Chomo-immunological Studies on Pneumococcus. Report of Dr. Avery with Drs. Tillett, Julianelle, Goebel, Dubos and Dawson. I. Theory of Antigenic Dissociation. Brief statement of theory. Attempts to prevent antigenic dissociation by cell fixation. Action of selective agents; Iodine, Fomalin. II. Autolysis: Chemical study of Enzyme Action. Relation to proteolyais, lipblpsis. Reaction accompanied by antigenic dissociation, inactivation of enzymes, oxidation of hemolysin, and format ion of purpura producing substance. Previous studies of intracellular enzymes on foreign substrates; what action have they on native protein, carbohydrate and lip- oids of cell itself? III. Bacterial Variation. 1) Pneumoc occus . a. Interconvertibility of R and 8 forms. b. Conversion of S 4 R. c. Revei-slon of R 4 S. 2) Effect of adaptation to growth at high temperatures. a. Attenuation of virulence. b. Production of variants. c. Differences in susceptibility of different types. 3) Friedlander's Bacillus. a. Occurrence of variants. b. Different f0rur.s of variants. C. Distribution of specific types and variants in human and animal infections. IV. Studies on Ratural Rosist.gco and Acquired Immunity to Fneumococcus:-. --- Natural resistance end immune response of normal rabbits to Tyi>e III. Bon-ty?Ie sgecific resistance induced by IIR1' form of Pnemococcus. Passive transfer of acquired resistmce. V. Anaphylaxis with Pneumococcus Polysaccharides. a. Nature of Pneumococcus haptens (protein-free polysaccha.rides). b. Active sensitization. c. Fassive sensitization. d. Fatal anaphylact ic shock. VI. Are the Specific Precipitable Substances of P~C~~OCOCCUF, Ea.otcns? a. Investigations at Foch Institute. b. Experimental evidence of non-ant igonic n.ature. VII. Immune Response of Rabbits to Intracutaneous VaccinatiE VI'II. Studies on Oxidation and Reduction by Fncumococcus'* .__-A 1) Bacterial ankagonism botwcen Fneumococcus and Sta$hylococcus aureu a. Antagonism in presence of oxygen. b. Symbhoeis in absence of oxygen. c. Reaction of phenomenon to cellular oxidation. 2) Relation of oxidation pr6oossea to viability. 3)_Relation of oxidation processes to initiation of growth. 4) Importa,nce af oxidation - reduction potent ial. of media* + IX. Nature `and Duration of Immunity Induced by Inhalation Method. X. Publications. SIyc?, *' .e 2;8 I. Antigenic Dissociation. In the last report the theory of sntigenic dissociation and its significance in Pneumococcus immunity was discussed in detail. During the past year new facts in support of this theory have been ac- quired end the principles involved have been cxgerimentally applied in the study of active and passive immunity. In order to establish the sequence of thought between the work previously recorded and that to be discussod in the present report, tho results of recent chcmo-immun- ological studies may be briefly summarized as follows: Chemically the bacterial cell is composed, among other things, of two constituents each belonging to two wholly different classes of substances and both forming togothor the'major part of the cell as a whole. Those two components, - one carbohydrate and the other protein - constitute a unique system which determines both the typo-sl?ccificity and the anti- gcnicity of the cell. The protoin is the somatic substance racially common to all pneumococcl, while the carbohydrate, the capsular ma- terial, is chemically different and serologically specific for each type. In this system theprotein is constantly present as a~ assent- ial constituent of the cell protoplasm; the carbohydrate, on the other hand, is the synthetic product of a specialized function which is independent of the purely vegetative processes of cell reproduc- tion and growth. By `reason of the chemical nature of the substance elaborated, this faction bestoplrs upon the cell highly differential properties which biologically endow it with type-specificity and morphologically distinguish it with a well defined capsule. This special function is inhibited or suppressed whenover the cell is sub- jected to an untoward cultural environment, and is on'hanced whenever 259 the organism is gram in the animal body. The presence or absence of carbohydrates determines the irnnunological specificity of the cell and modifies its antigcnic nature; for, pneumococci which by chemical or biological methods have suffered loss of this substance no longer exhibit the specific retions of the enca$aulatod forms from