At 1 hr., nega- || At 5% hrs., In 0.25 c.c. tube, a 0.5 0.25 tive. negative. slight precipitate on sides. except that of B. icteroides in dilutions of 1–5. The anticoli serum thus causes precipitation, filtrates of bacilli belonging to the intermediate group, but also in the filtrates of the dysentery bacillus. The anticoli serum thus precipitated the filtrates of various and distinct species of the intermediate and dysentery groups, staphylococcus filtrate it has no action. not only in the Upon the THE BACTERIAL PRECIPITINs 479 SUMMARY OF SERIES II. The results of the tests of Series II may be summarized briefly. The sera of three rabbits, immunized respectively to B. typhosus, the intermediate B. No. 1, and B. coli, precipitated in dilutions of 1–3 and also 1–6, not only their homologous, but also a large number of the filtrates of heterologous, species belonging to the same so-called colon-typhoid group. Precipitative relation- ships between the species of the same group are thus indicated. SERIES III. In order to confirm the observations made with the antisera of the rabbits of Series I and II, and also to enable us to widen the scope of our investigations, a third set of rabbits was adapted to the species of the colon-typhoid group, namely, B. No. 1, B. typhosus, and B. coli. ANTI-B. No. 1 SERUM. Rabbit VII received six agar cultures, between April 4 to May 26, 1902; bled June 1. Agglutinating value: 1–5,000 to 1–10,000. No agglutinating action upon the typhoid or colon bacillus. TABLE VI. THE PRECIPITATING ACTION OF THE ANTI-B, No. 1 SERUM UPON IT.s CoRRESPONDING FILTRATE AND UPON THE FILTRATES OF B. TYPEIOSUS AND B. COLI. B. No. 1 filtrate: Anti-B, No. 1 se rum : 0.5 c.c. 0.25 c.c. At 9% hr., slight cloudiness; at 4 hrs., fine 0.5 0.10 precipitation ; at 10 hrs., copious precipiº tation in both tubes, fluid being clear and sterile. B. typh. filtrate: tº §§ C. C. }; At 94 hrs., slight cloudiness; at 4 hrs., not as marked as in homologous filtrate; at 10 hrs., the amount of precipitum is sim- ilar to above set. B, coli filtrate . & a 0.5 c.c 0.25 No reaction at 4 hrs. ; very slight reaction in 0.5 0.10 both tubes at 24 hrs. ; at 48 hrs., the tube with the larger amount of serum contains a copious precipitum. ANTI-B. Ty PHOID SERUM. Rabbit VIII received, from April 10 to June 12, 1902, nine twenty-four- hour old agar cultures; bled June 17. Agglutinating value between 1–5,000 and 1–10,000. No agglutinating reaction upon B. No. 1 (1–10, complete in four hours; 1–100, negative). - - 480 - CHARLES NORRIS TABLE VII. THE PRECIPITATING ACTION OF THE ANTITYPHOID SERUM UPON ITS CORRESPONDING FILTRATE AND UPON THE FILTRATES OF B. No. 1 AND B. COLI, Typhoid filtrate: Antity ph. serum : 0.5 c.c. 0.25 c.c. Cloudiness in several minutes in both 0.5 0.10 tubes; at 1 hr., small flocculi; at 4 hrs., copious precipitum, fluid clear; at 29 hrs., amount of precipitum similar in both tubes. B. No. 1 filtrate: - 0 5 c.c. 0.25 At 1 hr., opalescence and cloudiness; at 4 0.5 0.10 hrs., precipitation not quite as marked as with the homologous filtrate; at 29 hrs., .# the two sets of tubes are similar. B. coli filtrate: 0.5 c.c. 0.5 Negative at 4 hrs. ; at 19 hrs., large flocculi in the 0.25 c.c. serum tube. In the 0.10 c.c. tube scanty precipitum at 48 hrs. The reaction was slower and much less copious than with the other filtrates. § ; § The tests with fresh anticoli serum could not be made owing to the death of the rabbits during the course of adaptation. The tests made with the antisera of rabbits of Series III con- firm the previous observations of Series I and II in regard to the precipitating action of the serum of a rabbit adapted to one species, upon the filtrates of two other species of the colon-typhoid group. - RECORD OF FURTHER TESTs MADE witH THE ANTISERA of SERIES II AND III, WITH THE SAME AND WITH ADDITIONAL FILTRATES. The precipitating action of the antisera of the rabbits of Series II and III adapted to the three species of the colon group upon filtrates of other species of this same group which have not as yet been referred to, will now be given, for brevity, in tabulated form, the sera of Series II having been kept two months; that of Series III, several weeks. SUMMARY. Table VIII indicates that a highly active antiserum, “anti-B. No. 1,” induces precipitation in a large number of heterologous filtrates (even after preservation for two months in cold storage). On the other hand, an antiserum, “antityphoid,” originally much less active than the above anti-B. No. 1 serum, as indicated by its low agglutination value, 1–1,000, possesses only a slight pre- cipitating action upon its corresponding and none upon heterolo- gous filtrates. THE BACTERIAL PRECIPITINS 481 TAHLE VIII. THE PRECIPITATING ACTION OF ANTI-B. No. 1 AND ANTI-B. TYPHosUs SERA UPON THEIR CORRESPONDING FILTRATES AND UPON THE FILTRATES OF HETEROLOGOUS SPECIES OF THE COLON-TYPHOID GROUP. (Test made June 18, 1902.) Filtrate Anti-B. No. 1 Serum. Rabbit W, Series II. Agglutinating Mºe, 1–20,000. Bled April 8, Antityphoid Serum. Rabbit IV, Series II. Agglutinat- ing Value, 1–1,000. Bled, April 2, 1902 0.5 c.c. B, coli: 1/3–4/10/1902. B. typhosus: 1/3–4/10/1902. B. No. 1: 1/3–4/10/1902. B. Schottmüller:* Stamm Seemann: 4/11– 6/6/1902. Stamm Müller: 4/11– 6/6/1902. B. Gwyn: 4/11–6/4/1902. B. Cushing: 4/11–6/4/1902. 0.25 c.c. .* Negative at 48 hrs. Cloudiness at 4 hrs. ; slight precipitum at 18 hrs. ; more marked at 48 hrs. Cloudiness at 1 hr. ; slight pre- cipitum at 4 hrs. ; abundant precipitum at 18 hrs. Cloudiness at 1 hr.; well-marked precipitum at 4 hrs. Cloudiness at 4 hrs. ; slight pre- cipitum at 48 hrs. Cloudiness at 1 hr.; well-marked precipitum at 4 hrs. Opalescence at 4 hrs. ; distinct precipitum at 18 hrs. 0.25 c.c. - Negative at 48 hrs. Negative at 18 hrs. ; slight precipitum at 48 hrs. Slight precipitum at 48 hrs. Negative at 48 hrs. Negative at 48 hrs. Negative at 48 hrs. Negative at 48 hrs. TABLE IX. THE PRECIPITATING ACTION OF ANTI-B. No. 1, ANTI-B. TYPEIOSUS, AND ANTI-B. COLI SERA UPON THEIR CORRESPONDING FILTRATES AND UPON THE FILTRATES OF HETEROLO- GOUs SPECIES OF THE COLON-TYPHOID GROUP. (Test made June 21, 1902.) Anti-B. No. 1 Serum. Rab- bit VIII, Series III. Ag- Antityphoid Serum. Rabbit VII, Series bit Anticoli Serum. Rab- WI, Series II. Filtrate g * –5– III. Agglutinating No Agglutinating fºº #: Vºiº | vºlugºi 3. % Bied June 17, 1903 || 7, 1903 0.5 c.c. 0.1 c.c. 0.1 c.c. 0.1 c.c B. Schottmüller: & g Seemann. At 4% hr., cloudiness; at At 9% hr., cloudiness; Slight precipitum at 19 hrs., copious pre- at 19 hrs., copious 48 hrs. cipitum, & precipitum. Müller. Nº. at 4 hrs. ; slight | Negative at 4 hrs. ; Negative at 4 hrs. ; cloudiness and pre- slight precipitum slight precipitum cipitum at 19 hrs. at 19 hrs. at 19 hrs. B. Gwyn. Slight precipitum at 19 Siº precipitum at | Negative at 48 hrs. 1I’S. TS. B. Cushing. Cloudiness at 4 hr. ; Cloudiness at 4 hr. ; Small amount of pre- B. psittacosis. B. typhimurium. B. icteroides. B. enteritidis. marked precipitum at 4 hrs. ; copious precipi- tum at 18 hrs. Cloudiness at 9% hr.; copious precipitum at 4 hrs ; most marked reaction of a '1, Same as above. Negative at 4 hrs. ; small amount of precipitum at 19 hrs. Cloudiness at 9% hr. ; moderate amount of precipitum at 4 hrs. marked precipitum at 4 hrs. ; copious precipitum at 18 hrs. Cloudiness at % hr. ; copious precipi- tum at 4 hrs. Same as above. Negative at 4 hrs. ; at 19 hrs., small amount of precipi- tum. Moderate amount of precipitum at 19 hrs. cipitum at 24 hrs. Same as above. Same as above. Negative at 48 hrs. Slight amount of pre- cipitum at 24 hrs. *The filtrates of Schottmüller, Gwyn, and Cushing were obtained from meat-infusion broth. The writer is personally indebted to Dr. Schottmüller for his cultures. 482 CHARLES NoFRIs The filtrates psittacosis, typhi murium, icteroides, and enteritidis, were alkaline to litmus paper, and were obtained from Liebig’s meat-extract peptone-salt broth, grown for three months at 37° C. and then passed through Berkefeld filters. SUMMARY. Table IX demonstrates the fact again that highly active anti- sera adapted to one species precipitate the heterologous filtrates of various species belonging to the same group. The anticolon serum which had been kept for seventy-five days produced no reaction, or only slight precipitates, in the heterologous filtrates. It must be borne in mind that, even when fresh, this anticolon serum did not produce copious precipitates in its corresponding filtrate. The Müller, Gwyn, and icteroides filtrates gave the scantiest reactions. The most copious precipitates were furnished by the Seemann, psittacosis, and typhi murium filtrates, which reacted with the antityphoid and anti-B. No. 1 sera in serum dilutions of 1–6 as copiously as did the antisera in their corresponding filtrates. The onset of reaction, cloudiness, is practically instan- taneous in the corresponding filtrates, whereas in the heterolo- gous filtrates it is delayed for fifteen to twenty minutes or IſlCI'ê. A STUDY OF THE PRECIPITATING VALUE OF SERA ADAPTED TO SPE- CIES OF THE COLON-TYPHOID GROUP FOR THEIR HOMOLOGOUS AND HETEROLOGOUS CULTURE-BROTH FILTRATES. A DETER- MINATION OF THE MAXIMAL LIMITS OF PRECIPITATION OF ANTI- SERA FOR THEIR CORRESPONDING AND HETEROLOGOUS FIL- TRATES. In order to determine the limits of the precipitating action of antisera for their corresponding and heterologous filtrates, the following series of tests were made, at first in serum dilutions of 1–50 (5 c.c. of filtrate and 0.1 c.c. of serum), November 8, 1902: THE BACTERIAL PRECIPITINs 483 TABLE X. THE PRECIPITIN REACTIONS OBTAINED wrTH ANTI-B. No. 1, AND ANTITYPHOID SERUM DILUTIONS OF 1–50, IN THEIR CORRESPONDING FILTRATES AND IN THE FILTRATES OF HETEROLOGOUS SPECIES OF THE COLON-TYPHOID GROUP. Anti-B. No. 1 Serum. Aggluti- Antityphoid Serum. Agglu- Filtrate nating Value, 1–10,000. Bled tinating Value, 1–20,000. Oct. 24 Bled Oct. 24. 5 c.c. g 0.1 c.c. 0.1 c.c. B. typh. Negative up to 24 hrs. At 9% hr., slight cloudiness; at 2% hrs., much cloudi- ness; at 22 hrs., abundant s - precipitate. B. No. 1. Slight cloudiness at 9% hr. ; at | Negative up to 4 and 24 hrs. 2% hrs., much cloudiness; at º 22 hrs., heavy precipitate. B. coli. Negative up to 22 hrs. and later. Negative up to 22 hrs. ; at 48 B. typh., Pfeiffer. B. dysenteriae : “Flexner.” “Kruse.” “Seward.” B. Cushing. B. Gwyn. B. Schottmüller: “Seemann.” “Müller.” B. psittacosis. B. typhi murium. B. enteritidis. B. icteroides. B. suipestis. Negative at 4 and 24 hrs. Negative up to 4 and 24 hrs. Negative up to 4 and 24 hrs. Negative up to 4 and 24 hrs. Negative at 2% hrs. ; at 22 hrs., moderate amount of precipi- tate. Negative HP to 22 hrs. ; at 48 º: small amount of precipi- 3.062, Negative at 2 hrs. ; , s m all amount precipitum at 22 hrs. ; at 48 hrs., considerable pre- cipitum. Nºſe at 22 hrs. ; perhaps slight cloudiness and small amount of precipitate at 48 TS. Slight cloudiness at 9% hr., which has in creased in amount at 2% hrs. ; at 22 hrs., a copious precipitate. At 2% hrs., very slight cloudi: ness; at 22 hrs., distinct and moderate amount of precipi- tate; more at 48 hrs. - Negative at 48 hrs, and later. Negative at 22 hrs. ; slight amount of precipitate at 48 hrs. Negative at 48 hrs, and later. hrs., a faint cloudiness. Identical reaction , to that given by the typhoid fil- trate above. At 22 hrs., slight cloudiness. At 22 hrs., slight cloudiness. At 22 hrs., slight gloudiness. No precipitate develops. Negative at 2% hrs., at 48 hrs., and later. Negative up to 4 and 24 hrs.” Negative at 22 hrs, and later. Negative at 48 hrs. At 2% hrs., slight cloudiness; at 22 hrs., cloudiness and small amount of precipi- tate, which at 48 hrs. has increased. Negative at 48 hrs. and later. Slight cloudiness at 22 hrs. ; no precipitate develops. Ngative at 22 hrs. ; at 48 hrs. Negative at 48 hrs. *Addition of 0.1 c.c. of both sera together gave copious precipitate at twenty-two hours. All filtrates strongly alkaline to litmus paper. SUMMARY. The table indicates that the sera of rabbits adapted to one species of the colon-typhoid group precipitates in dilutions of 1–50, the filtrate of many, but not all, of the species in this group. 484 CHARLES NORRIS DETERMINATION OF THE MAXIMAL PRECIPITATION LIMITS OF THE ANTI-B. No. 1 SERUM FOR ITs HomoLOGOUs BACILLUs. In order to determine the precipitation limits of the anti-B. No. 1 serum more accurately than was indicated by the above table, the following tests were made: To each tube was added 0.9 c.c. of filtrate and 0.1 c.c. of serum, or its dilu- tions in sterile 0.85 per cent. saline solution. TABLE XI. ANTI-B, No. 1 SERUM ; AGGLUTINATION LIMIT, 1–10,000, B, No. 1 filtrate: B. No. 1 serum : * 9/10 c.c. 1/20 c.c. Cloudiness marked at 1 hr. 9/10 1/40 Faint cloudiness at 1 hr. 9/10 1/60 Faint cloudiness at 1 hr. 9/10 1/80 Faint cloudiness at 1 hr. - 9/10 1/100 Negative at 1 hr. ; negative at 48 hrs. 9/10 1/120 Negative at 1 hr. 9/10 1/150 Negative at 1 hr. At 20 hrs., 1/100 c.c., a very faint reaction; 1/150 c.c., reaction is possibly present; at 48 hrs., a distinct small flocculus had formed. Reaction thus occurred at 1–1,500. The limit of the precipitating action of the anti-B. No. 1 serum for the homologous filtrate was thus shown to be 1–150. By a similar method the limits of the action of the serum upon the following filtrates of broth cultures of approximately similar lengths of incubation were found:* B. psittacosis, 1–150. . B. typhi murium, 1–120. B. Schottmüller (Seemann), and B. Cushing, 1–80. B. Gwyn, 1–40. - B. enteritidis, 1–50. - The serum did not precipitate in a dilution of 1–50, the filtrates B. typho- sus, B. coli, B. dysenteriae, “Flexner,” “Rruse,” and “New Haven” (non- mannite fractors) or B. icteroides. The B. Schottmüller (Müller) gave a ques- tionable reaction with the above serum dilution. DETERMINATION OF THE MAXIMAL LIMITS OF THE ANTI-B. TYPHOID SERUM FOR ITS HomoLOGOUS BACILLUS. - ANTITYPHOID SERUM. Agglutination limit, 1–20,000. The serum was tested as was the anti-B. No. 1 serum, as above described. The precipitation limit for the homologous typhoid filtrate was found to be 1–150; for the psittacosis, 1–140; the filtrates of three dysentery races and *Two factors, it may here be stated, were always taken into account in determining the precipitating value of an antibacterial serum: the degree of adaptation of the animal to the bacterial species, and the age and abundance of the growth of the broth culture from which the filtrate has been obtained. In the series of tests which were made to determine the limits of the precipitating action of the antisera upon their corresponding and heter- ologous filtrates, the various filtrates were accordingly obtained from broth cultures of approximately the same age. THE BACTERIAL PRECIPITINS 485 B. enteritidis, 1–50 (slight cloudiness), no reaction in the other filtrates being obtained with a serum dilution of 1–50. THE PRECIPITIN REACTIONS OBTAINED witH A FRESH ANTICOLI SERUM. - ANTICOLI SERUM. Agglutination limit, 1–250. Owing to unavoidable circumstances, no anticoli rabbit serum was at hand at the time the above tests were made. A rabbit which had received eight and a half agar cultures during a period of fifty-one days, however, furnished a serum of high precipitative powers. The serum agglutinated * the homologous colon bacillus completely in a dilution of 1–250 in seven hours. Another colon bacillus of the same type (non-saccharose splitter) agglutinated only in a dilution of 1–100. The B. typhosus “Coll.” and the intermediates were not agglutinated in dilutions above 1–10 and 1–50, respectively. DETERMINATION OF THE MAXIMAL PRECIPITATION LIMITS OF THE ANTICOLI - SERUM FOR ITs HomoLOGOUs FILTRATE. The following tests were made to determine the precipitating limit, that is, the maximal dilution, at which a reaction is obtained by the anticoli serum in its homologous filtrate: Each tube contained 5 c.c. of filtrate, and to each was added 1/10 c.c. of the diluted serum. TABLE XII. Filtrate Anticoli 2 HTS 6 Hrs 24 Hrs 72 Hrs Serum e * . e t; 5 c.c. 1/25 c.c. | Negative. Very Clear fluid with Fairly abundant pre- slight. slight precipitate, cipitate. 5 1/50 Negative. Negative. Negative. Clear, sterile, fluid; slight precipitate. 5 1/100 Negative. Negative. Negative. Clear fluid; slight, but typical precipitate. 5 1/250 Negative, Negative. Negative. Negative. The maximal dilution at which a reaction is obtained was thus found to be about 1–500. THE PRECIPITIN REACTIONS OBTAINED witH ANTICOLI SERUM DILUTIONS OF 1–50 IN ITS CoRRESPONDING FILTRATE AND IN THE FILTRATES OF HETER- OLOGOUs SPECIES of THE COLON-TYPHOID GROUP. The anticoli serum was next tested upon the following filtrates: Each tube contained 5 c.c. of filtrate and 1/10 c.c. of the serum. Tests were made at 37° C. . *The low agglutinative value of the serum for the B. coli the writer does not believe can be ascribed to the recent isolation of the bacillus from the cadaver, for it had been cul- tivated for a year, about twelve generations; and, furthermore, a stock colon bacillus was likewise agglutinated by the serum only in low dilutions. It is a well-known fact that freshly isolated cultures which themselves are more or less inagglutinable do give rise to abundant production of immune bodies in animals. The inagglutinability of typhoid and pyocyaneus cultures of recent isolation has been described in EISENBERG's interesting article, to which the reader is referred. 16 486 CHARLES NORRIS TABLE XIII, Filtrate 5 c.c. 3 Hrs. 6 Hrs. 8% Hrs. 12 Hrs. 24 Hrs. B. Schottmüller: & e & º Stamm Müller | Negative. Negative. Negative. Negative. Negative. Stamm Seemann Negative. Negative, Negative, Negative. Negative. B. enteritidis Very slight | . . . . . . Slight. Slight. Slight cloudiness; cloudiness. no precipitate. B. psittacosis Negative. Negative. Negative. Slight. | Considerable pre- cipitum ; fluid clear. B. icteroides Very slight | . . . . . . Slight. Slight. Considerable pre- .* cloudiness. cipitum ; fluid clear. B. typhi murium | Negative. Negative. Negative. Negative. Slight precipitum B. typhi Negative. Negative, Slight. Slight. Slight precipitum B. No. 1 Negative, Negative. Negative. Negative. Slight precipitum B. dysenteriae: * ge e tº ſº * Flexner’” Very slight | . . . . . . . . . . . . . Very slight Slight precipi- cloudiness? cloudiness. tum ; fluid r nearly clear: ** Kruse ’’ Very slight | . . . . . . Slight Cloudiness. Slight #. cloudiness. g quite marked ... tate; fluid clear. “Flexner’’ Very slight | . . . . . . Slight Slight. Considerable pre- (New Haven) cloudiness. cipitate. B. Gwyn Negative. Negative. . . . . . . . Negative, Negative. . B. Cushing Negative. Negative. . . . . . . . Negative. Considerable pre- cipitate; fluid clear. B. proteus” Negative. Negative. . . . . . . . Negative. Negative. . Sp. Metchnikovi | Negative. Negative. . . . . . . . Negative. Negative. Sp. cholerae Negative. Negative. i ...... Negative. Negative; pre- cipitate? Staphylococcus Negative. Negative. ...... Negative. Negative. - pyOgeneS aureuS - B. coli (homologous) clouding within fifteen minutes; at six hours, slight precipitation and marked clouding. - + The filtrates of the three dysentery races have nearly as copious precipita at twenty-four hours as the colon. increased in amount, in B. No. 1, Seemann, typhoid, icteroides, psittacosis (copious), typhi murium, enteritidis, and Müller (very slight); Cushing, quite copious; Gwyn, less so; staphylococcus, Sp. Metchnikovi, and B. proteus, filtrates absolutely negative. . • The anticoli serum, we have already shown, gives a reaction in the colon filtrate in a dilution of 1–500. In a dilution of 1–50 the serum precipitates the filtrates of B. enteritidis, psittacosis, icteroides, typhi murium, B. typho- sus, B. No. 1, B. dysenteriae, Flexner, Kruse, and New Haven (non-mannite fractors), and B. Cushing, after twenty-four hours at 37° C., B. Gwyn giving only a slight reaction. The filtrates of other bacterial groups, such as B. proteus, B. prodigiosus, the vibrios, cholerae and Metchnikovi, and the staphylococcus, are not acted upon. * Referred to later on. At seventy-two hours precipita have THE BACTERIAL PRECIPITINS 487 THE PRECIPITIN REACTIONS OBTAINED witH ANTICOLI SERUM DILUTIONs OF 1–10 IN ITS CoRRESPONDING FILTRATE AND IN THE FILTRATES OF HETEROLOGOUS SPECIES OF THE COLON-TYPHOID GROUP. Each tube contains 1 c.c. of filtrate and 1/10 c.c. of anticoli serum. TABLE XIV. 15–20 Minutes 1%. Hour 18 Hours. 72 Hours. Schottmüller: Seemann Negative. Cloudiness. Like typhoid, but a little less. B. enteritidis Cloudiness. Marked cloudi- | Same as above. IlêSS, B. typhi murium| Cloudiness. Marked cloudi- Hº, precipitum, like B. IlêSS. O. l. B. Gwyn Cloudiness. Cloudiness. Less than enteritidis. B. Cushing Cloudiness. Cloudiness. Less than enteritidis. B. pºcosis Cloudiness. Cloudiness. About like B. No. 1. B. No. 1 Marked cloudi- | Less than colon. Like colon tube, only slightly Il eSS. § in buil. Next to colon U1 Oe. B. § (collec- Marked cloudi- Less than colon. Like B. No. 1, perhaps less. IOIl In CSS, B. typhi Pfeiffer Cloudiness. Marked cloudi- Lº than other typhoid fil- In eSS, Tall;62 . -" B. coli Most marked | . . . . . . . . . . Marked precipitum on sides, cloudiness of which settles out on agi- * * all... tating tube. B. icteroides Negative. Negative. Very small amount; increased at 72 hrs. B. dysenteriae : “Flexner’’ Slight cloudi- || Marked cloudi-| Flocculent precipita. IleSS. ness, but less than B. No. 1 and B. typh. tº (col.). “Kruse” Slight cloudi- | Flocculi. DeSS. “New Haven” | Slight cloudi- | Flocculi. IleSS, sº Metchni- Negative. Negative. Negative. Negative. OWI Sp. cholerae Negative. Negative. Negative. Negative. Streptococcus Negative. Negative. Negative, Negative. pyOgeneS aureus & B. prodigiosus Negative. Negative. Negative. Negative. With one to ten anticoli serum dilutions, after eighteen hours, little difference in the amount of precipitum formed is made out between the colon and typhoid, B. No. 1, B. enteritidis, B. typhi murium, and B. psittacosis filtrates. The onset of the reaction with colon filtrate is, however, immediate, whereas in the typhoid and B. No. 1 filtrates the cloudiness reaction appears after fifteen minutes. At seventy-two hours all tubes containing filtrates of the colon-typhoid group contain a considerable amount of pre- cipitum. The icteroides filtrate contains the least; but this is accounted for by the fact that the growth obtained in the broth culture which furnished the filtrate was considerably less than 488 CHARLES NORRIS those of the other bacilli, with the possible exception of the B. dysenteriae “Flexner” filtrate, in which the amount of precipitum was also small.” SUMMARY OR TABULATION OF THE MAXIMAL LIMITS OF PRECIPITATION OBTAINED witH THE THREE FRESH ANTISERA. The maximal limits of precipitation of the three antisera in the various filtrates, so far as determined, are represented in the following table: TABLE XV. Limit of Pre- cipitation October 24: * B. No. 1 filtrate............................. Anti-B. No. 1 serum — aggluti- 1–150 nation limit, 1–10,000 1–150 . B. psittacosis filtrate....................... { { 1–120–150 B. typhimurium filtrate . . . . . . . . . . . . . . . . . . . { { 1- 80 B. Schottmüller, “Seemann "filtrate...... (, , 1- 80–100 B. Cushing filtrate . . . . . . . . . . . . . . . . . . . . . . . . . { { 1- 40 B. Gwyn filtrate ............................ { % 1— 50 B. enteritidis filtrate....... . . . . . . . . . . . . . . . . { % October 24: - B. typh, filtrate.................. . . . . . . . . . . . Antityphoid serum — aggluti- 1–150 - nation limit, 1–20,000 1–140 B. psittacosis filtrate. . . . . . . . . . . . . . . . . . . . . . . . . { { B. dysenteriae, Flexner, Kruse and New 1- 50 Haven filtrate . . . . . . . . . . . . . . . . . . . . . . . . . { % - - November 25 : B. colon filtrate ............................ Anti-colon serum —agglutina- tion limit, 1–250 1–500 B. enteritidis filtrate. . . . . . . . . . . . . . . . . . . . . . . {{ 1– 50 B. psittacosis filtrate. ...................... { { 1- 50 B. icteroides filtrate... . . . . . . . . . . . . . . . . . . . . . . { { 1– 50 B. typhi murium filtrate ................... { { 1– 50 B. typhoid filtrate...... . . . . . . . . . . . . . . . . . . . . { { 1– 50 B. No. 1 filtrate ............................. { % 1– 50 B. Cushing filtrate ......................... { { 1– 50 B. dysenteriae, three races, filtrate. . . . . . . . - { % 1– 50 IV. THE PRECIPITATING ACTION OF THE SERA OF RABBITS ADAPTED TO VARIOUS RACES OF THE B. DYSENTERIAE SHIGA, AND ALSO TO OTHER SPECIES OF THE SO-CALLED DYSENTERY GROUP, WHICH FERMENT MANNITE, IN THEIR RESPECTIVE CULTURE FILTRATES. - The tests made with antidysenteric rabbit sera are now to be described, for although the B. dysenteriae doubtless belongs to the colon-typhoid group, it presents many points of contrast, as regards its fermentative and pathogenic characters, to the species *The results obtained with this anticoli serum are in marked contrast to those recorded by KRAUS in the Wiener klimische Wochenschrift (1901, 14, p. 693) which emphasize the specificity of the precipitating action of the serum of animals adapted to various races of the colon bacillus. THE BACTERIAL PRECIPITINS 489 heretofore mentioned; namely, B. typhosus, B. coli, and the dex- trose splitters of the intermediate group. The three races of B. dysenteriae employed for the production of antisera and filtrates were a B. dysenteriae (known as Flexner), the B. dysenteriae, Rruse, and the B. dysenteriae, New Haven (Duval). Mannite is not fermented by these races, which accordingly belong to the Shiga type of the B. dysenteriae. Unfortunately, only a few experiments could be made with those species or races of bacilli which have been mistaken for the B. dysenteriae Shiga; namely, those species which ferment mannite and monosaccharids, and those which, unlike the former race, ferment in addition maltose, dextrin, and saccharose. Both these races, unlike the B. dysen- teriae Shiga, have only slight pathogenicity for rabbits.” Difficulty was experienced in immunizing rabbits to the three races of the B. dysenteriae Shiga in our possession, living cultures being pathogenic, whereas boiled cultures in our hands conferred only slight protection, and the serum of rabbits adapted to boiled cultures yielded only traces of agglutinating and precipitating substances. A precipitating serum was, however, obtained finally in a rabbit with repeated doses of living cultures of B. dysenteriae “New Haven.” ANTIDYSENTERY RABBIT SERUM. Rabbit received about ten agar cultures of B. dysenteriae “New Haven," between June 4 and August 12, 1902. Serum collected August 19. Agglutinating value: 1–1,000. 1–10: distinct agglutination at 94 hr.; complete at 4 hrs. * 1–100: distinct agglutination at 34 hr.; at 1% hrs., agglutination farther advanced than the dilution tube 1–10; complete at 4 hrs. . . . 1–500; at 134 hrs., indistinct clumps; at 4 hrs., settling, heavy deposit at bottom; complete at 8 hrs. 1–1,000: slight settling at 1% hrs.; at 4 hrs., like 1–500; complete at 8 hrs. No reaction in higher dilutions at 8 hours. In the hanging drop with a dilution of 1–100, Small clumps are formed at seven hours. With a dilution of 1–200, small clumps developed at one-half hour, which did not increase in size. The serum agglutinated the B. dysen- teriae “Kruse” in a dilution of 1–500. The filtrate of the B. dysenteriae “New Haven” was obtained from a broth culture of beef extract, grown for one hundred and two days at 37° C. *The reader is referred to the interesting article of HISS upon the differential fermen- tative peculiarities of the various species which have been thrown temporarily into the so- called dysentery group; 17 also LENTZ.18 490 CHARLEs NorBIs TABLE XVI. THE PRECIPITATING ACTION OF THE ANTIDYSENTERIC SERUM UPON ITS HOMOLOGOUS FILTRATES. Filtrate Antidysen. Serum ** New Haven” “New Haven’’ 1 c.c. 1 c.c. Immediate clouding, and precipitation in a few min- utes; complete at 4 hrs. 1 1/4 Immediate reaction. 1 1/10 Cloudiness developed after 15 min., at 37° C.; at 4 hrs., small flocculi on sides of tube. 5 1/10 No reaction at 10 hrs. ; at 22 hrs., a flocculent precipi- tum. The serum thus agglutinated in a 1–1,000 dilution and precipitated in a dilution of 1–50, which represents practically its maximal limit of precipita- tion.* TABLE XVII. - THE PRECIPITATING ACTION OF THE ANTIDYSENTERIC SERUM UPON Its HETEROLOGOUS FILTRATES. - 1 c.c. colon filtrate + 1 c.c. dysenteric serum: Immediate cloudiness; pre- . cipitum at 1% hrs. 1 “ . “ {{ + 1/4 “ 66 6% Immediate cloudiness; pre- cipitum at 1% hrs. 1/10 “ $6 {{ No instant reaction; faint cloudiness at 1% hrs.; - precipitum at 4 hrs. 1/10 “ 66 {{ Negative after 72 hrs. 1 {{ § { . 66 5 1 “ B. No. 1 “ 1/4 “ 66 $6 Immediate cloudiness; at 4 hrs., abundant precipi- tum. 1. “ 6% 66 + 1/10 “ 6 & 66 Immediate cloudiness; at 4 hrs., cloudiness and . slight precipitum. 5 “ {& 6 & + 1/10 “ 66 46 Negative at 72 hrs. The typhoid filtrate reacted like the B. No. 1 filtrate. A prodigiosus filtrate was tested, with negative results. SUMMARY. The reactions above noted are not quantitatively as marked as that obtained with the homologous filtrate. The antidysenteric serum in a dilution of 1–50 failed to precipitate the filtrates of colon, * Through the kindness of Professor Flexner, we were able to test the precipitating action of the serum of a horse which had undergone adaptation for four months with the dysentery bacillus—the culture “New Haven” isolated from a case of dysentery in an epi- demic at New Haven. - Filtrate of B. dysenteriae “Flexner,” 1/2 c.c. 1/10 and 1/4 c.c. horse serum. Filtrate of B. dysenteriae “ Kruse,” 1/2 c.c. 1/10 and 1/4 c.c. horse serum. Filtrate of B. dysenteriae “New Haven,” 1/2 c.c. 1/10 and 1/4 c.c. horse serum. Tubes kept at 37° C. No reaction at six hours, but after twenty-four hours a small amount of flocculent pre- cipitate had developed. The serum added in the same proportions did not precipitate the filtrates of any of the colon-typhoid group in our possession. The low agglutinating value, 1–200, by microscopic test, of the horse serum, reasonably accounts for the failure to obtain precipitation in any other than the dysenteric filtrates, especially as it has been noted that the typhoid, colon, and our intermediate B. No. 1 antisera precipitated dysenteric filtrates. THE BACTERIAL PRECIPITINS 491 typhoid, and B. No. 1. Thus a reaction was obtained with a serum dilution of 1–10, but not in 1–50, in these heterologous filtrates. The B. dysenteriae “New Haven” having been shown to pos- sess precipitative relationships for the B. typhosus, B. colon, and the intermediate B. No. 1, in the tests to be described the pre- cipitative relationships between several species or races of the dysentery group were studied. THE AGGLUTINATING AND PRECIPITATING ACTION OF THE SERA OF RABBITS ADAPTED TO THREE SPECIES OF THE DYSENTERY GROUP UPON THEIR CORRESPONDING AND UPON HETEROLOGOUS FILTRATES. The three species of the dysentery group are classified (Hiss) according to their fermentation reactions in the following manner: (1) The B. Shiga ferments dextrose and other monosaccharids; but not mannit, maltose, dextrin, or saccharose. (2) The B. “Y” (Hiss) ferments monosaccharids and mannit, but not maltose, dextrin or saccharose. (3) The B. “Baltimore” (same type as Flexner’s “Manila” and “Harris”) ferments monosaccharids, mannit, maltose, dextrin, and saccharose. Two dysentery filtrates were employed—a Kruse filtrate (90 days at 37°C.) and a B. “Y” (Hiss) filtrate (184 days at 37° C.). Through the kindness of Dr. Hiss, three antisera were placed at the writer's disposal—a “Shiga,” a B. “Y” (Hiss), and a “Balti- more” antiserum. The “Shiga” antiserum and the “Kruse” filtrate may be considered homologous, since the B. Shiga and the B. Kruse are non-mannit factors and have similar aggluti- nating limits. For the Baltimore antiserum no homologous filtrate was at hand. The following agglutination tests (Table XVIII) were made by Prof. Hiss with the three antisera of rabbits and also with an anti- typhoid serum” (Macroscopic method, readings at twenty-four hours at 37° C). SUMMARY. Agglutination relationships were thus found to exist between the mannit splitters, the Baltimore and “Y” cultures. Since the normal agglutinins of rabbit serum occasionally reach 1–100 for each of these species, an agglutination reaction below 1–100 has no significance. *These agglutination results were kindly placed at the writer's disposal for the sake of completeness of the experiments, although they are to appear later in a Il article by Prof. Hiss on agglutination in the “dysentery” group. 492 CHARLES NORRIS TABLE XVIII. AGGLUTINATION TEST.S. (Tests made June 23, 1903.) Shiga serum, bled June 17: Baltimore serum, bled May 20: “Y” serum, bled April 30: Typhoid serum, bled April and May: Shiga, 1–1,600–3,200. Baltimore, 1–400. “Y,” 100. Only a little above nor- Typhoid, 100. mal rabbit serum. Shiga, 1–100. Baltimore, 1–3,200. “Y,” + 400. Typhoid, -- 100. Shiga, -100. Baltimore, --1,600. “Y,” 6,400. Typhoid, +100. Shiga, –100. Baltimore, —-100–200. “Y,” +100. Typhoid, +6,000. PRECIPITATION TEST.S. The following precipitin reactions were made with the four antisera and the three filtrates: TABLE XIX. Fil. Kruse, 5.0 c.c. 0.1 c.c., “Y” serum. “ “ 5.0 “ 0.1 “ Shiga “ * { § { 5.0 & W. 0.1 yp g { % { { “ 5.0 “ 0.1 ** Balt. ** { { { { 0.9 { % 0.1 { { { { Y 3 * { { $ (, (, , 0.9 { { 0.1 { % Shiga { { { { { { 0.9 { { 0.1 { { typh. § { { % “ 0.9 “ 0.1 ** Balt. “ Fil. “Y,” 5.0 c.c. 0.1 “ “Y” “ { % “ 5.0 “ 0.1 “ Shiga { % $ $ § { 5.0 { { 0.1 ( { typh. C & “ “ 5.0 “ 0.1 “ Balt. “ { % { { 0.9 { { 0.1 { % { { Y } } { % { { “ 0.9 “ 0.1 “ Shiga { { tº * ( [. 0.9 { { 0.1 { % ty h. { { $ i. { { 0.9 § { 0.1 { % Ba t. { { Fil.typh., Pfeifferjº c.c. 0.1 : “Y” “ § { $ { { 5.0 § { 0.1 { { Shiga $ tº { % { { { { 5,0 { { 0.1 { { typh. { { { { { { { % 5.0 § { 0.1 { { Balt. { \ (, i. & 9 { { 0.9 € $ 0.1 § { * { Y } } { { { { { { { { 0.9 { { 0.1 { {, Shiga { { { % { { { { 0.9 { { 0.1 { { typh. ! { { { { { * { 0.9 { { 0.1 $ $ Balt. { { Cloudy at 6 hrs. ; same at 24 hrs. Cloudy at 6 hrs. ; same at 24 hrs. Negative at 24 hrs. and 48 hrs. Negative at 24 hrs, and 48 hrs. Cloudy at 1 hr.; small precipitum at 24 hrs. Clºudy at 40 min. ; small precipitum at 24 TS, - Negative at 24 hrs...and later. Negative at 24 hrs. and later. Clouding in 40 min...; at 6 hrs., cloudy; slight precipitum at 24 hrs. Clouding in 40 min.? at 6 hrs., cloudy; slight precipitum at 24 hrs. Negative at 24 hrs. Negative at 24 hrs. Almost immediate clouding, more at 40 min. ; at 22 hrs., small precipitate. Cloudiness at 40 min. ; at 22 hrs., small pre- cipitate. Negative at 24 hrs. Negative at 24 hrs. Negative at 24 hrs. and later. Negative at 24 hrs. and later. At 6 hrs., cloudiness; at 24 hrs., small pre- cipitate. Negative at 24 hrs. At 45 min., negative; at 6 hrs., negative; at 24 hrs., very small amount e At 45 min., negative; at 6 hrs., negative; at 24 hrs., small amount. At 45 min., negative; at 6 hrs., small pre- cipitate; at 24 hrs., sm a fi 'flocculent deposit. ' Negative at 45 min., at 6 hrs., and at 24 hrs. THE BACTERIAL PRECIPITINS 493 The “Kruse” filtrate is precipitated by the homologous Shiga and “Y” serum, but not by the Baltimore or typhoid serum. (The typhoid serum was of low value.) The “Y” filtrate is precipitated by “Y” and Shiga, but not by the Baltimore and typhoid antisera. The typhoid filtrate is precipitated only by its corresponding serum in serum dilutions of 1–50; in 1–10 dilutions, also by Shiga and possibly by “Y” serum. SUMMARY, Marked precipitative relationships were thus shown to exist between the Shiga and “Y” bacilli; whereas no agglutinative relationships, or only very slight ones, could be demonstrated with the help of the corresponding antisera, which possessed, however, moderately low agglutinating values. V. THE PRECIPITINS DEVELOPED IN RABBITS ADAPTED TO VARIOUS BACTERIAL SPECIES WHICH DO NOT BELONG TO THE COLON_TYPHOID GROUP. This part of our theme was undertaken, first, to determine the absence or presence of adaptation precipitins for species of bac- teria other than those of the colon-typhoid group in the sera of rabbits that had adapted themselves to these species; second, to determine the specificity of the precipitins thus formed for a spe- cies of a given group by testing the antisera upon the filtrates of species belonging to other groups as well as classes of bacteria. We thus endeavored to throw further light upon the extent of the biological relationships which exist among bacterial species of different groups and classes of bacteria, by means of the precipi- tin reaction, in a way similar to that which Nuttall" has so success- fully employed to demonstrate the biological relationships which exist between animals belonging to various orders, classes and groups. We shall now consider the precipitins which were developed by adaptation to various species of bacteria not hitherto or only cursorily recorded in the literature, as well as the ineffectual attempts that were made to obtain precipitins for certain species of bacteria. 494 CHARLES NORRIs I. THE PRECIPITINS DEVELOPED BY ADAPTATION OF RABBITs TO B. PROTEUS. The following tests were made with the serum obtained from a rabbit which had received six inoculations during a period of four weeks, at first with boiled and then with suspensions of liv- ing agar cultures. The agglutinating value of the serum was 1–5,000, complete in two hours; 1–10,000, incomplete at six hours. The precipitating action of the serum was slower and weaker than in the case of the antiprodigiosus and antistaphylo- coccus sera referred to below. One-fourth c.c. and 1/10 c.c. of the antiproteus serum added to 1 c.c. of the corresponding filtrate gave no reaction in five hours. Later a copious precipitum was formed in the tube containing the larger amount of serum and at twenty-four hours in the 1/10 c.c. tube the precipitum was still settling. One drop of serum to 1 c.c. gave no reaction at forty-eight hours at 37°C. The antiproteus serum added to each of the filtrates in our possession likewise gave rise to no reactions, with the exception of the filtrate of the Sp. Metchnikovi, a slight reaction in this case being obtained at the end of twenty-four hours. On the other hand, the proteus filtrate was not precipitated or made cloudy by the antityphoid, coli, and B. No. 1 sera added in the proportion of 1/4 c.c. of serum to 1/2 c.c. of filtrate, or by antistaphylococ- cus or antipyocyaneus serum. With another antiproteus serum, with an agglutinating value of 1–500, taken from a dying rabbit which had received two inocu- lations of proteus emulsion, a similar and confirmatory series of observations was made, the precipitating action of the serum, however, being slight. We may therefore conclude that no precipitating relationships exist between the B. proteus and the bacilli of the colon-typhoid group and the pyogenic cocci, but that in the class of spirillaceae the Sp. Metchnikovi may give reactions. II. THE PRECIPITINS DEVELOPED BY ADAPTATION OF RABBITS TO B. PRODIGIOSU.S. The antiprodigiosus sera yielded copious precipitates, and gave rise to the most rapid reactions which we have obtained with anti- bacterial Sera. THE BACTERIAL PRECIPITINS 495 The first serum was obtained from a rabbit inoculated at intervals, between April 4 and May 13, 1902, six, twenty-four-hour agar cultures being introduced into the peritoneum. The rabbit developed “snuffles” + May 16, and was therefore bled the next day. The serum had an agglutinating value of 1–10,000, complete in less than three hours. At 1–30,000 and 1– 50,000 incomplete reactions were obtained. THE PRECIPITATING ACTION OF ANTIPRODIGIOSUs RABBIT SERUM (No. 1) UPON ITs HomoLOGOUs FILTRATE. One half c.c. and 1/10 c.c. of serum added to 9% c.c. of filtrate gave rise to instant cloudiness, heavy flocculi being formed in one-half hour, the reac- tion being complete in one hour. THE PRECIPITATING ACTION OF WARIOUs ANTIBACTERIAL RABBIT SERA UPON THE PRODIGIOSUs FILTRATE, AND THE ACTION OF ANTIPRODIGIosus RAB- BIT SERUM (No. 1) UPON FILTRATES or VARIOUs HETERologous BAC- TERIAL SPECIES. The prodigiosus filtrate did not react with the various anti- bacterial sera in our possession; namely, anticolon, antityphoid, anti-B. No 1, and antidysentery; and no reaction followed the addition of the antiprodigiosus serum to filtrates of various bacilli, vibrios, and cocci; or this led to such small and inconstant precipi- tates, even when added in equal volumes, that these can be safely disregarded. A second antiprodigiosus rabbit received eight twenty-four-hour agar cultures in the peritoneum, between April 7 and June 7, 1902, the rabbit being bled four days later. The macroscopic agglutination tests were instantaneous up to 1–500; 1–1,000, complete in less than one hour; 1–5,000, complete in less than two hours; 1–15,000, complete in six hours. - THE PRECIPITATING ACTION OF ANTIPRODIGIOSUs SERUM (No. 2) UPON ITS HOMOLOGOUS FULTRATE. One-half c.c. prodigiosus filtrate pſus one drop of serum produced cloudiness and precipitation in two minutes, and at twenty-four hours a copious precipitum had formed. The limit of actual precipitation was obtained with a serum dilution of 1–300, the precipitum formed being very Scanty. * Adaptation with most of the cultures was rendered a tedious task during the win- ter months on account of a prolonged epidemic of “snuffles " in our rabbits. To fore- stall criticism, it may be mentioned here that only a few observations obtained with the sera of afflicted rabbits are recorded, the fact being noted in the text. Such Sera were of value, however, in substantiating the results obtained with the sera of the healthy rabbits. The afflicted rabbits did not yield as potent precipitating and agglutinating sera, but the same relative results were found as with the antisera of healthy animals. 496 - - CHARLES NORRIS THE PRECIPITATING ACTION OF ANTIPRODIGIosUs RABBIT SERUM (No. 2) UPON THE FILTRATES OF WARIOUs ANTIBACTERIAL SPECIES. The second antiprodigiosus serum was tested against various heter- ologous filtrates, each tube containing 3% c.c. of the filtrate and 34 c.c. of serum. The reactions obtained were slow, the twenty-four hour results alone being given. With filtrates of B. typhosus, B. typhi murium, B. No. 1, B. psitta- cosis, and Sp. Metchnikovi, a moderate amount of precipitum was formed, typical flocculi adhering to the sides of the tubes. The dysentery filtrates gave a small amount of precipitate. The filtrates of bovine and human tubercle bacilli, B. icteroides, and Sp. cholerae, were negative. At twenty- four hours all the fluids were clear and sterile, as found by cultural and microscopic tests of the fluid and precipitate. The precipitin of this antiprodigiosus serum thus showing a somewhat marked affinity for the precipitable substances of these bacillary filtrates, one might be led to suppose that a strong pre- cipitating serum would combine with the precipitable substance of any filtrate, whether of coccus or of bacillus, thus leading to the conclusion that there is no specificity in the precipitin reac- tion. This was readily disproved by the following test. The antistaphylococcus serum to be described later acted rapidly upon, and yielded a copious precipitum in, its corresponding filtrates. TABLE XX. Antiprodigiosus serum | Staphylococcus fil- || After 48 hrs., at 37° C., absolutely negative. (No 2) trate * .5 c.c. 0.5 c.c. g Antistaphylococcus se- | Prodigiosus filtrate | After 48 hrs., at 37° C., absolutely negative. I'llm 0.5 0.5 c.c. . O C. C. Here, although both sera were extremely powerful, there is no interaction. In the course of the reactions between the anti- prodigiosus serum and the above filtrates, but little of the pre- cipitating substances of the serum were fixed, since addition of the prodigiosus filtrate to the tubes, when freed of the first pre- cipitum that had formed after forty-eight hours contact at 37°C., gave rise to an immediate reaction. The reaction, although not delayed, was, however, never as copious. No attempt was made to calculate the amount of the precipitating substance of the serum which was thus fixed. This was the only antiprodigiosus serum obtained which precipitated heterologous filtrates of a differ- ent class, with the single exception of that of the filtrate of Sp. Metchnikovi. THE BACTERIAL PRECIPITINS 497 A less actively precipitating serum of a rabbit treated with a sterile and cell-free prodigiosus filtrate likewise precipitated the Sp. Metchnikovi filtrate, a perfectly characteristic precipitate being formed in seventeen hours. The serum had no action on any other of the above filtrates, possibly because of its slighter precipitating action. - Our first antiprodigiosus serum obtained from a rabbit adapted by inoculation with living cultures, although of greater precipitating value for the corresponding filtrate than the serum of the rabbit inoculated with the prodigiosus filtrate, never- theless failed to precipitate the same filtrate of Sp. Metchnikovi. From this fact it seems not unlikely that the B. prodigiosus devel- ops several distinct precipitable substances, depending upon its environment. Thus, the filtrate of old broth cultures that was inoculated may have contained a larger quantity of the precipi- table substance which gives rise to the formation of the special or partial precipitin of the Sp. Metchnikovi, than did the emul- sion of the living bacilli with which the other rabbits had been inoculated. On the other hand, it is conceivable that there exists a difference, both quantitative and qualitative, in the cell recep- tors of different rabbits which give rise to the various partial precipitin haptines.