which they were originally derived. In more recent studies on bacterial variation, to be discussed later in this report, the appearance of variants in a given culture is shown to be intimately, if not casual- ly, related to the loss of the physiolbgical function of elaborating this specific substance. Apart from the importance of the carbohydrate as the de- terminative substance of type specificity, this constituent is unique in that when chemically isolated in protein-free form, it still re- tains the capacity to react with antibodice induced by immunization with the whole cell but loses completely the powar to stimulate these same antibodies when injected by itself into the animal body. The polysaccharides of pneumococcus, thereforo, belong to the group of immunologically important aubetancos which Landstoinor has call@ haptans o In cellular union theee substances constitute the type- specific antigen, separately they are devoid of antigenic fbnction. Evidently, their antigenicity is dependent upon the nroscnce in the cell of some other substance from which they are easily dissociated in the form of haptens. Upon conditions determining the physical or chemical stability of this combination rests the antigenic effectivc- ness of any given type of pneur2ococcu8, and conversely the more eas- ily dissociable this complex is, the less efficient it is as antigen; th& is, the antigenic potency of any given type of pneur.ococcus is inversely proportional to tho rate and oxtcct of dissociation of the haptaphoro groun. This, in brief, is the essence of tho theory of antigenic dissociation. Its significance in the problems of antipneumococcus immunity at once becomes clear when it is reslizcd that tho only specific therapy of the disease at -present available is dependent upon t h turn d .I in its ant ITg .e 10 use of typo-specific serum, the potency of irhich is in rectly relntod to the ar,tigezic quality of the orgnisns used production. As pointed out in the last report the titre of ~ur~ococcus SOW in terms of type s?ocific antibodies is con- ditionod by a balance between thoso factors of the animal body which bring about antigen cleavage and those chemical properties of the bacterial co11 which detsrmino tho stability of the specific anti- genie complex of each type. The factors relating to the animal body ;vill be discussed in the experimental nork on the ViaturzA resistance and acquired im- munity". The relation between chemical constitution and ant igenic stability is being studied from two viewpoints: - 1) Selective fix- at ion of antigen by chemical agents, and 2) Chemical changes occurr- ing during autolysis. 1) Selective fiction. Assuming that the union between pnoumococcus haptan (polysaccharide) and its antigenic activator is a labile one, a search is being made for chemical roagcnts which might "fix" this linkage and by stabilizing the compound prevent or inhibit subsequent dissociation. One of the most promising ngents employed thus far in an attem?t to bring about rcr_tigenic fixation is Iodize. Qpxtitat ivc esti?ntion has shot-r, that living l:;ou .ococci susT,endcd ii1 ZL st Tnd2rd solution of iodine absorb surprisingly large ammts of the rcagm:t. After saturation with iodine the cells arc killed and remair; irtact, well fomed and Grm poeitive for r?or.ths. They ur.dcrgo no nutolgsis and the processes of oxidation cease. The inactivstion of lytic e;l zyne 9 ) the sugression of oxidativc reactions and the urcscrvatioi: of morphological integrity are of distinct advantage in conservi?.g the mtigen. Comparison of heat killed and iodized cells shows that i:i the former instame, death resulting fron exposure to 560 - 60oC. is 800~ followed by cell dlsintegrstion with the appearance of shru9e::, Grm. negative and shadow forms, due ir, part perhaps to imompletc L:- activation Of Cellular enzylr.88 at this teq8ratUre and in mart t0 CX traction. which subae+lently goes ori in the saline suspe?:sioE. On the pthar hand, the unheated, iOhiZ8d cells prepared fro? the sm8 mss culture remain intact and retain indefinitely their for% size -aud stain@ reactions. Moreover, the iodized vaccine is nOZi-toxic, md animle treated for weok with these prepnmtiozs ShOW ,?o ill effects. Immmlogicsl investigations by other WOrk8rS' havo demou- stratod that iodized proteinrs are altered in their antigenic specific- ity. Tho type-specificity of pneumococcus antigen, hoWeVer,.is de- temined not by $30 bwterial protein but