* Precipitative relationships between species of one group of Bacteriaceae and those of different groups of the same family and those of different families were thus found to exist. Since several equally actively precipitating antiprodigiosus Sera did not precipitate the filtrates of the colon-typhoid group, it seems reasonable to assume that the cells of different rabbits develop precipitins which may vary qualitatively. * * Ascoli and von Dungern have noted qualitative differences in the hemo-precipitins developed in the sera of animals of the same and of different species. Von Dungern 2" has obtained almost conclusive evidence that the precipitable substance of the blood plasma of the Major squinado is composed of at least several haptophors which have distinct affini- ties, and that corresponding to the composite nature of this substance several precipitins are developed. The precipitins are thus composed of a number of partial substances, which are distinct from each other and react only with the special precipitable substance which has given rise to them. Joos 21 has described two distinct agglutinins in anti- typhoid serum which, according to him, owe their origin to the presence in the typhoid bacillus of two distinct agglutinable substances. Joos's results seem capable of a differ- ent interpretation, however, from that which he has placed upon them. 498 CHARLES NORRIS III. THE PRECIPITINS DEVELOPED BY ADAPTATION OF RABBITS TO THE SPIRILLUM CHOLERAE AND METCHNIKOWI. In the class of Spirillaceae only two species were investigated. The first series of rabbits treated with cultures of Sp. cholerae asiaticae and Sp. Metchnikovi were unsuccessful, the sera having little precipitating action. The cholera culture was an old stock culture which had become non-pathogenic to guinea pigs and to rabbits. The Sp. Metchnikovi was, however, still pathogenic. Rabbits were finally adapted which yielded precipitating sera. The anticholera rabbit was inoculated with twelve agar cultures (twenty- four hours' growth) in the course of ten weeks. (Agglutination test was impossible on account of extensive pseudo-clumping in broth cultures and precipitation of the emulsions of cholera cultures.) The anti-Metchnikovi rabbit received nine agar cultures grown for twenty-four hours at 37° C., in the course of seven weeks. The serum possessed an agglutinating value of 1–500, complete in four and one-half hours. The cholera serum did not agglu- tinate the Sp. Metchnikovi above the limit of normal rabbit serum (1–10 being nearly complete in four and one-half hours; 1–100 negative at twenty- four hours). - The filtrates of both species were obtained from broth cultures grown at 37° C. for fifty-one days. TABLE XXI. THE PRECIPITATING ACTION OF THE ANTICHOLERA AND ANTI-METCEINIKOVI SERA UPON CHOLERA AND METCHNIKOVI AND TYPHOID FILTRATES. * D .*) C. C. Cholera fil.” .*) C. { s . cholera serum | No reaction up to 6 hrs. ; at 24 hrs., a small “ { { precipitum in all tubes. - 5 . c & { { { % { % c.c. Metch. fil, { % { { { % { { { % { % c.c. Metch. fil. & { { { { e ſº * e 3 , cholera serum | No reaction at 6 hrs. ; at 24 hrs., a slight { { \ { reaction, perhaps less than above homolo- { { * { gous Serles. \ 5 c.c. Metch. serum | Instantaneous clouding in all tubes; at 4 { % { { hrs., reaction complete in all tubes; at 24 š. s 5 § { % \ { % { { . { % hrs., copious precipitum. cg. Metch, serum | Instantaneous clouding, in all tubes. In { { { { ( & 0.5 c.c. and 0.25 c.c. tubes copious precip- itum, which did not form in 0.1 c.c. tube until 6 hrs. In the two sets of tubes at 24 * hrs., same amount of precipitum. .5 c.c. Typhoid fil. 0.5 c.c. Metch, serum | No trace of reaction after 48 hrs., at 37° C. .5 “ { % { { 0.5 c.c. cholera “ c.c. Cholera fi $ { { { % .# c 5 { { { % { % { { { { | { # § * In the set of tubes which contained cholera filtrate and serum, where a slight reac- tion was first obtained, the addition of 0.5 c.c. of Metchnikovi serum likewise produced immediate cloudiness and precipitation in a few minutes. At eighteen to twenty-four hours there was a copious precipitum, showing that the cholera serum had fastened no appre- ciable amount of precipitable substance in its own filtrate. t A further addition of 0.5 c.c. of Metchnikovi serum to the mixtures of cholera filtrate and Metchnikovi serum produced immediate cloudiness and precipitation at one and one- half hours; at eighteen to twenty-four hours, copious precipitum ; most in 0.1 c.c. tube. The precipita were slightly greater than those of the homologous series. THE BACTERIAL PRECIPITINS 499 DETERMINATION OF THE MAXIMAL PRECIPITATION LIMITS OF THE ANTI- METCHNIKov1 SERUM For ITs HomoLOGOUS AND FOR THE CHOLERA FILTRATE. The maximal limit of reaction of the anti-Metchnikovi serum on its corresponding filtrate was then determined. TABLE XXII. 5 c.c. filtrate 1/10 c.c. serum. Cloudiness within several hours; precipitation - occurs before 20 hrs. 5 c.c. filtrate 1/20 c.c. serum. Cloudiness is formed within 10 hrs. ; never reaches a distinct precipitation even at 72 hrs. 5 c.c. filtrate 1/40 c.c. serum. Negative at 72 hrs. Maximal dilution about 1–100. With 5 c.c. of cholera filtrate and 1/10 c.c. of Metchnikovi serum, no reaction at seventy-two hours. Maximal dilution at which precipitation still occurred in the cholera filtrate was considerably under 1–50, the exact limit not being determined. - SUMMARY. - From these tests the following conclusions can be drawn: The serum of an animal adapted to a given species of spirillum may precipitate the filtrate of another species of spirillum, although it does not agglutinate the spirillum. Distinct precipitative rela- tionships may thus be present unassociated with agglutinative relationships. IV. THE PRECIPITINS DEVELOPED BY ADAPTATION OF RABBITS TO B. DIPHTHERIAE. It has long been known that during the process of immuniza- tion to certain species of bacteria the animal organism does not respond, or only with difficulty, with the production of precipitins, although various other antibodies, such as antitoxins, may be developed. We shall briefly describe our ineffectual attempts to obtain precipitins for the B. diphtheriae, the B. pyocyaneus, and the human and bovine tubercle bacillus. No attempt was made to immunize rabbits with diphtheria toxin, as Rruse and Nicolle could not determine the presence of precipitins in diph- theria antitoxin (although agglutinins have been described in horse antitoxin serum). Although abundantly established that precipitable substances are present in filtrates, “toxins,” the amount of precipitins developed in the serum of animals treated with filtrates was found by the writer and by other observers to be insignificant in comparison with that obtained when living bacterial cells are introduced. - 500 CHARLES NORRIS Rabbits were accordingly immunized with emulsions of agar cultures, grown for twenty-four hours, in the hope of obtaining a precipitating or anti- bacterial serum. A rabbit was finally immunized against many times the fatal dose of diphtheria culture, preliminary doses of antitoxic horse serum being given. (The diphtheria bacillus was kindly furnished by the Depart- ment of Health of the city of New York, and is the one commonly designated throughout the world as the American bacillus, or Bacillus “8.”) Agglutination value of the serum: An emulsion was made from a thirty- hour agar culture in sterile saline solution. Macroscopic test: 1–10: reaction distinct in 40 min. 1–100, 1–500, and 1–1,000: complete in 4 hrs. at 37° C. 1–5,000: showed a greater settling than did the 1–10,000 or the control tube. A second test failed to confirm the result above obtained, the emulsion being unsatisfactory on account of the clumping and sedimentation in the control tube. A third trial on June 14, ten days after the first test, was com- pletely negative even in a 1–10 dilution. It seems reasonable to concede that the serum had a relatively low agglutinating value, about 1–1,000. The fail- ures recorded in the second and third tests may perhaps be attributable to the condition of the emulsion. THE PRECIPITATING ACTION OF THE ANTIDIPHTHERIA RABBIT SERUM UPON THE DIPHTHERIA FILTRATE. The addition of 1/4 c.c. and 1/10 c.c. of serum to 1/2 c.c. tubes of diph- theria antitoxin (acid reaction), grown for three months at 37° C., caused no precipitation after several days at 37° C. Negative results were likewise obtained with a filtrate of extracts in normal salt solution of fresh agar cultures, even when mixtures of equal amounts (1 c.c.) of the antiserum and the solution were used.* RíSUMé OF PREVIOUS STUDIES UPON THE PRECIPITINS DEVELOPED BY ADAP- - TATION OF ANIMALS TO B. DIPHTHERIAE. - Diphtheria precipitins have been developed by S. Wassermann” in the serum of rabbits treated with 0.1 per cent. ethylin-diamin extracts of dried and pulverized diphtheria bacilli, previously killed by heat, 60° C., any toxic action being neutralized by administration of suitable quantities of antitoxin. Wassermann thus claims to have obtained a precipitating antidiphtheria serum which, unlike antitoxin, contains a substance which reacts specifically upon the body substance (the precipitable substance) of the diphtheria bacillus. The advantage claimed by him for this method over that by which merely the untreated bodies of the bacilli are inoculated is that more of the precipitable substance can be introduced. The serum, he claims, furnishes a method of differentiating the pseudo- from the true diphtheria bacilli by agglutination and precipitation tests. Wassermann hopes that with such a “bactericidal” serum the diphtheria bacillus may be made to disappear from the throats of patients who have continued to harbor the bacilli during a prolonged convalescence. Wassermann's expectation is premature, unless *The diphtheria bacilli of three potato tube agar slants were grown for three days at 37° C. and emulsified in 5 c.c. of sterile normal saline solution and filtered. THE BACTERIAL PRECIPITINS 501 one concedes that the precipitin is an immune body on a par in its action with the lysins, or is always accompanied by such a body, on which the assumption of the bactericidal powers of the serum is based.* Lipstein” has shown that by a procedure similar to the one employed by us—intraperitoneal inoculation of living diphtheria bacilli mixed with anti- toxin to neutralize the toxin — a strong agglutinating serum is developed in rabbits. The subject of the agglutinin in antidiphtheria serum, and the employment of such sera to differentiate the true diphtheria races from the pseudo-diphtheria races or species, cannot be entered upon, the reader being referred to the articles of Lipstein and Lubowski,” and to the admirable article of Josef Schwöner.” Schwöner casually mentions that he has observed specific precipitation of broth diphtheria filtrates after twenty-four hours at 37° C., with anti- diphtheria horse serum. V. THE PRECIPITINS DEVELOPED BY ADAPTATION OF RABBITS TO B. PYOCYA NEUS. With our stock culture of the B. pyocyaneus we were unable to obtain precipitating substances in the serum of several immu- nized rabbits for its homologous filtrates. The bacillus employed corresponded in most of its characters to the classical description of the B. pyocyaneus. With cultures of other races of this bacil- lus precipitins have been developed by Eisenberg.” VI. THE PRECIPITINS DEVELOPED BY ADAPTATION OF RABBITS TO HUMAN AND BOVINE TUBERCLE BACILLI. Our efforts to obtain precipitating substances in the serum of rabbits which had been infected with living or inoculated with dead cultures of human and bovine tubercle bacilli were likewise unsuccessful, so that our experiments may be briefly given. Rabbits inoculated with cultures of human and bovine tubercle bacilli, and bled at various periods of time after inoculation or before death, failed to develop precipitins in their sera for tubercle filtrates. Five per cent. glycerin broth cultures grown for several months at 37° C., in which the growth of bovine and human bacilli had been luxuriant, were used as the test fluids. The rabbits were inoculated with virulent cultures or with large amounts of the surface scum from tubercle cultures. The Sera tested were obtained from rabbits in various stages of tuberculization, from *This is an interesting commentary on the loose methods of expression frequently employed in connection with the immune bodies, especially since A. Lipstein 26 was not able to obtain any evidence of the formation of a suitable amboceptor in the sera of animals inoculated with living cultures, although the same sera possessed marked agglutinating properties. 502 - CHARLES NORRIS a local abscess at the site of inoculation in the subcutaneous tissue, to those with an acute general miliary tuberculosis of the peritoneum with more or less involvement of the lungs, lymph modes, kidneys, and other organs. No suggestion of a precipitating reaction was obtained by addition of the serum to their corresponding filtrates. Negative results were also obtained with Koch's extract of tubercle bacilli, which he employed to determine the agglutinating action of the serum of animals and patients afflicted with tuberculosis The solutions were made with the preparation of 2erriebene Tuberkelbacillus prepared by the Farbwerk vom Meister Lucius und Bruning, Höchst a. M., the directions given for Koch's Agglutinations- Flüssigkeit being closely followed. Our sera, added in various proportions, did not precipitate or cause cloudiness of this solution. VII. THE PRECIPITINS DEVELOPED BY ADAPTATION OF EABBITS TO THE STAPHYLOCOCCUs PYOGENES AUREUs, TO THE PNEUMO- COCCUS, AND TO THE STREPTOCOCCUS. - In the class of cocci three species were investigated as to their capacity for developing precipitins in the serum of rabbits which had been adapted by inoculation. of their cultures. The species were a Staphylococcus pyogenes aureus, a pneumococcus, and a streptococcus. The staphylococcus precipitin is first described. The pneumococcus and the streptococcus are described together for brevity, since by their inoculation no precipitating sera were formed. 4 a) THE PRECIPITINS DEVELOPED BY ADAPTATION OF RABBITS TO THE STAPHY- LOCOCCUS PYOGENES AUREUS. The staphylococcus used during the course of the experiments was highly pathogenic for rabbits, and it was only with care that these were finally rendered immune.* With the serum obtained from a highly immunized rabbit, the following experiments were made: - The rabbit received from March 10 to April 7 nine agar cultures into the peritoneum. The agglutination tests with the serum were made by the macroscopic method with emulsions of eighteen to twenty-four-hour agar cultures. 1–100 and 1–500: complete in 5 hrs. 1–1,000: nearly complete in 5 hrs. 1–5,000: complete in 18 hours. 1–10,000: negative after 24 hrs. 1–5,000 may be said to be the limit of the agglutination value of the Sé I’ll Iſl. *I am indebted to Dr. Wadsworth for the culture, which was obtained from a case of severe phlegmonous inflammation of the nose. THE BACTERIAL PRECIPITINs 503 TABLE XIII. THE PRECIPITATING ACTION OF THE ANTISTAPHYLococcus SERUM Upon THE STAPHYLo- - - COCCUS FILTRATE. Staphylococcus fil- || Antistaphylococcus Immediate cloudiness, well marked after 5 trate: Serll'In min. ; in 1 hr., distinct precipitate in sets of 1/2 c.c. 1, 4 c.c. tubes at room as well as at incubator tem- 1/2 c.c. 1/10 c.c. peratures; reaction complete at 5 hrs., the filtrate being clear. DETERMINATION OF THE MAXIMAL PRECIPITATION LIMITs of THE ANTISTAPHYLococcus SERUM UPON THE STAPHYLOCOCCUS FILTRATE. - The maximal limit of dilution of the antistaphylococcus serum at which precipitation occurs in the staphylococcus filtrate was found to be a serum dilution of 1–60. The serum likewise precipitated the filtered normal saline extracts of fresh agar cultures, as shown below.” THE PRECIPITATING ACTION OF ANTISTAPHYLococcus SERA UPON VARIOUS HETEROLOGOUs FILTRATES. The above serum and other antistaphylococcus sera were tested upon the filtrates already referred to of the colon-typhoid group, and also upon the filtrates of B. proteus, B. prodigiosus, Sp. cholerae, and Sp. Metchnikovi, with negative results. b) THE PRECIPITINS DEVELOPED BY ADAPTATION OF RABBITs To THE PNEU- MOCOCCUS AND TO THE STREPTOCOCCUs. The precipitins for the pneumococcus and streptococcus are evidently, as shown by the failure to obtain any evidence of their formation in immunized rabbits, only obtained, if at all, by special procedures. Our experiments may therefore be briefly summarized. Three rabbits were rendered highly immune to virulent cultures of the pneumococcus and of the streptococcusi grown upon agar. The antisera of the rabbits failed to produce cloudiness or pre- cipitates in their corresponding filtrates when tested in the follow- ing proportions: 0.5 c.c. filtrate + 0.25 c.c. serum. It might be * Emulsions of the three-day growth on three ag ur tubes in 5 c.c. of sterile normal saline solution, after standing for several hours at room temperature, were passed through Berkefeld filters. Equal parts of the filtered solutions and antiserum, when mixed, gave rise to no reaction after several hours; at eighteen hours (37° C.) a copious precipitum was present. - tſam indebted to Prof. Hiss for the streptococcus which was obtained from a severe case of cellulitis of the foot. The pneumococcus was virulent to rabbits, and was obtained from Dr. A. Wadsworth, to whom my thanks are also due. The filtrates were obtained from 5 per cent. glycerin meat infusion broth, grown for two months at 37° C. 504 CHARLES NORRIs concluded from these tests that the pneumococcus and strepto- coccus do not possess precipitable substances, or that, if present, suitable receptors are not present normally, or are sessile, or are not formed during the process of adaptation of rabbits to these species of cocci. . . . The following tests with Staphylococcus pyogenes aureus, indi- cate, however, that precipitable substances are present in the strepto- coccus and pneumococcus, but that these substances do not diffuse readily from their cells into their glycerin-broth cultures. This assumption furnishes an explanation of the non-precipitation of the filtrates by their corresponding antisera, although the same antipneumococcus and antistreptococcus sera precipitated the staphylococcus filtrate: THE PRECIPITATING ACTION OF THE ANTIPNEUMO- AND ANTISTREPTOCOCCUS SERA UPON THE STAPHYLOCOCCUS FILTRATE. 0.5 c.c. staphylococcus filtrate + 0.25 c.c. serum. The tests carried out in the above proportions show the formation of a Small quantity of whitish, Somewhat granular precipitate after eighteen to twenty-four hours. The tests were repeated several times with confirmatory results, the precipitum obtained being found free from bacteria on cultiva- tion and on examination of the coverslips. In confirmation of the specific character of the precipitum obtained may be cited the fact that neither the antipneumococcus nor antistreptococcus sera produced precipitates in any of the filtrates of the colon-typhoid group or the other filtrates in our posses- Sion. If the slight reactions obtained above are granted to be those of precipitation—and there seems to be no reasonable ground for doubting the correctness of this supposition—it seems fair to suppose that the precipitable substances exist in the pneumococcus and streptococcus. These substances, however, do not diffuse into the filtrates, being retained in the bodies of the cocci. When the living cultures are inoculated, they are set free and induce the formation of precipitins in the sera of the rabbits. These precipitins, however, possess only a relatively slight affinity for the precipitable substance in the staphylococcus, and hence the small reaction. The antipneumococcus and streptococcus sera do not therefore, produce precipitates in their corresponding filtrates, since little or no precipitable substances are present, these sub- THE BACTERIAL PRECIPITINS 505 stances being retained in the bacterial cell protoplasm and not diffusing throughout the ambient fluid of the broth.* Furthermore, it is reasonable to assume that no, or only an insufficient amount of, precipitable substance was present in the filtrate upon which the precipitins of our antipneumococcus and streptococcus sera could act with the formation of a precipitum, for the strongly precipitating antistaphylococcus serum does not yield precipitates in the pneumococcus and streptococcus filtrates. The growth in the broth from which these filtrates were obtained was not as abundant as that formed in the staphylococcus or in other bacillary filtrates. The filtrates of staphylococcus, on the other hand, are rich in precipitable substances, and the precipi- tins of the pneumo- and streptococcus antisera possessing a slight affinity for the staphylococcus precipitable substance, precipita- tion follows. - - RESUME OF PREVIOUS STUDIES UPON THE PRECIPITINS DEVELOPED BY ADAPTATION OF ANIMALS TO THE PNEUMOCOCCUS AND TO THE STREPTOCOCCUs. Pneumococcus.-Neufeld” obtained evidence of the formation of precipitat- ing substances in the pneumococcus antisera by means of special procedures. According to him, the precipitable substance resides exclusively in the bacterial cells—a statement apparently true for the pneumococcus. By means of the Solvent action of normal bile on a fresh pneumococcus culture, which does not alter the specific nature of the constituents of the cocci, a drop sufficing usually to dissolve the cocci of several cubic centimeters of broth culture, a solution is obtained to which, if a serum which agglutinates pneumococci in a dilution of 1–15 is added in the proportion of 1–2, 1–4, 1–8, refracting hyaline masses, the size of red corpuscles, are seen to form in the hanging drop after a quarter of an hour. The masses increase in size and clump, forming later a precipi- tate visible to the naked eye. Normal rabbit serum does not bring about this reaction. According to Neufeld, a serum which agglutinates specifically a bacterial species will always cause a precipitation in a solution containing a sufficient amount of the bacterial constituents. He leaves it uncertain whether one substance induces both reactions, as Kraus first claimed, or whether two distinct substances which are formed simultaneously in the serum induce these phenomena. Neufeld adheres to Bordet's theory of agglutination. Under the influence of the agglutinin a peculiar coagulation process takes place on the surfaces of the bacterial cells, which in the case of the pneumococcus is accompanied by visible changes in shape, and, secondly, * A somewhat similar explanation has been given by Neufeld for the pneumococcus, by Aronson for the streptococcus, and by Schwöner and Wassermann for the precipitable sub- stances of the diphtheria bacillus which is extracted from the cell protoplasm with diffi- culty. In our explanation the supposition is that the separation is made by the animal organism. 506 - CHARLEs NORRIS by an alteration in the physical properties of bacteria which induce clump- ing and change of equilibrium. - Dr. A. Wadsworth,” in our laboratory, has confirmed Neufeld's observa- tion concerning the precipitation by antipneumococcus serum of the solution of the pneumococci which have been dissolved by bile. Streptococcus.-Aronson” has shown that antistreptococcus sera which do not precipitate the ordinary streptococcus broth cultures nevertheless precipitate extracts of the cocci in 1 per cent. ethylin-diamin solutions. With a powerful antistreptococcus horse serum (for details the reader is referred to the original) he obtained typical agglutination of streptococcus cultures. The agglutination test, made by the macroscopic method, was absolutely typi- cal and caused only by a serum of high valency. Complete reactions were obtained with dilutions of 1–30; those of 1–40 were slightly cloudy, and the 1–50 dilutions were negative. The clumps formed in the effective dilutions, after Shaking, Soon settled again, unlike the sedimentation occurring in cul- tures which remain cloudy. A 1–40 reaction was obtained with Aronson's So-called twentieth normal serum. Aronson believes that the agglutinating power of serum is independent of its protective power. Besides the agglutinins, the serum contained a specific precipitin, but Special methods for its demonstration were necessary, as the serum did not precipitate streptococcus broth filtrates. The bacterial cell-bodies were dried and crushed in a little water, or better, with ethylin-diamin solution, as described by Aronson in a previous article.” The extract was filtered through paper until clear. Addition of his immune serum, in proportion of 1–10, to the Solution thus prepared, caused precipitation in several hours; whereas normal serum caused no cloudiness. The method resembles that described by Robert Koch * for the demonstration of the agglutinating and specific substances in the serum of individuals and animals. Aronson justly remarks upon the impropriety of Koch's designating this method one for the demon- stration of agglutination instead of precipitation. How a supposedly sterile and cell-free solution can agglutinate is certainly incomprehensible. Aronson believes that his process of extraction will reveal precipitins in the various antibacterial sera which have hitherto not been successfully demonstrated by merely testing the sera upon broth filtrates.* VI. THE PRECIPITATING ACTION OF THE ANTI-B. NO. 1, ANTI- TYPHOID, ANTICOLI, AND ANTIDYSENTERY SERA UPON THEIR HOMOLOGOUS AND HETEROLOGOUS SALINE SOLU- TION EXTRACTS OF YOUNG AGAR, CULTURES OF SPECIES OF THE COLON-TYPHOID GROUP. - Pick claims to have extracted from the typhoid bacillus two distinct substances, which he designated by the term “bacterio- coagulin.” The bacterio-coagulin A was obtained from old broth filtrates, and is precipitated by alcohol. The bacterio-coagulin R was extracted from the normal saline extracts of young agar *Tchistovitch reports that Marmorek has observed the precipitation of streptococcus filtrates by antistreptococcus serum. THE BACTERIAL PRECIPITINS 507 cultures, and was not precipitated by alcohol from its aqueous solutions. Although Pick was not able to develop precipitins in the serum of rabbits which had undergone adaptation by the use of these substances when purified, his absorption tests led him to conclude that each of the two bacterio-coagulins nevertheless developed separate and distinct precipitins in the serum of ani- mals adapted in the usual way to the typhoid bacillus. The action of these precipitins, sero-coagulins A and K, were specific, since they reacted only with their corresponding bacterio- coagulins. Granting that Pick's statements are correct, it would follow from our observations, referred to below, that precipitative affinities or relationships exist between the bacterio-coagulins A and K obtained from various species of the colon-typhoid group, the A body being found only in old broth filtrates, the K body only in the saline extracts of young agar cultures. * Our observations may be briefly summarized. The sera pre- viously referred to, which had been adapted to several species of the colon-typhoid group—namely, B. typhosus, B. No. 1, B. coli, and B. dysentery, “New Haven”—were employed. It was found that these antisera precipitated, not only the filtered normal saline solution extracts of agar cultures, grown for three days at 37°C., of the homologous species, but also the extracts of heterolo- gous species of this group. The reactions were slower in onset, and less copious precipitates were formed than in the case of old broth filtrates of these species. VII. RELATIONSHIPS EXISTING BETWEEN THE BACTERIAT, AGGLUTININS AND PRECIPITINS. A résumé of the various arguments advanced concerning the identity or non-identity of the bacterial agglutinins and precipi- tins cannot be entered upon here. The following series of obser- vations made with the antisera mentioned above and with other antisera may be briefly described, since they furnish further proof of the early views of French observers (Bordet, Tchistovitch) that *Pick’s statements regarding the specificity of bacterio-coagulins are in direct oppo- sition to the results which the author has obtained in a series of experiments which will be referred to in a later article. †The extracts were allowed to stand at room temperature or in cold storage, 5–8 C., for from 2–24 hours before filtration. 508 CHARLES NORRIS the bacterial agglutinins and precipitins are distinct substances. The antiserum of the B. No. 1, a dextrose fractor belonging to the intermediate “Durham” or hog cholera group, precipitated, it will be recalled, in high dilutions, not only the filtrates of the various species of this group, which has been subdivided into two types, corresponding to the two species of bacilli isolated by Schottmüller, and known as “Seemann” and “Müller” (Types A and B), but also in lower dilutions than the above, the following species, namely, the B. coli Escherich “non-saccharose fractor,” B. typhi, and various species of the indefinite dysentery group. The species of the intermediate group of the Seemann type, to which the B. No. 1 belongs—namely, B. “Seemann,” B. psitta- cosis, B. Cushing, B. typhi murium, B. icteroides—possess marked agglutinative affinities, the species agglutinating with high dilu- tions of anti-B. No. 1 serum, the agglutinative and precipitative affinities exhibiting a striking parallelism. The species of the Müller type of the intermediate group—namely, B. Schottmüller, Müller, B. Gwyn, and B. enteritidis (type determination ?)—have slighter, and with some antisera practically no agglutinative affinities. The precipitative affinities of the species of this group, although less marked than those of the species of the “Seemann” type (1–10–40 instead of 1–80–150), nevertheless are relatively of a higher grade and are more general than are their corre- sponding agglutinative affinities, as mentioned above. Our contention as to the distinctness of the agglutinins and precipitins is confirmed by our observations that several anti-B. No. 1 sera did not agglutinate B. typhi or the B. coli (Escherich type), although these sera precipitated copiously the typhoid and colon filtrates in dilutions above 1–10, but not in dilutions of 1–50. In this connection the agglutinative and precipitative relationships of antisera doveloped for the two tyyes of bacilli isolated by Schottmüller, for several species of the intermediate and dysentery groups are of interest. The anti-‘‘Seemann” serum agglutinated B “Seemann’’ 1–100,000; B. paracolon [Libmann] and B. No. 1, 1–64,000; B. typhi murium, B. psittacosis, and B. icteroides, 1–32,000; B. Cushing and a hog cholora bacillus designated by Professor Smith * as “motile,” 1–16,000; and the following species which belong possibly to the Müller type of the intermediate group: B. “Müller,” * We are indebted to Professor Theobald Smith for his motile and non-motile species of hog cholera, both cultures being dextrose splitters belonging to the intermediate group. THE BACTERIAL PRECIPITINS 509 1–50 (large motile clumps and scanty sedimentation, negative 1–200, micro- scopic method); B. Gwyn, 1–50, negative 1–200; and B. enteriditis, 1–200; also the following species: B. suipestifer (Král's) (hog cholera-Salmon-Smith), 1–16,000; B. suisepticus (Král's), 1–1,000 (?); and the non-motile species of hog cholera of Professor Smith, negative 1–50; B. dysenteriae “Y” Hiss, 1–250; B. dysenteriae Kruse, negative 1–50; B. coli (Escherich), 1–100; B. typhi, 1–100.* - Owing to unavoidable circumstances we were able to test the precipi- tating action of the anti-Seemann Serum upon only two filtrates of the species mentioned above. The maximal limit of precipitation for the homologous B. “Seemann” filtrate was 1–100, for the B. “Müller” filtrate, 1–40. We see again illustrated the independence of the agglutinating and precipitating activities of antisera, the “Seemann’’ serum precipitating the “Müller’” filtrate in dilutions of 1–40, whereas it fails to agglutinate the B. “Müller” in dilutions of 1–50, the homologous bacillus “Seemann” being agglutinated in dilutions of 1–10,000. - - With the anti-‘‘Müller” serum the following observations were made : The serum agglutinated B. “Müller”. 1–16–32,000, and B. Gwyn 1–8–16,000, but did not agglutinate in dilutions of 1–50, B. “Seemann,” B. No. 1, B. psittacosis, B. icteroides, B. typhi murium, B. paracolon Libman, and B. Cushing. The motile species of Professor Smith were positive 1–50, negative 1–500; the non-motile species, negative 1–50; likewise B. enteriditis; B. coli “Escherich,” positive 1–500; B. typhi murium, positive 1–500; B. dysenteriae, “Y” Hiss and B. dysenteriae, “Diamond’” Hiss, both negative 1–50; B. dysenteriae, Kruse, positive 1–50. On the other hand, the antiserum developed cloudiness and flocculi within ten hours in dilutions of 1–10 in the following filtrates of the species above mentioned, which it did not agglutinate in dilutions of 1–50: B. Schott- müller “Seemann,” B. Cushing B. No. 1, B. psittacosis, and B. dysenteriae “Y” Hiss, as well as in its homologous filtrates.[ We thus see that, although B. No. 1 and B. Schottmüller “Seeman" cannot *The tests were made in tubes of narrow caliber with agar cultures grown for 24 hours at 37° C., suspended in normal salt solution, 10 c.c. for one agar culture, readings being made at 2, 10 and 19 hours. Those recorded represent the extreme limits — microscopic clumps seen with low-power lens at 19 hours. Equal parts 9% c.c. of emulsion and serum dilution were employed. t The filtrates were obtained from broth cultures of similar periods of incubation, 79 days at 37° C. f Although the writer is of the opinion that his contention as to the distinctness of the bacterial agglutinins and precipitins, and consequently of the corresponding substances (agglutinable and precipitable substances), has been placed on a sound basis, nevertheless he realizes that many points underlying the intimate processes of agglutination and precipi- tation are still unknown. To illustrate, the following agglutination limits of an anti-B. No. 1 serum kept in cold storage for two years and greatly concentrated through evaporation, may be cited: B. No. 1, 1–40,000, practically complete and clear at eighteen, hours at 37° C.; microscopic clumps seen with aid of hand lens at 1–60,000. (No proagglutinoid zone at 1–20.) B. psittacosis: complete, 1–20,000; microscopic, 1–80,000. * B. Schottmüller “Seemann '': complete, 1–40,000; microscopic, 1–80,000. B. Schottmüller “Müller”; complete, 1–4,000; microscopic, 1–80,000. B. Cushing: complete, 1-8,000; microscopic, 1–80,000. B. Gwyn: complete, 1–2,000; microscopic, 1–80,000. \ 510 CHARLES NORRIs. be differentiated by their agglutinative reactions or biochemical characters, they nevertheless develop antisera with marked agglutinating differences indicating differences in the precipitable substances of the two species of bacilli, which may be explained, possibly, by the multiplicity of their precipi- table substances, or of their haptophor groups. The following observations made upon the agglutinating properties of the anti-B. No. 1 serum for the typhoid bacillus are of interest. An anti-B, No. 1 serum, with an agglutinating value of 1– 10,000, agglutinated the typhoid bacillus in dilutions up to 1– 2,000; another anti-B. No. 1 serum agglutinated the same typhoid bacillus incompletely in dilutions of 1–300; whereas a third antiserum failed to agglutinate in dilutions above 1–10. On the other hand, three different antityphoid sera possessed no agglu- tinating properties for the B. No. 1, although the three antisera possessed marked precipitating action upon the filtrates of the B. No. 1.% - Observations similar to those recorded above concerning the distinctness of the two substances as seen by the non-parallelism of their activities, expressed in terms of serum dilutions, were made in the case of the antityphoid sera which in dilutions of 1–10, and in greater dilutions in the case of some filtrates, pre- cipitated all the filtrates of the species belonging to the colon- typhoid group, enumerated in the text, whereas the serum (1–20,000) agglutinated the B. enteritidis (limit,1–300), the other species being clumped in much lower dilutions or in such as approach approximately the limits of normal rabbit sera. To conclude: an anticolon serum (agglutinating value, 1–250) precipitated the colon-filtrate in a dilution of 1–500; and in a dilution of 1–50 all the filtrates of the colon-typhoid group mentioned above, except B. Gwyn and B. Schottmüller (Seemann). *Many observers have called attention to numerous examples of the specificity of the agglutinins of various species of the intermediate group. Durham, Castellani, and other observers have, on the other hand, noted numerous examples of agglutinative relationships between the members of the colon-typhoid group. See BELJAEFF, “Ueber Paratyphus- Erkränkungen.”33 . • The reader is referred to a most suggestive article upon the agglutination affinities of related bacteria by Theobald Smith and Reagh.33 According to them, “there exist aggluti- nation relationships between the pathogenic groups of bacilli which ferment dextrose. . . . . This relationship is not brought out clearly unless the agglutinative limit of the various cultures is worked out with a serum agglutinating its specific bacillus in dilutions of one to one thousand and above.” THE BACTERIAL PRECIPITINs 511 On the other hand, the serum agglutinated only in low dilu- tions the homologous colon cultures, and was without action upon another culture of B. coli, as well as upon the intermediate B. No. 1 and the B. typhosus. - - A consideration of these observations leads us to the conclu- sion that the bacterial agglutinins and precipitins are distinct substances which owe their origin in the serum of adapted animals to different bacterial receptors or haptophor groups. According to Ehrlich's theory, the haptophor groups are alone responsible for the development of the immune bodies; hence the assumption that the substances in question are identical would leave wholly unexplained the fact that an antiserum may possess precipitative affinities of a high degree without or with only slight agglutina- tive affinities for a given heterologous species, or even for its homologous species, as in the case of one of the anticoli sera, or only in low dilutions with the other sera. On the other hand, if it be assumed that the substances owe their origin to different sets of receptors, we have only to assume that the haptophors (precipitable substances) which give rise to the partial precipitins* are more constantly present in the bacterial cell than are the cor- responding haptophors (partial agglutinable substances) which give rise to the partial agglutinins. VIII. CONCLUSIONS. The following conclusions may be drawn from our researches: 1. The normal serum of the rabbit and of the Ox, and various antibacterial rabbit sera, exert no precipitating action upon the usual peptone-salt meat infusions “or Liebig’s extract,” broth of slight alkaline or acid reaction. * - Normal rabbit serum and ox serum do not precipitate bacterial broth filtrates. We are therefore warranted in drawing the con- clusion that bacterial precipitins are absent in the serum of the rabbit and of the Ox. 2 a.) The sera of rabbits adapted respectively to four species of the colon-typhoid group, which species are distinguished from * Ehrlich and Morgenroth have assumed the existence of partial hemolysins; Durham and Wassermann, of partial bacterial agglutinins; and von Dungern, of partial hemopre- cipitins. - - 512 CHARLES NORRIS each other by definite and constant biochemical characters— namely, B. typhosus, B. No. 1, a member of the hog-cholera or intermediate group of Durham, B. coli communis, “Escherich,” and B. dysenteriae Shiga (New Haven “Duval”)—precipitate not only their homologous filtrates, but also the filtrates of the species above mentioned, as well as the filtrates of all the species of the colon-typhoid group which were tested by us; namely, the filtrates of B. psittacosis, B. typhi murium, B. enteri- tidis, B. icteroides, B. paracolon Gwyn, B. paracolon Cushing, B. Schottmüller, “Müller,” and “Seemann.” The reactions obtained by the sera of rabbits adapted to one species of the colon-typhoid group in its homologous filtrate and in the heterologous filtrates vary in the time of onset of the reac- tion, as denoted by the appearance of cloudiness, and in the copiousness of the precipitum which finally develops, cloudiness developing more quickly, and the amount of precipitum formed being greater, in the homologous than in the heterologous fil- trates. These differences are most strikingly exhibited when actively precipitating antisera are tested in higher dilutions; for an antiserum invariably precipitates its homologous filtrate in higher dilutions than the filtrates of heterologous species—a fact which may be made use of in the differentiation of bacterial species. The precipitins for species of this group, cannot, however, be considered specific in the strict sense, but rather generic; for, as mentioned above, a serum adapted to one species precipitates in low serum dilutions the filtrates of heterologous species. The bacterial precipitins may thus serve to indicate genus or group relationship. The term “specificity” has merely a quantitative and not a qualitative value, and should be applied only in this broad manner, for the bacterial precipitins. - The quantitative differences in the precipitating action of the sera of rabbits which have been adapted to various species of the colon-typhoid group for the filtrates of homologous and heterologous species of this group are best explained, we believe, at least for the present, by the assumption that mutual bacterial receptors—precipitable substances—exist among the species of THE BACTERIAL PRECIPITINS 513 one group of bacteria, which receptors give rise to the forma- tion of partial precipitins in the sera of adapted rabbits. b) The sera of rabbits adapted respectively to two species of the group of Spirillaceae—namely, Sp. cholerae asiae, and Sp. Metchnikovi—precipitate not only the homologous filtrate, but also the filtrates of the other species. The statements made above concerning the precipitins of the colon-typhoid group would seem to apply to the precipitins developed by the use of species of the genus of Spirillaceae. - c) In regard to the Coccaceae, we have reason to believe that precipitative relationships also exist between certain groups of this family, for slight precipitates were developed in the filtrates of Staphylococcus pyogenes aureus by antipneumococcus and antistreptococcus sera. -* 3. Precipitative relationships between species belonging to one group of Bacteriaceae for those of different groups of the same family, or for those of different families, such as the Coccaceae and Spirillaceae, as a rule, were found not to exist. Exceptions to this general rule were, however, noted. Thus, the serum of a rabbit adapted to a species of one group (B. prodigiosus) may pre- cipitate the filtrates of many different species belonging to other groups (colon-typhoid), and also those of species (Sp. Metch- nikovi) of a different family (Spirillaceae). The precipitative relationships are, however, limited. Thus the same antiprodigi- osus serum had no precipitative action upon filtrates of species (B. proteus) of other groups of the same family (Bacteriaceae), or of a different family (Staphylococcus pyogenes aureus). These facts lead us to believe that the precipitin reaction may serve to detect biological relationships between various groups and families of bacteria analogous to those which Nuttall and others have shown to exist in the animal kingdom. 4. From a consideration of the agglutinative relationships among members of the colon-typhoid group described by various observers, and from our limited number of observations and of the precipitative relationships described in our paper, the follow- ing conclusion may be safely drawn: The precipitative relation- ship between various species of one group of bacteria is a much 514 * CHARLEs Norris more intimate and constant one than the agglutinative relation- ship. The correctness of this assumption is supported and emphasized by the fact that the reaction of precipitation, as applied to bacterial filtrates, occurs only with low serum dilutions —as a maximum 1–500, but usually much lower; whereas agglu- tination readily occurs with thousandth dilutions. Furthermore, although agglutination and precipitation are usually closely associated phenomena, they are independent. In other words, the agglutinins and the precipitins are distinct substances. In conclusion, the writer desires to acknowledge his debt of gratitude to Professor T. Mitchell Prudden and to Professor Philip Hanson Hiss, for much help and many suggestions in the prep- aration of this article. REFERENCES. 1. KRAUs. Wien. Helin. Wehmschr., Aug. 12, 1897, p. 736. Preliminary com- munication, ibid., April 30, 1897, p. 431. 2. WASSERMANN. Samml. klin. Vortr., April, 1902, p. 336, footnote. - 3. TUPNIx. Münch. med. Wehmschr., March 10, 1903, p. 444. Also, TUPNIx AND PosNER. Prag. med. Wehmschr., 1903, 28, p. 205. 4. LINossIER AND LEMOINE. Comptes rendws de la Soc. de Biol., 1902, 54, pp. 85,276, 320,369. 5. Noguchi. Univ. of Penn. Med. Bull., 1902, 15, p. 295. 6. LAMB. Lancet, 1902, 2, p. 431. - - 7. OBERMEYER AND PICK. Wien. klin. Rundschaw, 1902, No. 16. * 8. MYERs. Lancet, 1900, 2, p. 98; also Centralbl. f. Bakt., 1900, 1. Abt. 28, pp. 237, 244. - 9. TOHISTOvITCH. Annales de l’Institut Pasteur, 1899, 13, p. 406. (See especially p. 415.) - 10. BUCHNER AND GERET. Münch. med. Wehmschr., 1901, 48, pp. 1163, 1275. 11. MICHAELIs. Deutsche med. Wehnschr., 1902, p. 733. 12. OBERMEYER AND PICK. K. k. Ges. der Aerzte, Mai, 1903. 13. ROSTOSKI AND SACCONAGHI. Berl. klin. Wehnschr., 1903, p. 114. Paper read Nov. 27, 1902. 14. LEVENE. Med. Rec., March 28, 1903. - 15. BRUNER AND KAYSER. Ztschr. f. Hyg. w. Infectionskr., 1903, 43, p. 401. 16. EISENBERG. Centralbl. f. Bakt., 1903, Th. 1. Abt. orig., 34, p. 739. 17. HIss AND RUSSELL. Med. News, N. Y., Feb. 14, 1903. 18. LENTZ, Ztschr, f. Hyg., 1903, 41, p. 559. 19. NUTTALL. Blood Immunity and Relationship: Precipitin Test. Cam- bridge University Press, 1904. - 20. von DUNGERN. Die Antikörper, 1902; also Centralbl. f. Bakt. u. Path., 1903, 1. Abt., orig. 34, p. 355. º 21. Joos. Centralbl. f. Bakt., 1903. THE BACTERIAL PRECIPITINs 515 22. 23. e . LUBOwsRI. Ztschr. f. Hyg., 1900, 35, p. 87. 25. 26. 27. 28. 29. WASSERMANN. Deutsche med. Wehnschr., Oct. 30, 1902, p. 785. 31. 32. LIPSTEIN. Deutsche med. Wehnschr., Nov. 13, 1902, p. 821. SCHwöNER. Wien. Jelin. Wehmschr., 1902, 48, p. 1274. LIPSTEIN. Centralbl. f. Bakt., 1. Abt. orig., 1903, 34, No. 5, p. 421. EISENBERG. Centralbl. f. Bakt., 1. Abt. orig., 1903, 34, p. 739. NEUFELD. Ztschr. f. Hyg. u. Infectionskr., 1902, 40, p. 54. WADsworth. Jour. of Med. Research, 1903, 10, pp. 228–42. ARONSON. Berl. klin. Wehnschr., 1902, 39, p. 979. ARONSON. Arch. f. Kinderheilkwºnde, 1900, 30, p. 23. Roch. Deutsche med. Wehmschr., 1901, 48, p. 829. BELJAEFF. Centralbl. f. Bakt., Referate, 1903, 33, p. 87. SMITH AND REAGH. Jour. of Med. Research, 1903, n. S. 4, p. 270. HEMORRHAGES IN THE EYE, PRESENT AT BIRTH. By DR. EDWARD B. COBURN, ASSISTANT SURGEON AND PATHOLOGIST TO THE NEW AMSTERDAM EYE AND EAR HOSPITAL, NEW YORK, (With three figures on 7 ext-Plate XZ.) HEMORRHAGES IN THE EYE, PRESENT AT BIRTH. By DR. EDWARD B. COBURN. ASSISTANT SURGEON AND PATHOLOGIST To THE NEW AMSTERDAM EYE AND EAR HOSPITAL, NEW YORK. (With three figures on Text-Plate XI.) / | * HIS study was undertaken to ascertain, if possible, the reason for poor vision in certain eyes which pre- sent no appreciable changes in the fundus of the eye com- mensurate with the diminution of vision. Ophthalmoscopic examination of amblyopic eyes some- times reveals a peculiar stippling or an unusual arrangement of the retinal pigment, but not necessarily beyond the range of mere physiological variation. Often these eyes squint, and, if in adults, ea anopsia is usually assigned as the reason for the poor sight. A restoration of the visual axes by operation and glasses do not increase the visual acuteness. I have had the opportunity to examine the eyes of a num- ber of young infants, the unnatural behavior of whose eyes had attracted the parents' attention. Peculiar retinal pig- mentation, diminution in the size of the retinal vessels, pallor of the disc, feeble pupillary reflex phenomenon, with pupils sometimes large and with uncertain or nystagmic movements of the eyes, were some of the conditions noted, either one or some combination of the above complex of symptoms being present. With the above changes appear- ing at such an early age, it seemed necessary to seek for the cause at the very threshold of life—at childbirth or before. While such investigations have been made before, the details surrounding each case have not been sufficiently Reprinted from the ARCHIVES OF OPHTHALMOLOGY, Vol. xxxiii., No. 3, 1904. Płemorrhages in the Eye, Present at Birth. 257 recorded to enable a proper estimation to be made of the causes of these changes. Laceration of the lids, injuries to the cornea, expulsion, avulsion, and rupture of the eyeball, retroversion of the lens and vitreous body without rupture of the eye, and hemor- rhages into various parts of the eye are some of the injuries to the eye recorded as occurring during birth. In addition, there may occur to more distantly related parts, indentation of the skull from pressure of the sacral promontory, fracture of the bones of the skull, epidural, subdural, and intracerebral hemorrhages. These injuries resulted during normal and instrumental delivery in mothers with a deformed pelvis. The prenatal period of the infant's life is often a strenu- ous one from the moment that labor begins. In normal labor of a primipara, the dilatation of the parts is accom- plished by enormous uterine vis a tergo, acting through the foetus, whose head serves as a battering ram to force the passage. In turn, the head is greatly compressed and suffers a moulding of the bones in its passage through the par- turient canal. Hours and days are often passed before the child is born, during all of which time the pressure is main- tained and periodically increased. A multigravida with roomy pelvis and relaxed tissues offers little resistance to the passage of the foetus, and in a few hours, or perhaps minutes, the process is terminated in the birth of the child. With deformed pelvis, or where for any reason instrumental delivery is necessary, the pressure is greater and is perhaps maintained longer and the danger is correspondingly increased. A large foetus with a large head, or, in fact, any condition that may make labor protracted, is a possible cause of retinal hemorrhages. At the termination of labor, the transition from the birth canal to the outer world is sudden, causing additional strain. With the exit of the head (in vertex presentations), the pressure on the body is maintained and the blood is forced into the head, which is now the only part of the child in a relieved state and free from pressure. The congestion of the foetal head produced by com- 258 Edward B. Coburn. pression during its passage through the obstetrical via vita is extended to the eyes also, and dilatation of the minutest capillaries is the usual occurrence. Diapedesis and rupture of the vessels lead to extravasa-. tion of blood into the tissues. The tissues are separated, distended, and compressed in different ways, suffering from the abnormal presence of blood in the tissues outside of the blood-vessels. Ophthalmoscopic Examination by Other Observers. In 281 personal observations Königstein found retinal hemorrhages in 29 cases, about IO per cent. Among I 50 children ranging in age from birth to eight days, Schleich found 78 eyes affected in 49 cases, about 33 per cent. In a week absorption of blood had usually occurred and only slight cloudiness of the retina remained. In one case he found hemorrhage of the retina eleven days after birth, but the blood corpuscles were changed. Bjerrum found only 3 per cent. of the cases he examined affected with retinal hemorrhages. De Vries examined II2 cases with the pupils dilated and found retinal hemorrhages in II per cent. of the cases. In I 5o new-born children examined by Montalcini from ten minutes to two days after birth, 25 per cent. had hemor- rhages in the retina, while nearly 18 per cent. had the hemorrhages in both eyes. These hemorrhages, which were mostly at the posterior pole of the eye are, according to these observers, monocu- lar or binocular, and roundish, irregular, striped, or radiate. Montalcini accepts the opinion of Schleich, who believes that venous stasis during the normal labor and, still more, dystocia play the principal rôle. He also considers the constitution of the foetus of importance, for most of the infants showing retinal hemorrhages were in bad health ; all infants showing retinal hemorrhages were head presenta- tions, and prolonged labor also seemed to be a factor; while these hemorrhages occurring by diapedesis absorb quickly, he states they may often represent the origin of functional troubles of the eye, and suggests therefore that we ought Hemorrhages in the Eye, Present at Birth. 259 to attend to the mother to ascertain faulty conformation of the pelvis and regulate the mechanism of labor. Microscopical Examination by Other Investigators. Naumhoff made an examination of the eyes of 47 chil- dren who had died shortly after birth, and found changes in I2, or about 25 per cent. The hemorrhages were chiefly in the nerve-fibre and ganglionic layers, with blood infiltra- tion of the tissues. In the macular region, hemorrhages had pressed apart the fibres in the inner nuclear layers. The blood was in fine stripes in the outer layers of the retina, and caused detachment of the retina. In the retinal capillaries tears were sometimes found. Extravasated blood was only observed three times in the choroid and once near the macula with detached retina. Naumhoff believes in a stasis due to cerebro-spinal fluid blocking up the sheath of the optic nerve (as in brain tumor) from the pressure. This causes both diapedesis and rupture of the capillaries. Von Hippel found hemorrhages in IO out of 24 cases, or about 41 per cent. The blood was chiefly in the anterior part of the retina in the region of the Ora Serrata, and some- times distributed over the whole fundus; seldom in the optic nerve or choroid, and never in the ciliary body, iris, anterior chamber, or sclera. He found the region of the fovea often involved. In some sections blood was observed to have forced its way outward from the inner nuclear layers to the vitreous membrane and had collected there, causing detachment of the retina. Hemorrhages were also found in the papilla. He calls attention to the possibility of the formation of blood cysts from this cause. The hemorrhages in the retina are frequent, and may be the cause of congenital amblyopia or of the so-called macular coloboma. In a case reported by Wintersteiner of a child whose eye was injured in instrumental delivery, a microscopic examina- tion showed that the anterior chamber was filled with blood, and the ciliary bodies, the suprachoroidal space, the choroid, and Sclera contained extravasated blood. He states that hemorrhages are usually found in the posterior segment of 26O Edward B. Coburn. the eye, and that hemorrhages in the anterior part were probably due to the pressure of the forceps. Axenfeld states that he has seen abundant hemorrhages in the eyes of infants a few days old, but they all contained hematoidin, which seemed to prove that they did not occur during birth but during intrauterine life. Wintersteiner could find no rupture of the vessel walls, and thinks the rupture may have closed by the time of death. In view of Schleich's observations as to the rapid absorption of the extravasated blood, Axenfeld's conclu- sion does not seem to be warranted. Ammann, in speaking of retinal hemorrhages in general, states they may occur from diapedesis, or by rupture of the vessels, the latter being mostly due to trauma. They may also occur from muscular strain, but in these instances he presupposes vascular changes. He also states that the effect on the retina is probably due to mechanical pressure. Tepljaschin, in experimental injuries of the retina, shows that division of the retinal nerve fibres causes ascending and descending degeneration of the nerve fibres, with degenera- tion of the ganglion and other cells of the retina. Chas. S. Bull, in discussing retinal hemorrhages, states that degeneration of the vascular coats in the young causes subhyaloid hemorrhage. If in the vitreous and extensive, the prognosis is grave, as blood comes from the peripheral vessels. Any fundamental disease, as inherited Syphilis, which causes abnormal brittleness of the vessel walls, is influential. Simple retinal hemorrhage, without retinitis, indicates increased blood pressure, a disease of the blood- vessel walls, or a morbid condition of the blood, or all three. Recurrent retinal and subhyaloid hemorrhages in the young are of slight prognostic importance, especially if due to in- herited or acquired syphilis. In an article on “Retinal Hemorrhages as Diagnostic of Fracture of the Base of the Skull and Subarachnoid Hemor- rhages,” Ballantyne concludes that a subarachnoidal hem- orrhage, if sufficiently rapid, will cause retinal hemorrhage, or, if the effusion is unilateral, the hemorrhage will be con- fined to the same side. It is difficult to explain why the Hemorrhages in the Eye, Present at Birth. 26 I sinus basalis does not distribute equally to both optic-nerve sheaths, though he believes that this peculiar fact may be explained by hydrostatic principles. Blood enters from the sinus basalis, and not from vessels of the pial sheath. Presumably a sudden distention of the inner sheath space just behind the eyeball causes great increase of pressure in retinal vessels and rupture. Unilateral hemorrhages occur in unilateral subarachnoidal hemorrhages. The Author’s Studies. These studies were conducted in the laboratory of the College of Physicians and Surgeons, and for the opportuni- ties afforded I am greatly indebted to Dr. T. M. Prudden. The eyes of 37 infants who had died at a maternity hos- pital were examined. These were either still-born or had lived for varying periods up to twenty-two days. The eyes were removed and placed in Orth's fluid, and after harden- ing in successive alcohols of increasing strength were im- bedded in celloidin and cut in serial sections. Staining was either en bloc in alum carmine or in sections with hema- toxylin and eosin. Author’s Observations. Macroscopical,—On opening the eyes it was often possi- ble to see the hemorrhages with the unaided eye, or by viewing the divided eyeball with transmitted light. These had the appearance of brownish spots, usually circular in form, though often flame-shaped and striated, occasionally presenting the appearance of several having coalesced. Con- trary to the findings of other investigators, most of the hemorrhages were found in the equatorial part of the eye, up to the ora serrata, the posterior part, in particular the macular region, presenting fewer extravasations. In only one case did the blood appear to have penetrated into the vitreous. Detachment of the retina in the usual papillo- macular fold was generally observed, and where other de- tachments were noted they were usually linear, following the course of the blood-vessels, while the vessels themselves could frequently be seen at the top of the fold. 262 Edward B. Coburn. Microscopical.—A fact worthy of note is the customary intense congestion of the eyes of the new-born, fully four- fifths of the eyes examined showing the vessels gorged with blood even where there had been no extravasation. This natural injection of the vessels and capillaries of the retina reveals the fact that the retina is not so sparsely supplied with blood-vessels as is ordinarily supposed. Hemorrhage anterior to the iris is comparatively rare, blood on the anterior surface of the iris and in the meshes of the pectinate ligament being observed but once. Blood in the outer layers of the sclera was seen five times. Cho- roidal hemorrhage was not observed, though in one instance the lamina vitrea was ruptured at the ora serrata (see Fig. I, Text-plate XI), and blood, evidently from the choroid, had passed through the break, causing detachment of the retina. Suprachoroidal hemorrhage was observed once. The retina was the chief site of the extravasation in all of the 17 cases in which hemorrhage was found. Hemorrhage in the optic-nerve head was noted 4 times, in the optic-nerve sheath 5 times, in the subretinal fluid 3 times, and in the subhyaloid 4 times. The vitreous was involved only twice, though blood corpuscles were found in some cases where the extravasation appeared to be subhyaloid only. Hemorrhages were most frequent in the inner layers of the retina, especially in the nerve-fibre layer (see Fig. 2, Text-plate XI) (in which the large blood-vessels lie), and in the layers contiguous thereto—the ganglion and the molecu- lar or plexiform layers. The extravasation in the nerve-fibre layer was usually somewhat linear in form, the blood prob- ably following the line of least resistance, the cleavage being parallel to the course of the nerve fibres. Next in frequency the hemorrhages occur in the outer molecular and the ganglion layers in which the smaller blood-vessels are located. Differing, however, from the hemorrhages in the nerve-fibre layer, these extravasations were usually cir- cular or oval when they were of considerable size. The limitans externa seemed to form a natural barrier to the escape of blood toward the columnar layer. This layer was ruptured in several cases, and blood was observed in a con- Hemorrhages in the Eye, Present at Birth. 263 tinuous line from the middle region of the retina to the pigment layer and the lamina vitrea, where it was heaped up, causing detachment of the retina at that point. In every instance but one where this had occurred, the pig- ment cells were undisturbed except in places where the lamina vitrea seemed to be ruptured. The subhyaloid hemorrhages (see Fig. 3, Text-plate XI) were usually massive with rounded outlines and had occurred as the result of rupture of the blood-vessels lying close to or on the inner surface of the retina, as the most common form of extrava- sation was diffuse, consisting of only a few red blood cells scattered here and there through the tissues. It is quite possible to believe that these were caused by diapedesis, but rupture of the blood-vessels must have occurred to pro- duce the large collections of blood seen in some eyes. The disrupting effect of these hemorrhages is another proof of the rupture theory, as large cavities entirely filled with blood were seen wherein were no signs of retinal elements to be seen. The extravasations of blood in the sheath of the optic nerve were not large but consisted usually of a few corpuscles lying free in the intervaginal space. There was no evidence of distention of the sheath, though this might very well have existed during birth and subsided after the removal of the compression of the head from the birth canal and the length of time which elapsed between the birth, death, and final examination of the eyes. Unless the extravasation had been local or there had been profuse hemorrhage in the cerebro-spinal fluid, marked evidence of such extravasation could not be expected. The blood must have been very greatly diluted by the relatively large quantity of the cerebro-spinal fluid, and on the return of the fluid to the subdural lymph spaces only a few corpuscles would become stranded in the meshes of the lymph spaces surrounding the optic nerve. REMARKS." Childbirth usually causes intense congestion of the head and its contents, and the eyes are probably the greatest * Deduced largely from the tables, 264 - Edward B. Coburn, sufferers from hyperaemia occurring at that time. Retinal hemorrhages occur chiefly in vertex presentations, as the head is then the part which causes dilatation of the par- turient canal and suffers most. In breech presentations the head follows in the canal already dilated. Extravasations are more apt to occur in congenitally weak children. This is shown by the fact that the average life of these children is less than that of children not so affected. Long and dry labors, infants with large heads, mothers with deformed pelves, and instrumental delivery are contributing causes. Conversely, Small children, short labors, large and roomy pelves and vaginae, are conditions which favor birth without ocular hemorrhages. This applies also to premature and induced labors, as in these cases the infants are not fully developed. While it has not been proven that forceps delivery is a cause (and statistics on this point ought to be secured where forceps are used without de- formity of the pelvis), still, forceps are probably used where labor is prolonged, the head is large, or the pelvis is de- formed, all being conditions where unusual or long-continued compression of the head is maintained. In many lying-in. hospitals forceps are always used if descent of the head does not follow in two hours after the os uteri has been suffi- ciently dilated for its passage. - - Subjects with hemophilia or with hemorrhagic tendencies, should be regarded with suspicion, and birth should be facilitated as much as possible. Hemorrhages in the eyes of colored infants are more frequent than in those of white parentage. Extravasations are often visible two weeks and longer after birth, a point which must not be forgotten, and which may be of some diagnostic importance in making microscopical or ophthalmoscopic examinations of infants' eyes. The frequency of retinal hemorrhages is remarkable and somewhat difficult to explain when contrasted with the infrequency of choroidal hemorrhage and the exceedingly vascular structure of that membrane. The free anastomosis of the choroidal vessels compared with the terminal arteries in the retina with their lack of intercommunication should not be forgotten. Another element tending to cause these Hemorrhages in the Eye, Present at Birth. 265 hemorrhages (and which we must remember are limited solely to the retina) is suggested by our knowledge of the effects of cerebral tumors, and the dilatation of the optic-nerve sheath (which is a cul-de-sac of the cerebral lymph spaces) with the cerebro-spinal fluid, and the consequent compression of the optic nerve and the central vessels of the optic nerve and retina. A consideration of the circulation of the eyeball will convince us that this is the only place where such a compression of the vessels supplying the retina can take place without effecting a corresponding change in the ves- sels which supply the other coats of the eye. The occa- sional finding of a few blood corpuscles stranded in the sheath of the optic nerve is partial evidence of this state- ment. As has already been suggested, the cerebro-spinal fluid returns immediately to the cerebral lymph spaces after birth, leaving practically no trace of its former presence, ex- cept its secondary effect in producing retinal hemorrhages. Relation of Retinal Hemorrhages to Impaired Vision. Hemorrhages in the eyes of new-born infants suggest difficult or protracted labor, and are apt to be found in infants of low vitality. Of course the traumatism of birth may be so severe that life may not be possible, or even desirable, with injuries resulting in paralysis, blindness, or other defects. These hemorrhages, if numerous and sufficiently exten- sive, may cause defects in the retina, with destruction of the ganglion cells and nerve fibres, and ascending and de- scending degeneration of the nerve fibres, detachment of the retina, and colobomata—all of these conditions pro- ducing blind spots (scotomata) in the field of vision. If the hemorrhages are subhyaloid, they are probably absorbed, leaving no trace of their presence behind. If they pene- trate the vitreous, they may disappear without any sign being left. It is a question whether they may not be the origin of certain forms of angioid streaks or connective- tissue formation in the retina. Whether retinal hemor- 266 Edward B. Coburn. rhages have any relation to retinitis pigmentosa is also a matter of interest. Hemorrhages at the macula may be sufficiently destructive to cause great impairment of vision, loss of central fixation, and subsequent strabismus. If the hemorrhage is subretinal, absorption or disturbance of the retinal pigment may occur, destruction of the layer of rods and cones, and even detachment of the retina. Conclusions. While the cases here recorded are too few in number to permit definite conclusions, certain probabilities are suggest- ed. The statistics herein presented are taken from cases which were congenitally weak or unable to survive the trau- matism of birth. The proportion of those affected with retinal hemorrhages ought rather to be taken from observa- tion of living children. Although the ophthalmoscope does not reveal the finer or more peripheral changes, only the microscopic examination of the eyes of infants who have died shortly after birth, and in whom the effects of birth may be excluded, will give the proper proportion. Childbirth causes congestion of the head and eyes. Vertex presentations produce greater congestion than breech presentations. Long and complicated labors are causative factors. The retina is the structure most frequently affected, and is often the site of extravasation of blood from the retinal vessels. * The probable cause of retinal hemorrhages is a disturb- ance of the retinal circulation resulting from compression of the optic nerve and its central blood-vessels due to a disten- tion of the optic-nerve sheath with cerebro-spinal fluid from compression of the head occurring at birth. The colored race and persons of a hemorrhagic diathesis are most susceptible to this form of traumatism. While the percentage of children in whom these hemor- rhages occur is relatively large, yet their destructive effect is very limited, the blood generally being absorbed without Hemorrhages in the Eye, Present at Birth. 267 leaving any trace ophthalmoscopically, or any appreciable defect in the field of vision. These hemorrhages may cause sufficient injury to affect the visual acuteness or the field of vision, and may leave ophthalmoscopic evidence of their previous existence. DETAILS OF CASES. CASE I.-Twenty-two; mother white; second pregnancy; one miscarriage; flat pelvis; R. O. P.; duration of labor, twenty-nine hours; high forceps, long second stage; child large and deeply asphyxiated; died in forty-five minutes; firm pressure on fundus to expel. EYEs: Contain profuse hemorrhages in all the layers of the retina, subretinal, subhyaloid, optic-nerve head and sheath, sclera; rupture of lamina vitrea. - CASE 2.—Mother, twenty-nine years, colored; sixth pregnancy; L. O. A.; duration of labor, nine hours; child lived two days. Autopsy: status lymphaticus; congestion of lungs and a cloudy swelling of kidneys and liver. EYEs: Few small diffuse hemorrhages in inner layers of retina. CASE 3.−Mother, twenty-six years; fourth pregnancy; two still-born; L. O. A.; duration of labor, thirty-six hours; child asphyxiated, and poorly nourished; lived one and one-half hours. Autopsy: atelectasis, with Oedema and congestion; cloudy swelling of liver and kidneys; subpleural and subpericardial hemorrhages; hyperplasia of Malpighian bodies. EYEs: Large and small diffuse hemorrhages in inner layers of retina, vitreous, episcleral. CASE 4.—Mother, twenty-one; primipara; L. O. A.; labor, thirty- one hours; asphyxia neonatorum; hydramnios; lived fourteen days; green stools; infection of navel. Autopsy: purulent in- flammation of the cord and peritonitis; hyperplasia of mesenteric lymph nodes and spleen; cloudy swelling of kidney and liver; atelectasis and congestion of lungs. EYEs: Large circular hemorrhages in periphery and smaller ones in equatorial region. Hemorrhages in inner layers of retina, episcleral, and sheath. CASE 5.—Mother, twenty; second pregnancy; L. O. A.; labor, one hour; child lived eighteen days; green stools; purulent dis- charge from eyes; pneumonia. 268 Edward B. Coburn. EYES: Numerous hemorrhages in two inner layers of retina and in sheath of optic nerve. . CASE 6.-Mother, aged twenty-nine; third pregnancy; L. O. A.; labor, eleven hours; narrow pubic arch; low forceps; child, deep asphyxiated; lived fifteen days; green stools; abscess of occiput. . Autopsy: lobular pneumonia; purulent pericarditis; abscess right kidney; congestion and hyperplasia of spleen and mesenteric lymph nodes. EYEs: Hemorrhage in all layers of retina, subretinal, scleral, and episcleral; suprachoroid and sheath of the optic nerve. CASE 7-Mother, twenty-seven; primipara; slight albuminu- ria; R. O. A.; labor, thirty-three hours; child lived four days. Autopsy: lobular pneumonia; multiple punctate hemorrhages of pleura and pericardium. EYES: Hemorrhages in three inner layers, subretinal, and sub- hyaloid. CASE 8.—Mother, thirty-one; second pregnancy; slight albu- minuria; L. O. A.; labor, fourteen hours; still-born. EYES: Extravasations in two inner layers of retina and in sheath of optic nerve; great Oedema of retina and choroid. CASE 9.—Mother, twenty-six; multipara; labor, fifteen hours; L. O. P.; still-born. EYES: Large; hemorrhages in two inner layers of retina, in optic-nerve head, and episcleral. - - CASE Io.—Mother, twenty-one, colored; primipara; L. O. A.; duration of labor, fourteen hours; lived fifteen days; green stools. Autopsy: gastro-entero-colitis. EYEs: Hemorrhage of nerve-fibre layer of retina and sub- hyaloid. CASE II.-Mother, thirty-one, colored; seventh pregnancy; suspected syphilis; rickets; five still-born children; R. O. P.; labor, thirty-seven hours; child died suddenly on third day; con- gestion and atelectasis of lungs and subpleural hemorrhages. EYES: Show a few small hemorrhages in periphery. CASE 12.-Mother, twenty-nine; second pregnancy; justo- major pelvis; breech presentation; labor, seventeen hours; child lived twelve hours. Autopsy: atelectasis of lungs. EYEs: Intense congestion of retinal and choroidal vessels; diffuse hemorrhages in inner layers of retina, in optic-nerve head and sheath. CASE 13.−Mother, twenty-nine; third pregnancy; L. O. P.; Hemorrhages in the Eye, Present at Birth. 269 labor, seven hours; child lived four days. Autopsy: bronchial pneumonia. EYEs: Hemorrhages in inner layers of retina and in vitreous. CASE I4.—Mother, thirty-seven; seventh pregnancy; two mis- carriages; L. O. A.; dry and premature labor, lasting eleven hours; child lived four days. Autopsy: atelectasis and conges- tion of lungs; cloudy swelling of kidneys and liver; punctate hemorrhages of pleura. EYEs: Show hemorrhages in inner layers of retina. CASE 15.-Mother, twenty-two, colored; primipara; albuminu- ria; L. O. A.; labor, eleven hours; child lived six days. Autopsy: hemorrhages in internal organs. EYEs.—Hemorrhages in periphery of one and generally dis- tributed in inner layers of retina of other eye. Hemorrhages in anterior portions of iris and in angles of anterior chamber. CASE 16.—Mother, nineteen, colored; primipara; rickets; R. O. A.; premature labor, four hours; child lived four days; jaundiced; ecchymosis of skin; prematurity and hemophilia. EYEs: Only one eye had few small hemorrhages in periphery. CASE 17.-Mother, twenty; primipara; L. O. P.; low forceps; labor, thirty-two hours; child lived sixteen days; cough. Au- topsy: atelectasis; congestion and Oedema of lungs; cloudy swell- ing of kidneys and liver. EYEs.—Several small hemorrhages in peripheral portion of retina of one eye. 27O Edward B. Coburn. An investigation of the records reveals the following interesting facts: I. With reference to the child. No hemorrhages. Hemorrhages. Cases. . . . . . . . . . . . . . . . . . . . 2O I7. Average life. . . . . . . . . . . . . . 9 days 7 days Longest life. . . . . . . . . . . . . . 22 ** - I8 ‘‘ Shortest life. . . . . . . . . . . . . . I still-born I still-born Lived less than 12 hours.... I 3 Lived 3 days and over... . . . I7 . I I Lived I4 days and over. ... 5 - 5 Twins... . . . . . . . . . . . . . . . . . 2 Jaundiced. . . . . . . . . . . . . . . . 2 I Large head, no moulding (I9 hours' labor) . . . . . . . . I Large child (labor 29 hours) I Cord around neck. . . . . . . . . I Child lived (maternal pelvis deformed). . . . . . . . . . . . . . I I still-born; I IO days; I2 days; I5 days hour One eye only affected. . . . . . 2 Breech presentation. . . . . . . . 5 I (deformed pelvis) Vertex presentation. . . . . . . . - • - Face presentation... . . . . . . . I (44 hours) In the cases with ocular hemorrhages the following presenta- tions were noted: L.O.A., II; L.O.P., 2; R.O.A., 2; R.O.P., 2. 2. Maternal statistics. - No hemorrhages. Hemorrhages. Youngest. . . . . . . . . . . . . . . . . . 20 IQ Oldest . . . . . . . . . . . . . . . . . . . 37 37 Average age. . . . . . . . . . . . . . 26 , 26 Primipara. . . . . . . . . . . . . . . . 6 6 Multipara . . . . . . . . . . . . . . . . I4 II Duration of labor, longest. .. 56 hours 37 hours ( \| { { shortest... I hour I hour { { “ average. I6 hours I8; hours Over average labor. . . . . . . . 5 7 Albuminuria. . . . . . . . . . . . . . 5 3 Syphilis. . . . . ... e. g. s. ſº tº º e º e º º 2 I Rickets . . . . . . . . . . . . . . . . . . 2 2 Deformed pelvis. . . . . . . . . . . 2 (49 and 9 hours) 3 (II, I7, and 29 hours) Colored . . . . . . . . . . . . . . . . ... 2 5 Version . . . . . . . . . . . . . . . . . . 6 Forceps. . . . . . . . . . . . . . . . . . I I high; 2 low (“worst” cases; I, deformed - pelvis) Placenta previa. . . . . . . . . . . I Premature, dry, or induced labor . . . . . . . . . . . . . . . . . . 6 3 Strong pressure on fundus - to expel child. . . . . . . . . . . I (As breech presentations are said to cause less compression of the head, it is interesting to note that where the pelvis was deformed retinal hemorrhages were also found.) .# Hemorrhages in the Eye, Present at Birth. 271 REFERENCES. BALLANTINE, Manual of Antenatal Pathology and Hygiene. THOMPson and BUCHANAN, Trans. Oph. Soc. U. K., vol. xxii. SCHAPRINGER, AV. Y. med. Monatsschr., xii., II, p. 591. JARDINE, Brit. Med. }our., 1901, p. 401. SNELL, Meeting Oph. Soc. U. K., Nov. 4, 1902. MORRIS and OLIVER, System of Diseases of the Eye, vols. iii. and iv. KöNIGSTEIN, Wiener med. Şahresbericht, 1881, p. 47. SCHLEICH, Mitteilung aus der A limić, Tübingen, 1884, vol. i., p. 44. BJERRUM, Congress, Copenhagen, 1884. p NAUMHOFF, Arch. f. Augenheilk., vol. xxxvi., 3, p. 180. WINTERSTEINER, Zeitschr. f. Augenheilä., Oct. 11, 1899, p. 443. MONTALCINI, Rivista de Ostetricia, Ginicologia et Pediatria, Marzo-Aprile, 1897. - * FLEMING, Edinburgh Med. }our., April, 1903. TRUC, Ann, d'oculistique, cKix., p. 161. AMMANN, Beiträge z. Augenheilk., 1898, 33. TEPLJASCHIN, Arch. f. Augenheilk., xxviii. VON HIPPEL, Arch. f. Ophth., clv., 2, pp. 286 and 313, 1898. German Ophth. Society at Heidelberg, Aug., 1897. DE VRIES, AVed. Zydschr. d. Genees&., 1901, I, No. 6. STEINHEIM, Deutsche med. Wochenschr., xvii., p. 249, 1883. BAAS, Arch. f. Ophth., 1898, 641. BULL, Med. Alec., Feb. 3, IQoo. Arch. of 0phthal., Vol. XXX///. Text-Pſafe Yſ. FIG. I. | FIG, 2. FIG. 3. FIG. I.-Retinal hemorrhages in the nerve-fibre layer, displacing and compress- ing the normal tissue. FIG. 2.-Retinal and choroidal hemorrhage; rupture of the lamina vitrea. The retina is detached by blood coming from the choroid through a rupture in the lamina vitrea. Diffuse hemorrhage in the nerve-fibre layer. FIG. 3.-Subhyaloid hemorrhage, in its usual rounded form. AN ADENOMA OF SEBACEOUS GLANDS OF THE ABDOMINAL WALL.” WITH A REVIEW OF THE LITERATURE OF ADENOMIA AND EIYPERPLASIA OF THE SEBACEOl.JS GLANDS. BY WILLIAM COGSWELL CLARKE, M.D., OF NEW YORK. THIS tumor is of surgical and pathological interest both from its extreme rarity and from the difficulty of diagnosis. Dr. Emil Bock, in 1880, probably gave the first good descrip- tion of a real adenoma of sebaceous glands, accompanied by drawings of microscopic sections. Many writers have de- scribed hyperplasia of the glands and also primary lesions of the skin where the glands have become secondarily involved, and called them “Adenomata Sebacea.” But if one takes into consideration the classification of Unna, which is quoted in the latter part of this article, the difference is manifest. Present Case.—In July, IQO3, a Jewess, eleven years of age, was referred from the Medical to the Surgical Class of the Out- Patient Department of the Roosevelt Hospital, for the removal of a tumor of the abdominal wall. Past History, She had always been in good health and had never received an injury of any kind. Menstruation had not com- menced, although pubic and axillary hair was fully grown and her breasts were well developed. Present History.—Three years ago she noticed in the skin a tumor about the size of a pea, situated three centimetres above and five centimetres to the left of the umbilicus. It lay in a line drawn from the nipple downward and slightly inward. Both she and her mother were positive that the growth had not always been present, and also that from the time of its first appearance it had * A study from the Department of Pathology of the College of Phy- sicians and Surgeons (Columbia University), New York. 488 WILLIAM COGSWELL CLARKE. slowly increased in size, at the same time becoming dark in color. . Three months ago it suddenly began to grow more rapidly. There had never been any pain or unpleasant sensation connected with it, and she had not lost flesh or strength. Physical Examination.—Inspection. The patient was well nourished, apparently in perfect health, large and strong for her years. No growths were seen on any other part of the body. The tumor was oval, smooth, and prominent, about six centimetres by five centimetres in diameter, bulging forward three centimetres. There was some wrinkling of the skin at the angle where it turned to pass over the tumor. On the dome of the growth for an area of three centimetres, the skin was thin, smooth, and dark, on account of the presence of many enlarged blood-vessels. Palpation.—The tumor was a soft, fluctuating mass appar- ently made up of two cysts, one considerably larger than the other. It was intimately adherent to the deeper layers of the skin. The superficial layers of the latter were movable over the growth, showing not even the slightest point of attachment. The growth was freely movable on the deeper parts; the abdominal muscles put upon the stretch did not at all limit its mobility. Operation.—Ether anaesthesia. An elliptical incision sur- rounding the tumor and exposing its base allowed an easy dis- section of the tumor from its bed in the fat of the subcutaneous tissue of the abdominal wall. After its removal, a certain amount of fat and subcutaneous tissue still remained undisturbed, over- lying the abdominal muscles. The wound was readily closed by three deep catgut sutures, and by black silk for the skin. The wound healed by primary union and the sutures were removed on the eighth day. Morphological Examination.—Macroscopic. The material consisted of seven cysts, which composed about four-fifths of the tumor, and the balance was an irregular wedge-shaped mass of firm tissue situated between the two largest cysts. The average size of the cysts was about two centimetres in diameter, the largest, four centimetres; the smallest, three millimetres. Five of these cysts contained a thin, watery, dark-brown fluid, while the smallest (Plate I, P) was filled with a dark, firm, sebaceous material. The true skin was not attached to the surface of any of the cysts, although the mass was indirectly adherent to the skin because of its situation in the subcutaneous connective tissue. PLATE I. PLATE II. PLATE III. | || ~ ~\，\！！！！！！！！$$（） / \ （Š№ , \ \！！！ſae ！\，， ºſſºſ ~ ſae £§§ PLATE IV. ADENOMA OF SEBACEOUS GLANDS. 489 Microscopic Examination.—Two-fifths of the firm tissue was stroma and the balance epithelial masses and small cysts. The stroma consisted of rather dense connective tissue in bands, whose fibres were arranged in parallel rows. These bands ran between the epithelial alveoli and at times in the alveoli, acting as trabe- culae. In places, the connective-tissue bands, especially some of the connective-tissue trabeculae found in the alveoli, had under- gone hyalin degeneration. Blood-vessels were scanty through- out the stroma. The adventitia of the vessel walls was somewhat thickened. The endothelial cells of the intima were markedly swollen and in places increased in number. Sweat-glands were present, and both the stroma and epithelium were normal in appearance. Epithelial Cells.-The epithelial cells were arranged in cylin- ders and oval masses, sections of which in places were much elongated, in others, round or oval (Plates I and II). The width of the alveoli on an average ran from 75 to 800 microns, the majority being about 500 microns. The cells in the alveoli were arranged for the most part in a number of small nests, each consisting of from twenty to thirty cells, concentrically placed. Towards the outside of an alveolus the cells were arranged in rows parallel to its boundary. This regular arrange- ment of cells was especially marked in the outermost row. It is worthy of note that in no instance was there an extension of the epithelial cells into the stroma itself, but they were massed together and surrounded by a firm band of connective tissue. The cells themselves were of large size, cuboidal, polyhedral, columnar, or even in places cylindrical in shape. The most regular and symmetrical measured from thirty to thirty-five micromilli- metres in diameter. The cell-body stained easily with eosin, taking on a fairly deep color, and appeared faintly granular. The nucleus was large and oval, occupying about one-third of the cell-body. It stained very readily with haematoxylin, showing many rather coarse, deeply stained granules. A nucleolus was seen in each nu- cleus. No mitotic figures were observed in the cells of any of the sections. Spine or so-called “prickle cells” could not be found. Cell Degeneration.—Scattered here and there, usually towards the centre of the alveoli, or else in the epithelial layers of the cyst walls, were many isolated cells which stained deep purple, both with haematoxylin-eosin and picro-acid-fuchsin (Plate III, A). 490 WILLIAM COGSWELL CLARKE. The cell-bodies, staining purple, were often elongated, either homo- geneous or vacuolated, with no nuclei showing; apparently mu- cous degeneration of the epithelial cells. Besides the above cells, Some alveoli contained an area whose cells were large, measuring forty-three micromillimetres in diameter, and vesicular in shape. The cell-body did not stain with either haematoxylin-eosin or picro-acid-fuchsin. It was Occupied by fine vacuoles, very regular in shape. Occasionally larger vacuoles were seen. The cell-wall was thin, sharply outlined, at times broken down as though the cells had ruptured. This condition was most often observed towards the centre of the alveoli. The nucleus was round and measured eight and six-tenths micromillimetres in diameter. It was situated in some near the centre of the cell, in others near the cell-wall. After staining, the color of the nucleus was a rather faint homogeneous blue, not granular (Plate IV, A). Cysts—Walls and Contents.-In addition to the seven cysts whose gross appearance has already been described, many others were noted in the microscopic examination, measuring from 160 to 64O microns in diameter. They were for the most part round, at times oval, or very much elongated. The walls of the cysts varied considerably. At times, when the cyst was situated in the centre of an alveolus, the limiting walls were sharp and clean-cut, and the epithelial cells were even and regular in shape, no degenerated or ruptured cells appearing (Plate II, D). Again, especially in the larger cysts, the walls were made up of fifteen to twenty-five layers of epithelial cells, the innermost of them being broken and disintegrated (Plate I and Plate II at A, A, A, A). Here and there appeared the cells mentioned above as undergoing mucous degeneration. Some of the small cysts also had uneven walls, a few of them bounded by the large, clear staining cells undergoing fatty metamorphosis (Plate IV, C), others by jagged walls in which were many cells undergoing mucous degeneration (Plate III, F, F). The contents of some of the smaller cysts, when not dissolved while fixing and hardening the specimen, stained a deep purple color, and invariably there were noted in the walls cells which were undergoing the so-called mucous degeneration (Plate III, B). In other cysts the material stained a deep pink, granular or vacuolated in appearance (Plate I, B). In this material, especially towards the edge, many cholesterin crystals together with groups of cells and isolated cells resembling those PLATE + - \'. £&&！！！！！!！！！！！！！！-|- ſ?|×· … - - Å##§§§º ·----|- | ºſ:|-··ſaeſº，~ ！！！--|- -№ſſº|-£ſº （ - ）。！！！！！！ …·ſae|-· ·-------|×---- ſº º ſae， ſºſae^，：，，--~~~~.| «…， |-ſaeſº， ·- … . . . . . .---- -|-|- -|×§.* （）|-|- |-|- ſae. ADENOMA OF SEBACEOUS GLANDS. 491 undergoing fatty metamorphosis were noted; also a few nuclei, which stained a pale blue, the cell-body apparently having dis- integrated. In addition to the above cysts, a few openings similar to gland ducts were seen. One opening branched twice, and these branches also divided, resembling a compound tubular gland (Plate V., A). This duct-like arrangement existed in several serial sections, dem- onstrating that the lumen was of considerable size. The walls, which were sharp and clean-cut in outline, consisted in every case of a number of layers of epithelial cells, not like the ragged walls of the cysts described above. The Skin covering the Tumor.—The epithelial layers were normal, except that at the summit of the tumor the papillae of the true skin had all been much shortened and the spaces between the papillae widened. Between the epithelial layers of the skin and the epithelium of the tumor there was a band of connective tissue, the pars papillaris and pars reticularis of the true skin. Diagnosis of Present Case.—An epithelial cystic growth of this nature, with cells undergoing fatty changes, could have Originated from any one of the following: (A) The glands in the wall of a dermoid cyst; (B) the epithelium of an in- clusion cyst; (C) a supernumerary mammary gland; (D) the Sweat-glands; or (E) the sebaceous glands of the normal skin. (A) The first of the above possibilities is ruled out by the fact that no hair-follicles or other remains of the wall of a dermoid cyst existed, such as stratified squamous epithelium, spine, or prickle cells. (B) The presence of compound gland ducts, the absence of stratified epithelium, and the large size of the epithelial cells, discredit the growth as having originated from the wall of an inclusion or epidermoid cyst; also, there are no glands found in the walls of inclusion cysts from which an adenoma might Spring. (C) The tumor could not have originated from a super- numerary mammary gland, since it did not have any of the characteristics of such a growth. 492 WILLIAM COGSWELL CLARKE. (D) The fourth possibility, that the tumor originated from a sweat-gland, can be easily ruled out because in the present tumor the epithelial cell-body did not stain as deeply as Sudoriparous cells; and the gland-ducts were compound and thick walled, whereas in tumors growing from sweat- glands the ducts do not usually branch and their walls are lined by only one or two layers of cells. Klauber * has recently described sweat-gland adenomata in which there were many Small cysts and gland tubules, whose walls consisted of a thin layer of epithelial cells. (E) Therefore, having eliminated the four preceding possibilities, the existence of a pathological development of the infundibula of Sebaceous glands remains for consideration. In this connection, attention is called to the deep situation of the tumor and its apparent separation from the skin; to the complex alveolar arrangement; to the formation of compound ducts; to the sparseness of ducts in a given section; to the formation of sebaceous material in one of the cysts, probably from cells in the wall, more like cells found in the normal seba- ceous glands undergoing fatty metamorphosis; to the absence of laminated material such as is found in cysts formed from cells, which ultimately would have produced pavement or strati- fied squamous epithelium; and to the large size of the cells in the alveoli similar to sebaceous cells, and not like cells of the epidermis, except the basal cells, which are similar in char- acter and continuous with the germ or basal cells of sebaceous glands. RÉSUME OF THE LITERATURE OF TRUE ADENOMATA OF SEBACEOUS GLANDS. Bock,” in 1880, gave an exceedingly good description of the condition. In speaking of adenomata of sebaceous glands, he said the literature was very scanty. He mentioned Schmidt's Jahrbücher, in which there is an article by Porta 8 on tumors of sebaceous follicles. The date of this article is 1856; in it there is given a list and the situation of 384 tumors; 238 were cystic, twenty-six of them being sebaceous. No detailed ADENOMA OF SEBACEOUS GLANDS. 493 microscopic description was given. Bock also mentions Förs- ter’s “Handbuch der Allgemeine pathologische Anatomie,” S. I79, as referring to hypertrophy of sweat and sebaceous glands in contradistinction to that of sebaceous gland tumors. Bock further quotes Rindfleisch, “Lehrbuch der pathologische Ge- weblehre,” S. 288, as describing a case of pure hypertrophy of a sebaceous gland tumor which had a broad base covered by the hairy scalp. The tumor was the size of a pigeon's egg, and its structure resembled that of a mammary gland. Birch Hirschfeld is also mentioned as citing the preceding case, and considers the true hypertrophy of sebaceous glands as probably very rare. Bock's own specimen was eight centimetres in length, six centimetres in breadth, and three and five-tenths centimetres in height. Cut sections showed many epithelial alveoli separated by connective-tissue bands. The skin was apparently normal, except in places the papillae were somewhat elongated. The lobules of the tumor had diameters of from one to six millimetres and were either round or elongated. In the gland substance were openings, microscopic in size, some cut Squarely across, others obliquely. In places he found the centre of the lobules filled with cells undergoing fatty degeneration, with fat drops and granular masses. Towards the centre of the alveoli the cells were larger, appearing always epidermic in character and dimmed with fat; these changes in the cells were the causes of the construction of fat drops and granular masses in many acini. The sweat-glands were normal and regular. He gives a drawing showing the general topography of the epithelial alveoli with the central openings and with the normal condition of the skin covering the tumor; also a drawing of one-half of an alveolus, showing the outside cells clear and regular in arrangement, with fat drops increasing in number towards the centre of the alveolus, which is filled with a homogeneous II13.SS. Eve,” in 1881, gave a good description of a partly cystic and partly adenomatous tumor which had become calcified. His plate certainly shows the lobules of an adenoma. The tumor measured one and one-quarter by one-half inches. It was seen to consist of columns, made up of epi- thelial cells, which presented every variety of size, from minute round masses of cells, one-fortieth of an inch in diameter, to large columns or cylinders one-half to three-quarters of an inch in diameter, having the appearance of alveoli in transverse section. In the sections the alveoli were cut across in various directions and were round or oval; the smaller columns were in places irregular, tortuous, branched, and occasionally uniting. The cells were small, round or oval, closely packed together with- out intercellular substance. Occupying a position near the centre of many 494 WILLIAM COGSWELL CLARKE. of the larger columns, and in a few instances of the smaller, unstained granular masses were observed, which were probably the seats of the cal- careous matter. This material was found, on further examination, to consist largely of fatty matter containing crystals of cholesterin, and may be regarded as an attempt at the formation of sebum by the gland tissue. Eve thinks it may be regarded as an adenoma associated with the sebaceous glands. Krauss,” in 1884, in an article concerning giant cells in epithelial growths, mentions an adenoma of sebaceous glands, a tumor, 2.9 centi- metres by 2.5 by I.7 centimetres, on the left arm of a man twenty years of age. This fairly solid tumor was soft and contained a doughy mass in the centre. The skin was quite thin and separate from the growth, a firm connective-tissue capsule enclosing it. The epithelium was arranged in lobules, round or oval in shape, at times tubular. The centres of the lobules and tubules were filled with a firm necrotic mass from degenerating cells. A cut is given showing several alveoli with centres filled with the necrotic masses. He mentions sweat-glands as if they had no connection with the turnOr. Poncet,” in 1890, described a case of a man of fifty-three years of age, who, for thirty years or more, had over sixty tumors on his head alone. A cut is given showing a man's head covered with large-sized, closely set tumors; a very prominent one is seen on his forehead. Microscopic ex- amination of one of these tumors by Band showed it to be made up of round lobules separated by connective-tissue bands. The lobules are composed of epithelial cells. The cells have the same characteristics as young sebaceous gland cells. The peripheral cells of the lobules are in regular rows. In the centre of most of the lobules there is an unstained part, light gray or colorless; this part looks like a sebaceous mass. He concludes that it is a tumor made up of epithelial cells similar to sebaceous gland cells. Jarisch " gives a few cases of skin tumors, with many good cuts, and also reviews the literature of sebaceous adenoma. Richard Barlow,” in 1895, reviewed the whole subject thoroughly. He describes a case of multiple tumors of the Scalp, Of six years’ duration, in a person sixty years of age, and calls it a case of adenomata of the sebaceous glands. He speaks of the definition of adenoma, and quotes Förster, Cohn- heim, Ziegler, and others. He also speaks of the different types of lesions of these glands. In conclusion, he says an adenoma may occur in sebaceous glands, may become calcified, and also may become cancerous. He gives several plates of adenoma of sebaceous glands of the scalp. Curtis and Lambret,” in 1900, review the subject up to ADENOMA OF SEBACEOUS GLANDS. 495 date, and describe a large tumor of eight years' growth, which clinically was thought to be malignant; histologically it was thought to be a simple adenoma of the sebaceous glands and similar to Bock's, Monti's, and others. Borst "" in his book on tumors shows a picture of a Sebaceous adenoma which was made up of lobules of epithelial cells separated by bands of connective tissue. The lobules are very irregular in shape, their centres occupied by masses of Sebaceous material. He says that a pure adenoma of sebaceous glands, either a solitary, nodular, fleshy, or sometimes ulcer- ated, encapsulated tumor, may exist, which is made up of epi- thelial lobules, the centres of which are filled with fatty masses. The connective tissue and vessel walls may undergo hyaline degeneration; also, in an adenoma the peripheral germ cell layers, which shut in or enclose the epithelial cells of the fol- licle, spread out, while the old central layers of cells are mostly overwhelmed by fat infiltration, and, in consequence, they are not of a typical form, so that the enclosed fat mass in the centre grows from the rupture of these cell-bodies. Krompecher,” in 1903, mentions the relation that may exist between the basal cells of hair-follicles, or of sweat or Sebaceous glands, and flat-celled cancers, also the similarity of . structure of sebaceous glands to breast glands; and gives a diagram of different types of sebaceous gland hypertrophy, showing how different types of tumors may develop from the different epithelial layers of the skin. Levings,” in 1903, speaks of adenomata of sebaceous and sudoriparous glands. “These growths are of rare occurrence. When taking origin from the sebaceous glands they are closely adherent to the skin, are round or slightly lobulated, elastic, and reasonably hard growths. They are made up of an aggregation of alveoli, and are to be differentiated from the ordinary wen and from the tumors which are entirely subcutaneous. In size they are usually small, often not exceeding that of a hickory- nut, but in exceptional cases they have become quite large.” Ribbert 4° mentions adenoma of the sebaceous glands and speaks of Barlow's article. 496 WILLIAM COGSWELL CLARKE. Unna’s “Review of the Relation of Adenoma and Hy- perplasia of the Sebaceous Glands.-Under hypertrophy of the Sebaceous glands, Unna says that simple hypertrophy of the Sebaceous glands appears partly as an accompaniment of other processes, partly as an independent tumor. The independent tumors are rare; much more frequent is the more or less diffuse hypertrophy of the glands, which chiefly accompanies the chronic forms and results of seborrhoeic catarrh, especially of the nose and the middle of the forehead. Adenoma of sebaceous glands, Unna calls “Steatade- noma.” “The definition of adenoma can certainly not be ap- plied, where the sebaceous gland epithelium produces solid bodies, which grow into solid epithelial masses or epithelial cords; we have, then, some form of acanthoma proceeding from the sebaceous gland. The participation of the sebaceous gland in an epithelial growth, proceeding from the follicle or the surface epithelium, is a very frequent occurrence accom- panying the most varied inflammatory and hypertrophic con- ditions.” “It thus appears as if we must abandon the defini- tion of steatadenoma in the strict sense of the word, since the benignant tumors of the sebaceous glands are either hyper- trophies or acanthomata of the sebaceous glands. But there might also be tumors occupying a median position between the condition just mentioned, in that, for example, solid bodies first come from the sebaceous epithelium, which, after they have developed in a tumor-like fashion, again recall their point of origin by fatty degeneration of their cells. Then we would have a growth in a new direction agreeing with the definition of an adenoma, if we retain it, and at the same time the proof that the cells of the growth had not lost their character as sebaceous cells during growth. If no such tumors were to be found, then it would be better to abandon altogether the term ‘adenoma of the sebaceous glands; but there is the case recorded by Bock which agrees with the following strict defi- nition of steatadenoma, benign tumor-like growths of irregular formation proceeding from the epithelium of the sebaceous glands, in whose outgrowths fatty but no colloid metamor- ADENOMA OF SEBACEOUS GLANDS. 497 phosis takes place.” Unna quotes from Bock, and says, “The development of the glandular lobules took place from below upward. Bock founded the diagnosis steatadenoma on the triple recognition, first, of the normal behavior of the sweat- glands; second, of the absence of cancerous processes, and, third, of the presence of acini, which in their development, evolution, fatty degeneration, and calcification corresponded with sebaceous glands.” In conclusion, he says that one must admit that this origin would have been still clearer, if, at the more recent parts of the growth, a true formation of sebaceous glands had been recognized, with a reticular spongy character of the cell protoplasm, as in the fatty changes in true sebaceous cells. RÉSUME OF THE LITERATURE OF HYPERPLASIA OF SE- BACEOUS GLANDS, EITHER ALONE OR SECONDARY TO HYPERTROPHY OF THE SKIN. Pringle,” in 1890, reported a case of many small, indolent, firm, whitish, solid papules or little tumors of the face embedded in the skin, the largest the size of a small pea and near the nasolabial fold. When pricked with a pin, white sebaceous matter could be expressed. The superficial layers of epithelium along the whole section were rather thin ; the rete showed excessive and irregular involution, dipping deeply into the subjacent derma. The chief pathological changes were found in the corium; its upper papillary layer was enormously hypertrophied, but it was in the deeper layers that the essential lesions lay. They consisted of an enormous increase in number and complexity of the sebaceous glands. The gland epithelium was everywhere well formed, the acini at no spot exhibiting retrogressive changes. A good microphotograph was given; also a chrome lithograph of a girl's face studded around the nose and on the cheeks below the eyes with small flat tumors. He cites Balzer's “cases, and agrees with him in using the term “adenomata.” Caspary,” who followed Pringle in 1891 with a very similar case, described a girl’s face covered with small nodules. He showed a good lithograph and microscopic sections of the conditions. He gave Pringle's as the sixth and his as the seventh case on record. Pollitzer,” in 1893, described and gave a clear photograph of a group of thirty small adenomata, as he calls them, on the forehead of a man twenty-five years of age, in whom the condition had lasted six years. Each nodule was as large as a medium-sized barley-corn. Two were cystic and sebaceous matter could be expressed from almost all. The microscopic examination of excised specimens showed that almost the entire specimen was made up of sebacous gland tissue, which had preserved the lobulated 498 WILLIAM COGSWELL CLARKE. arrangement of the normal gland. The resemblance in both the appear- ance of the cells and their arrangement to normal sebaceous glands was so close that he suspected it to be hypertrophy of the gland. Actual measurement of the individual gland cell showed no increase in size beyond the normal, and there were far more than the normal number of lobules. The condition, then, could only be called adenoma. He gives a cut, and says, “Two club-shaped processes of epithelium may be seen shooting out from the side of the root-sheath of a hair, exactly like the buds from which the glands are formed in the embryo. Near the end of one of them a small island of epithelial cells appears as if cut off from the growing shoot. Such an isolated group of cells would, as they under- went their physiological fatty degeneration, necessarily give rise to a sebaceous cyst. Sweat-glands appear normal and far below the tumor.” He said there were scarcely twenty cases of this rare disease observed, and this was the first and only one in America. He mentioned Crocker’s, Jameson's, and Caspary's cases. Unna says that if any one attempted to apply the term adenoma to peculiar forms of simple sebaceous hypertrophy in such cases as those of Pringle and Caspary, it would be im- possible to draw the line, and we should have to call every enlargement of the sebaceous glands adenoma. Pezzoli,” in 1900, described a very similar condition, and mentioned that Balzer and Ménétrier were the first to de- scribe it in 1885. He gave a case of a great number of very small discrete tumors on the face of a young girl, showing a picture of the same, also a cut of a microscopic section. Walther Pick,” in 1901, described with great care and detail a diffuse mass of small cystic skin tumors on the face of a man suffering from acne rosacea. The cysts were filled with a homogeneous material, and some of the cells were degen- erating; also, some of the cell-bodies were much swollen, and at times fine protoplasmic scaffolding was seen which is found in normal sebaceous gland cells. The nature of the degenera- tion he attempted to prove by Orcein stain and Orange tannin. In normal sebaceous gland ducts, the dye stained the material a light yellow, and the cyst contents the same color. He also gave several very good lithographs of a diffuse hypertrophy or adenomatous condition of the sebaceous gland and hyper- trophy of the skin, with no real tumor formation. ADENOMA OF SEBACEOUS GLANDS. 499 SUMMARY. It is seen from a consideration of the cases cited by Bock and others, that, although denied by some observers, true ade- nomata of the sebaceous glands, without associated lesions of the skin, do exist, and that these adenomata undergo fatty metamorphosis with a formation of cysts. Tumors of this sort are rare, large ones especially so. On the other hand, adenoma, or an hyperplastic condition of the gland, secondary to or together with hypertrophy of the skin, is not so uncom- 111O11. True adenomata of the sebaceous glands may become calcified or carcinomatous. The stroma of the tumor may undergo hyaline degeneration. Giant cells may occur in them. The epithelial cells may undergo mucous degeneration and form cysts similar to those derived from the cell under- going fat metamorphosis. s It has been shown that the tumor which is the subject of this paper is neither a dermoid cyst nor a simple inclusion cyst, nor is it derived from a supernumerary mammary gland, or the sweat-glands. It would therefore appear that the growth is a true adenoma starting from the infundibula of the sebaceous glands and retaining the type of the secreting portion. It thus has all the characteristics of a steatadenoma or adenoma of the sebaceous glands; namely, the presence of acini, which in their development, evolution, fatty metamor- phosis, and calcification correspond to Sebaceous glands. I am indebted to Professor T. M. Prudden, Dr. John H. Larkin, and Dr. Charles Norris for many suggestions in connection with this article. BIBLIOGRAPHY. * Klauber. Beiträge zur klinischen Chirurgie, Band xli., S. 31.I. * Bock. Virchow's Arch., Band lxxxi, S. 503. * Porta. Schmidt’s Jährbuch, Band xcvi, S. I27. * Eve. Transactions of the Pathological Society, London, Vol. xxxiii, p. 335. * Krauss. Virch. Arch., Band xcv, S. 249. * Poncet. Revue de Chirurgie, I890, p. 244. 5OO WILLIAM COGSWELL CLARKE. ‘Jarisch. Arch. f. Dermat, u. Syph., Band xxviii, S. 163. * Barlow. Deut. Arch. f. klin. Med., Band lv, 61. - * Curtis and Lambret. Revue de Chirurgie, Vol. xxii, p. 147. * Borst, Geschwulstlehre, Band ii, S. 577. * Krompecher. Der Bazalzellenkrebs. * Levings. The AEtiology, Pathology, Diagnosis, and Treatment of Tumors, p. 405. * Ribbert. Geschwulstlehre, S. 384. *Unna. The Histopathology of the Diseases of the Skin, p. 814. * Pringle. British Journal of Dermatology, 1890, Vol. ii, p. 1. * Balzer et Ménétrier. Arch. de Physiologie, 1885, p. 564. * Caspary. Arch. f. Derm. u. Syph., Band xxiii, S. 371. * Pollitzer. Journal of Cutaneous and Genito-urinary Diseases, 1893, Vol. ii, p. 475. * Pezzoli. Arch. f. Derm. u. Syph., Band liv, S. 193. * Pick. Arch. f. Derm. u. Syph., Band lyiii. CONTENTS OF VOLUME IX. A Study of a Bacillus Resembling the Bacillus of Shiga, from a Case of Fatal Diarrhea in a Child; with Remarks on the Recognition of Dysentery, Typhoid, and Allied Bacilli. By PHILIP HANson HIss, Jr., M.D., of New York, Instructor in Bacteriology and Hygiene, College of Physicians and Surgeons, Columbia University, New York; and F. F. RUSSELL, M.D., Assistant Surgeon, U. S. Army. Experiments Relating to the Question of Fixation of Strychnine in Animal Tissues. By S. J. MELTZER and G. LANGMANN. A Resumé of Some Recent Researches Relating to Cytolysis and Immunity. By T. MITCHELL PRUDDEN, M.D., New York, * A Case of Multiple Fibroma (Fibro-Neuroma) of the Nerves of the Lower Extremities, with Diffuse Enlargement of the Sciatics—Complicating Sarcoma and Metastases in the Lungs. By JOHN H. LARKIN, M.D., Tutor in Pathology, College of Physicians and Surgeons, Columbia University, New York. The Prognostic Value of the Diago-Reaction in Pulmonary Tuberculosis. By FRANCIS CARTER WooD, M.D., of New York; Instructor in Clinical Pathology, College of Phy- 'sicians and Surgeons, Columbia University; Pathologist to St. Luke's Hospital. A Simple and Rapid Chromatin Stain for the Malarial Para- site. By FRANCIS CARTER WooD, M.D., of New York; Instructor in Clinical Pathology, College of Physicians and Surgeons. The Mixed Tumors of the Salivary Glands. By FRANCIS CARTER WooD, M.D., of New York; Instructor in Clinical Pathology, College of Physicians and Surgeons, Columbia University; Pathologist to St. Luke's Hospital. CONTENTS. I O. II. I 2. I3. I4. I5. I6. The Agglutination of the Pneumococcus with Certain Normal and Immune Sera. By AUGUSTUs WADsworth, M.D., Alumni Fellow in Pathology; Assistant in Bacteriology and Hygiene, College of Physicians and Surgeons, Colum- bia University, New York. Experimental Studies on the Etiology of Acute Pneumonitis. By AUGUSTUs WADsworth, M.D., Alumni Fellow in Pathology; Assistant in Bacteriology and Hygiene, De- partment of Pathology, College of Physicians and Sur- geons, Columbia University, New York. A Simple Method of Obtaining Blood from the Rabbit. By AUGUSTUS WADSWORTH, M.D. A Convenient Method for Growing and Storing Virulent Pneu- mococci. By AUGUSTUs WADsworth, M.D. Haemolymph Nodes. By HUGHEs DAYTON, M.D., Alumni Fellow in Pathology; College of Physicians and Surgeons, Columbia University. A Bacteriological Study of Trachoma, with Remarks on the Occurrence of the Influenza Group of Bacteria in Conjuncti. vitis. By ARNOLD KNAPP, M.D. The Bacterial Precipitins. By CHARLEs NoFRIs, M.D., In- structor in Bacteriology and Hygiene, College of Phy- sicians and Surgeons, Columbia University; Former Fel- low of the Alumni Association of the College of Physicians and Surgeons, New York. Hemorrhages in the Eye Present at Birth. By EDWARD B. CoBURN, M.D. An Adenoma of Sebaceous Glands of the Abdominal Wall. With a Review of the Literature of Adenoma and Hyper- plasia of the Sebaceous Glands. By WILLIAM CoGSWELL CLARKE, M.D., of New York. CONTENTS OF PREVIOUS volumES OF THE STUDIES FROM 2 7à3. .4.: : . . gº Y O, 3. 4e 6 tº 7. THE DEPARTMENT OF PATHOLOGY. VOLUME I . A Contribution to the Pathology of the Laryngeal and other Crises in Tabes Dorsalis. Alºgº Paralysis. By IRA. T. VAN GIEson, M.D., First Assistant in Normal His- tolo - gy. . A Case of Spina Bifida with º Spinal Meningitis and Ependymitis due to Bac- teria. By L. EMMEtt Holt, M.D., and IRA. T. VANGIEson, M.D . Observations on Prurigo. By R. W. Taylor, M.D., and IRA. T. VAN Gieson, M.D. The Anatomy and Physiology of the Faucial Tonsils with Reference to the Absorption of Infectious Material: and . Actinomycosis of the Lungs. By EUGENE HopBNPyL, M.D., First Assistant in Pathology. . Studies on the AEtiology of Diphtheria. Second Series. By T. Mitch ELL PRUDDEN, M.D., Director of the Laboratory. . Studies on the Action of Dead Bacteria in the Living Body. By T. Mitch ELL PRUDDEN, M.D., and Eugene HopenPyL, M.D. VOLUME II. . A Study of Experimental Pneumonitis in the Rabbit. The Element of Contagion in Tuberculosis. By T. Mitch ELL PRUDDEN, M.D. . A New Material for Models. By GeoRGE C. FREEBORN, M.D., Instructor in Normal Histology. . A Case of Traumatic Myelitis in the Lower Dorsal Region. By IRA VAN GIEson, M.D., First Assistant in Normal Histology. . A Report of the Gross and Microscopical ºrigions in Six Cases of Death by Strong Electrical Currents. By IRA VAN GIEson, M. . A Study of Typhus Fever—Clinical, Pathological, and Bacteriological. By JoHN WIN- TERS BRANNAN, M.D., Assistant Pathologist to St. Francis Hospital, and TIMOTHY MATLACK CHEESMAN, M.D., Assistant in Bacteriology. . On the Sterilization of Milk at Low Temperature. . By Rowl.AND GopFREY FREEMAN, M.D., Assistant. Physician to Roosevelt Hospital, Out-Patient Department; Pa- thologist to St. Mary’s Hospital. * . Diphtheria and Allied Pseudo-Membranous Iºftºpºl. A Clinical and Bacterio- *:::g Study. By WILLIAM HALLock PARK, M.D.: an On the Poisonous Products of the Tubercle Bacillus. By T. Mitch ELL PRUDDEN, M.D. . A Study of the Artefacts of the Nervous System. By IRA VAN GIEson, M.D., First Assist- ant in Normal Histology. VOLUME III. . On the Importance of Vibration to Cell Life. . By S.J. MELTzER, M.D. . Diphtheria and other Pseudo-Membranous Inflammations. Second Paper. By WILLIAM ALLOCK PARK, M.D., Alumni Association Fellow in Pathology. . Report of a Recent Sanitary Inspection of one of the Sources of the Croton Water Supply. By TIMOTHY MATLACK CHEESMAN, M.D., Assistant in Bacteriology. . Observations on a Case of Recurrent Amaebic Dysentery with Successive Large Hepatic Abscesses. By John WINTERs BRANNAN, M.D. . A Contribution to the Pathology of Traumatic Epilepsy. By IRA VAN GIEson, M.D. Assistant in Histology. * = * * g Sterilization of Milk at 75° C. (Pasteurization), and its Efficiency in Destroying Patho- genic Organisms. By Rowland GoDFREY FREEMAN, M.D . A §tudy on the AEtiology of Exudative Pleuritis. By T. Mitchell PRudden, M.D., Pro- fessor of Pathology. . A Study of the Leucocytosis of Lobar Pneumonia. By JAMEs Ewing, M.D., Assistant in Histology. . On the %gy of Appendicitis. By EUGENE HopenPyL, M.D., First Assistant in Pa- #; & A Consideration of Artesian Well and Surface Waters from the Standpoint of Bacteriology and Public Health. By TIMothy MAtlack CHEEsMAN, M.A. VOLUME IV. . Concurrent Infections and the Formation of Cavities in Acute Pulmonary Tuberculosis. An Experimental Study: and . A Case of Pyamia Following the Incision of an Urethral Stricture, with a Note of Certain Biological Peculiarities in the Staphylococcus Pyogenes Aureus. By T. Mitch ELL PRUDDEN, Professor of Pathology, College of Physicians and Surgeons, Columbia Uni- versity, New York. . Van Gieson's Picro-Acid Fuchsin as a Selective Stain for Connective Tissue. By G. C. $ºoks, Instructor in Normal Histology, College of Physicians and Surgeons, New OTK. On the New Relations of Pathology and Practical Medicine as Bearing upon the Patho- logical Departments of our Hospitals: an he Dominant Influence in the Medical Progress of the Nineteenth Century. By John SLADE, ELY, M.D., Assistant in Pathology in the College of Physicians and Surgeons; Pathologist to Bellevue Hospital. Some Considerations on Different Types of Exudative Inflammation. A Study Based on Bacterial Examinations from One Hundred and Thirty-five Surgical Cases. By CHARLEs N. DQWD, M.D., Assistant Surgeon to the New York Cancer Hospital. Pasteurized Milk as Supplied to the Poor by the Straus Milk Depot of New York. By Rowland GoDFREY, FREEMAN, M.D., Pathologist to the Foundling Hospital; Pa- thologist to St. Mary's Free Hospital for Children; Assistant Physician to Roosevelt Hospital, Out-Patient Department. I I. . On Certain Bacteria from the Air of New York City. By HARRIson G. DYAR, A.M. I 3. I4. I 5. I 6. I O. I I . I 2. I 3. I4. I S. I 6. I 7. I 8. I Q. 2 O. Acute Pneumonia in Childhood. By Thomas S. Southworth, M.D., Pathologist to the Nursery and Child's Hospital; Assistant Medical Department Vanderbilt Clinic; In- structor in Diseases of Children, New York Polyclinic. . Toxic Hypoleucocytosis. By JAMEs Ewing, M.D., Tutor in Histology, College of Phy- sicians and Surgeons; Assistant Pathologist to Roosevelt Hospital, New York. he Leucocytosis of Diphtheria under the Influence of Serum Therapy: By JAMES EWING, M.D., Tutor in Histology, College of Physicians and Surgeons; Assistant Pathologist to Roosevelt Hospital, New York. Notes on the Preparation of Diphtheria Antitoxine. By CHARLEs B. FITZPATRICK, M.D. Observations on an Uncommon Form of Cutaneous Tuberculosis. ... By CHARLES C. RAN: som, M.D., Clinical Assistant to the Chair of º College of Physicians and Surgeons; Visiting Physician to the City Hospital, New York; and IRA VANGIESQN, M.D., Instructor in Normal Histology, College of Physicians and *. Visiting Neurologist, Randall's Island Hospital; Curator to the City Hospital, New York. Notes on the Fixation of Nerve Fibres by Formalin. By EDw1N M. KITCHELL, M.D., Assistant in Normal Histology, College of Physicians and Surgeons, New York. . Malignant Adenoma of the Uterus. By WILLIAM S. Stone, M.D., Alumni Fellow in Pathology, College of Physicians and Surgeons; Assistant in the Department of Dis- eases of Women, Vanderbilt Clinic, New York. $ A Clinical and Bacteriological Study of the Gonococcus (Neisser), as Found in the Male Urethra and in the Vulgo-Vaginal Tract of Children. By HENRY HEIMAN, M.D., At- tending Physician to the Children's Department of the Mt. Sinai Hospital Dispensary, New York. - - VOLUME v.–PART I. Milk as an Agency in the Conveyance of Disease; and . Low Temperature Pasteurization of Milk at about 68°C. (155° F.). By Rowland GoD- FREY FREEMAN, M.D., Pathologist to the Foundling Hospital; Pathologist to St. Mary's Free Hospital for Children, New York. . A Machine for Tubing Bacterial Media. By HARRIson G. DYAR, A.M., Ph.D., Assistant § Bacteriology, Department of Pathology, College of Physicians and Surgeons, New ork. . A Contribution to the Study of Acute Ascending (Landry's) Paralysis. By PEARce BAILEY, M.D., A.M., Assistant in Neurology, Vanderbilt Clinic; Attending Physician to the Almshouse, Workhouse, and Incurables' Hospital, New York; and JAMes Ew ING, A.M., M.D., Tutor in Histology, College of Physicians and Surgeons, Colum- bia University, New York; Assistant Pathologist to Roosevelt Hospital. . . A Further Study of the Biology of the Gonococcus (Neisser), with Contributions to the Tech- nique. A Paper Based on the Morphological and Biological Examingtion of Exudates tn Cases of Chronic Urethritis. . By HENRY HEIMAN, M.D., New York, Attending Physician to the Out-Door Children's Department of Mount Sinai Hospital. . Bacterial and Allied Tests as Applied to the Clinical Diagnosis of Typhoid Fever. By John SLADE ELY, M.D., Professor of Histology and Pathological Anatomy in the Woman's Medical College of the New York Infirmary; Assistant in Pathology in the College of Physicians and Surgeons, New York; Pathologist to Bellevue Hospital. . Dangers of the Domestic Use, other than Drinking, of Contaminated Water, with Special Reference to Milk and Oysters as Carriers of Bacteria. By Row LAND GODFREY FREE- MAN, M.D., Pathologist to the Foundling Hospital; Pathologist to St. Mary's Free Hospital for Children, New York. . Report upon Two Cases of Tumor of the Spinal Cord:Unaccompanied by Severe Pain. By PEARCE BAILEY, M.D., Assistant in Neurology, Vanderbilt Clinic; Attending Phy- sician to the Almshouse, Workhouse, and Incurables' Hospital, New York City, . The Lymphatic Constitution and its Relation to Some Forms of Sudden Death. By JAMEs Ewing, A.M., M.D., Clark Fellow in Pathology, Columbia University; Assistant Fºlºgist to the Roosevelt Hospital; Pathologist to the Sloane Maternity Hos- pital, etc. Notes on Polychromatic Photo-Micrography, with Plate in Color. By EDw ARD LEAMING, M.D., Instructor in Photography. VOLUME. V.-PART II. Some Laboratory Moulds. By SMITH ELY JELLIFFE, A.B., M.D. On a Method of Isolating and Identifying Bacillus Typhosus, Based on a Study of Bacillus Typhosus and Members of the Colon Group in Semi-Solid Culture Media. By PHILIP HANSON HISS, Jr., M.D., Assistant in Bacteriology, College of Physicians and Sur- geons, Columbia University; Assistant Bacteriologist to the New York City Health Department. The Bactericidal Action of Lymph Taken from the Thoracic Duct of the Dog. By S. J. MELtzER, M.D., and CHARLEs Norris, M.D. On the Occurrence of Typhoid Fever without Characteristic Lesions of the Smaller Intestine: and - A Modification of Cullen's Method of Preparing Fresh Sections for Microscopic Work. By EUGENE HODENPyL, M.D., Instructor in Pathology, College of Physicians and Surgeons, Columbia University, and Pathologist to the Roosevelt and St. Francis Hospitals, New York. Further Studies (third series) on the Gonococcus (Neisser). By HENRY HEIMAN, M.D. A Case of Acute Leukaemia. By W. H. THQMPSQN, M.D., .D., and JAMEs Ewing, M.D., Assistant Pathologist to Roosevelt Hospital. Suture of the Cornea after Removal of the Lens. By W. H. BATEs, M.D. Studies of Ganglion Cells. A Preliminary Communication. By JAMEs Ewing, M.D., ºJºw in Pathology, College of Physicians and Surgeons, Columbia University. GW YOrk On#5 Therapeutic Value of Blood-Letting. An Experimental Study. By Isaac Levin, volume VI." 3. A Case of Apparent Absence of the Spleen, with General Compensatory Lymphatic Hyper- plasia. By EUGENE HopenPyL, M.D., Instructor in Pathology, College of Phy- sicians and Surgeons, Columbia University, New York; Pathologist to Roosevelt and St. Francis Hospitals. . An Experimental Study of the Direct Inoculation of Bacteria into the Spleen of Living Ani- mals, and q Contribution to the Knowledge of the Importance of a Lesion in Animal Tissue for the Lodgment and Multiplication of Bacteria within it. By T. M. CHEEsmAN, M.D., Instructor in Bacteriology, College of Physicians and Surgeons, Columbia Uni- versity, New York, and S. J. MELTzER, M.D . On the Influence of Fasting upon the Bactericidal Action of the Blood. By S. J. MELtzER, M.D., and CHARLEs Norris, Ph.B., M.D., Tutor in Pathology, College of Physicians and Surgeons, Columbia University, New York. . A Report on Six Cases in which the Bacillus Aérogenes Capsulatus was Isolated. By CHARLES NORRIS, Ph.B., M.D., Tutor in Pathology, College of Physicians and Sur- geons, Columbia University, New York. . Haemosporidia in American Reptiles and Batrachians. By Gustave LANGMANN, M.D. . Pathological Report of a Case of Akromegaly, with Especial Reference to the Lesions in the : I 2. I 3. Hypophysis Cerebri and in the Thyroid Gland, and a Case of Hemorrhage into the Pitui- tary. By PEARCE BAILEY, M.D., Consulting Neurologist to St. Luke's Hospital; At- tending Physician to the Hospital for Incurables, New York City. . Hemiatrophy of the Brai:1 without Disturbances of Intelligence or Personality. By PEARCE BAILEY, M.D., Physician to the Hospital for Incurables; Consulting Neurologist to St. Luke's Hospital; Assistant in Neurology, Columbia University, New York. . Preliminary Report on the Results of Blood Examinations at Camp Wickoff, August and September, 1898; and . Studies on Ganglion Čeils. By JAMEs Ew ING, A.M., M.D., Clark Fellow in Pathology and Instructor in Clinical Microscopy, College of Physicians and Surgeons, Columbia University, New York. VOLUME VII. . Miliary Tuberculosis of the Pleura without Other Tuberculous Involvement of the Lung. By EUGENE HopBNPyL, M.D., Instructor in Pathology, College of Physicians and Surgeons, Columbia University, New York; Pathologist to Roosevelt Hospital. . A Study of Lesions of the Liver in Young Children. By Rowland GoDPREY FREEMAN, M.D., Attending Physician to the New York Foundling Hospital and the Sea-Side Hospital of St. John's Guild. . Report of a Case of Alcoholic Multiple Neuritis, with Autopsy. By J. H. LARKIN, M.D., Assistant in Pathology, College of Physicians and Surgeons, Columbia University, and SMITH ELY JELLIFFE, M.D., Ph.D., Clinical Assistant in the Department of Ner- yous Diseases, Columbia, University. . Micro-organisms in the Air of Public Buildings and Conveyances, Due to Improper Methods of Cleaning. By ELMER. W. FIRTH, C.E., A.M., Fellow in Sanitary Engineer- ing, Columbia University. . Secondary Cataract. An Experimental Study. . By W. H. BATEs, M.D. . Qn the Histology and Etiology of Posterior Lenticonus. By R. DENIG, M.D., New York. . Poisonous Snakes and Snake Poison. By GUSTAv LANGMANN, M.D., New York. . Tetanus.--A Study of the Nature, Excitant, Lesions, Symptomatology, and Treatment of the Disease, with a critical Summary of the Results of Serum Therapy. By ALExis V. MoscHow ITz, M.D., Adjunct Attending Surgeon, Mt. Sinai Hospital, New York. . Traumatic Hemorrhages into the Spinal Cord. By PEARCE BAILEy, M.D., Consulting Neurologist to the St. Luke's, Orthopaedic, and Babies' Hospitals, New York, and to St. John's Hospital, Yonkers; Assistant in Neurology, College of Physicians and Surgeons, Columbia University. . An Experimental Contribution to the Knowledge of the Toxicology of Potassium Chlorate. By S. J. MELTzER, M.D., New York. . Serum-Globulin and Diphtheric Antitoxin.—A Comparative Study of the Amount of Globulin in Normal and Antitoxic Sera, and the Relation of the Globulins to the Anti- toxic Bodies. By PHILIP HANson Hiss, Jr., M.D., Instructor in Bacteriology, Col- lege of Physicians and Surgeons, Columbia University, and JAMES P. ATKINson, M.S., Assistant Chemist, Research Laboratory, Department of Health, New York City. Primary Splenomegaly—Endothelial Hyperplasia of the Spleen—Two Cases in Children –Autopsy and Morphological Examination in One. . By DAvro Bov AIRD, Jr., M.D., of New York, Pathologist to the Foundling Hospital; Clinical Instructor in Diseases of Children, University and Bellevue Hospital Medical College. Two Cases of Necrotic Broncho-Pneumonia with Streptothrix. By CHARLEs Norris, M.D., and JoHN H. LARKIN, M.D . Is Living Animal Tissue Capable of Neutralizing the Effects of Strychnine and Venom 2– An Experimental Study. By S. J. MELTZER, M.D., and G. LANG MANN, M.D., of New OTK. VOLUME VIII. . Studies on the Morphology of Ganglion Cells in the Rabbit. By FREDERIck RANDoDPH BAILEY, A.M., M.D., Tutor in the Normal and Pathological Histology of the Nervous System; Alumni Association Fellow in Pathology, College of Physicians and Sur- géons, Columbia University, New York. . On Puerperal Infection: With Special Reference to Douchingºnd the Practical Value of Bacterial Examinations. By AUGUSTUs WADsworth, . Haemolysis. By S. J. MELTzER, M.D. . Some Experimental Data on , the Significance of Concentration and of Multiplicity of 4rea in Hypodermic Infections. y S. J. MELT7 ER, M.D . Cell Proliferation under Pathological Conditions with Special Reference to the Etiology of Tumors. By Is AAc LEVIN. * Owing to a mistake of the binder, this volume was designated on the back of the cover as Vol. VI., Part I. I O. I I. I 2. I 3. I4. I 6. I 7. 18. . A Study of a Case of Gonorrhaea . Suppurative Pylephlebitis Associated with Anaerobic Micro-Organisms. By CHARLEs Norris, M. . The Bacillus Colt Communis in Human Infections. By AUGUST JERomE LARTIGAU, M.D., Tutor in Pathology, College of Physicians and Surgeons, Columbia University, New York . A Study of Čhronic Hyperplastic Tuberculosis of the Intestine, with Report of a Case. By AUGUST JEROME LARTIGAU, M.D., Tutor in Pathology, College of Physicians and Surgeons, Columbia ; New York. b * * * Ulcerative Endocarditis, with Cultivation of the Gono- coccus. By AUGUST JEROME LARTIGAU, M.D., Tutor in Pathology, College of Phy- sicians and Surgeons, Columbia University; Pathologist to the Sloane Maternity Hospital; Assistant Pathologist to the Roosevelt Hospital. Primary Sarcoma of the Thyroid Gland. By AUGUST_JEROME LARTIGAU, M.D., Tutor in Pathology, College of Physicians and Surgeons, Columbia University, New York. Multiple Leiomyomata of the Kidney. A Study of Two Cases. By AUGUST JEROME LARTIGAU, M.D., and JoHN H. LARKIN, M.D., Tutors in Pathology, College of Phy- sicians and Surgeons, Columbia ...}} New York. A Study of Hyperplasia of the Pharyngeal Lymphoid Tissue (Adenoids) with especial Reference to Primary Tuberculosis of the Pharyngeal Tonsil. By AUGUST JEROME LARTIGAU, M.D., Tutor in Pathology, College of Physicians and Surgeons, Columbia University; Pathologist to the Sloane Maternity Hospital; and MATTHIAS NICOLL, Jr., M.D., Pathologist to New York Infant Asylum and Founding Asylum; Attend- ing Physician to Seaton Hospital for Consumptives. The Influence of the Spleen on Natural or Acquired Hemolytic Properties of Blood Serum. By ISAAc LEVIN, M.D. The Detection of Typhoid Bacilli in the Feces as a Diagnostic Test of Typhoid Fever and a Comparison of this Test with the Widal Reaction. By HENRY A. HIGLEY, M.D., of Brooklyn, N. Y. . Glaucoma—An Experimental Study. . By EDwARp B. CoBURN, A.M., M.D., New York, Clinical Assistant, Vanderbilt Clinic, College of Physicians and Surgeons; Assistant Surgeon and Pathologist, New Amsterdam Eye and Ear Hospital. - Studies in the Bacteriology of Typhoid Fever, with Special Reference to its Pathology, Diagnosis and Hygiene. . By PHILIP HANSON HISs, Jr., M.D., of New York; Instrug- tor in Bacteriology and Hygiene, College of Physicians and Surgeons, Columbia University, New York. º - - New and Simple Media for the Differentiation of the Colonies of Typhoid, Colon, and Allied Bacilli. By PHILIP HANSON HIss, Jr., M.D. Primary Carcinoma of the Vermiform Appendix. By D. S. D. Jessup, M.D., New York, ãº Pathologist to the Woman’s Hospital; Assistant Surgeon to the Vander- ilt Clinic. UNIVERSITY OF MIC |||||||| & H|GAN ||||| | 3 9015 O • ’ ”~*·|- ------· și º √·---- * .--，-:•{• # …«*|-- -&.---- *|- -- **** }|------}--- ----* - -·+·aº..? . -**·*- »+|-}.*-- - - --* •'*|-ș---- ' .****•-- *ą~-- --.- +: ~~~.ae· ·!-------•--- ?N* ·^ <· -s. t-.**-·º . ^, & * *… *• .- -|×---- ·---- t º:*::::::::::::::: - - - ** . ****** ** *- - a . . .” sº * T* ~ * 2. . t ſ W. º ſº N | º R MA 3 * * - º º, W A. \!. * N º | || ^ |º |ſºſsº- ſº* , º\\ aº º||-|i. §:|º|ºf ',|Uſ.S º cº, & “ * - - ºf . , ſº : ºº . . . º.º. & ºf . ; ſ º g Fºº, sº º, ſº " ' " sº , Kº Hº ; º º º * . . . . * †. fº ---- sº. ºf ºxy. • , ** -- I " ... ... º. - ºº: , sº 3 . . . . º.º. º Pºrk * ºf Š - º ſº º Ú * * * ºr 3 º ſ " " ," & A. a ". . ' ' a . º º; Sº . . e” º º , ‘A Wº º As º, ºf . , ; º a lº * - || w Lº ſº. º - sº tº *, * * & Q ºz. A | A. ſº * fº º º Pºſ. d - - -> - g º v . . .” - º / /* S W . º. A, siſ. A ºgº". º * * * * * º ſ ſº - * f : hº ſ s 3. * : *: o ſ' / ". º a . º, 4.4%, 4. . A ! . f Aſſº"/", º º - Z" d º . º 4. s'ſ ! . -- A - º 9. * * /* ** * - ; - - */ ,” A . ºf 5 tº -- d - *º-r º s's lº & º º ... .º. s ºr a l. º: * , º Nº N \R^\