by the carbohydrate substzl;co which envelopes the cell. Moreover, the polysaccharidee of the threo specific types differ fror, th8 starch-glycogon group of carbohydrates in giving no color rosction with iodine. Such combination as iodim rakes with the ~011, therefore, does not chmicnlly alter this import- mt and dotcminative constituent of the mtige=. That the specific- ity of the iodized ccl1 rornaics wtige3ic?lljr umltcrcd is shown by yy` r' a2 the fad that saru?! of immnized animals agglutic3tee equally ml1 both the treated and untr8atOd orgmisns and protects r.ice qai;lst icfectioz with virulent organiszs of the homologous type. Th8 use of fO~~lii~ aa f lxat ive md the increased produc- tion of antibodies when fomolized cells of Type 1 are Used as ati- gOC, nas discussed in the procedicg report. Experkents now ir, pro- gress will shortly yield results whereby the coqarative ceritg of the various fixntivce for the different bacterial types my be fir,- ally dctcminod. Apnrt fro!`? the ktcrosticg thoorotiaa~ considcrn- tions of the rolatior, of these obsorvatiom to the nctual chcrricsl constituents of the ir?zurAzing antigen, the eethods evolved my evectually come to have practical value in the production of potent antisera for therapeutic purposes. II. Autolysis. (Dr. &38b8ll As part of the problem of deteminin& the nature and chemical 6tmCtur8 Of the immunizing a?tig8n Of PnemOCOCcus, a study is being mde of cell autolyels, During spontaneous dissolu- tion of pnemocooci there is an accomanying dissociation by the ef- f8CtiW typ8-antigen, Therefore, to how whether proteolytic md lipolytic cleavage of cell substaqea occurs during autolysis, and to what extent, if any, theaa chmges are' aeeociated with the 10~s of antigenicity of the autolyeate, would, indirectly &t least, ctf- ford some clut to the mchanisn of antigenic dissociation and to the zature of the linkage b8tW88n the carbohydrate and the sub- stance which confers antigenicity u?on it. Fro? the exyW%ent%l evidoncc ~ro.?.dy nvrLilable, it is apparent Chat autolysis is rot to be co:.fused wit?. protoolPiS, yY+y 263 sir,co only slight increzsc is clnino nitrogen occurs 2nd t:lc cell protein Zzftor prolonged autolysfs still retains its mtigenicity. The ratios of total citroger. to non-con.guable and mino nit rogcn are zlso bcir,g detcr9ncd since it is knoan that during nutolysis sme dogradztior. product is relcmed, probably in the fom of a protcoso, which gives rise to the purpura producing substmce in pnamococcus zutolysntes. so far, little or no appreciable hydrolysis of the lip- oidsl substmca of the co11 has beon deterrzined, although final judgmect on this phase of the autolytic process roust await the use of the Cars exact mthods now being perfoctod. The outcome of them studies on =lutolysis, nhethor they throrr light on the processes of antigonlc dissocistion or not, vi11 z& lc?.st add to the knowledgo of t?m %&ion of the intracellular eczyms on the nstivo constituents of the cell itself and sup@cr:ent our earlier studios on the action of these sm.8 Iokents on forcig:l substrates. III. Bacterial Variation. 1) Pm&.ococcus. a. Icterconvertibility of IIRfl amA ItSn forms. (Dr. Dawson) (Conversion of IfS't to LIP.) The process of conversion of virulent, type-specific, capsulated l*Sfl types of pnem.ococcus into avirulent, non-type specific, non-capsulated "R" form ha8 bocc stud- ied in furthcr dotail. The nest effective rooms r,f brir-ging about this chmge is by growth in hoeologous imur,e sorb and differences have been fomd to exist i:: the rcadinoss wit5 qhich the trmsforrn- tioc cm be cffcctod ia the various typos. Type I ttStt has bcor- found to be the most difficult to convert to the "IQ' forms requiring ten to fifteen transfers in serum dilutions of 1:2 or 1:4. Even after this number of transfers `lintermediaten colonies exist which readily revert to the "S" tme. On the other hand, Types II "$7` and III "`3 are readily converted to the "R" variety, and in each instance the change is more abrupt and complete than in the case of Type I "S". (Reversion of `R" to "S"). Reversion of "R" forms to their homologous "S" type has been effected both by an in vivo and an in vitro method. (a) In Vivo. By animal passage IIR" forms, de- rived respectively from Types I, II, and III `S", have been trans- formed into the `IS" type. Differences have been found to exist in the constancy of the "R" variant. Some nR" forms readily revert to the virulent "Sn type while one strain has been studied which `has remained completely refractory to all attempts to effect the trans- formation. In the earlier experiments the intraperitoncal route was used exclusively but. more recently the subcutaneous injections of large amounts of "Rn cultures has given a much greater number of positive results. Single-cell as well as mass cultures have been employed in all experiments and in every instance the single cell cultures have reacted in the same met as the mass cultures from which they were derived. lb) In Vitro. The in vitro method of effecting the "RN -+ "St transformation is by growth of the ttR1' forms in anti-"R" sera, the optimal concentration of which is 10 Fer cent. As in the in vitro experitihs both mass and single-cell cultures were employed. Such a finding argues against the hypoth- esis that the virulence of a culture depends upon the relative num- ber of lIRfl and Ifs" forms of whic.h it is coqosed. . yy) 265 Ir? all icstazccs ir, nhich roversion has beer, effected both in viva tsr,d ic vitro t!m "RI' forzs h>ve reverted to tho I~S~l typo from which thoy mere originally derived. RoversloE of 1W to lisle is almys accompanied by the ac- quisitior. of all the charsctoristics of tho IIS1 type, imluding m,x- inal virulallco. 2) Effect of Adaptation to Growth at Highor Tczpemturcs. Prolicimry observztiocs have been mle on tho cromth of Typos I and III pnou~ococcus at higher tct?perntures (39OC). The re- sults of onrly invcstigntions have bcaen such as to suggest that fur- ther work my rovoal slgnlficant differences in the behavior of tho various types. After fiftoon trmsfcrs nt 39OC the straig of Type I IIS1' employed had coqletely chw&ed to the aRVB fom, while after thirty trmsfers at the s1yTe tcqomturc ZL Type III IIS strsir. still produced only sSs colonies. Further aork is now being undertaken to detemine the effect of growth at 39oC on a variety of serological types. 3) FriedMnderbs Bacillus. (Dr. Julianelle). During the course of studies on the biological and izzunological properties of Friedlllnder~s bacillus, three sharply defined types were found to exist among different strains of this organim. The types were de- signated A, B, C, and into one Group, X, were placed several hetero- genous 9 t rains. Later studies revealed that, under certain condi- tions, variant form za.y be induced in cultures of Friedltlnderls bac- illus. The typical colonies of the organism- are identified as Smooth . (S) and the variant colonies as Rough (R). The `IS" strains produce capsules, solubls specific substa;lce, are vi rulcnt and type specific. _. The R strains, on the other ha& produce no capsules, nor soluble specific substance, are not pathogenic and are group, specific; that is, they are serologically undifferentiated regardless of their type derivation. Further study has since disclosed that there exist arnon;j: the "FP variants of Friedl8nderls bacillus additional forms which exhibit definite differences in morphology and antigenicity. Three different forms of R colonies have been studied and designated as %' I$, I&. The RS variety is extremely unstable and since it was never obtained in pure culture, it was studied only corL>hologiczlly. Rl and RZ. however, have boen observed in greater detail and they have been found to differ both grossly by colony formation, and mi- croscopically by the size and arrangement of the individual cells. Roth variants (IL1 and Rz) moreover, may be differentiated further by serological reactions. Roth forms are agglutinated in antisera prepared by the inJectIons of lack the capacity of complete Each variant adsorbs -from the rabbits with either strain, but they reciprocal agglutinin adsorption. homologous antiserum agglutinina for both homologous and heterologous organisms; from tho heterologous serum, however, antibody is removed only for the strain employed in the adsorption. The R forms differ considerably from their ant- ecedent rS" strain in colony appearance morphology, virulence and ant igenicity. A number of methods have been adopted to induce reversion of the R forms to the parent "Sll type. Kowever, whether the tech- nique or its aplication was inadequate, the results were uniformly negzt ive . This does not me,an that all W' cells of FriedUnder's bacillus are irreversibk, but that under the conditions of the cxpcriment, the method employed did not supply the proper stimulus. to reversion. The spontaneous development of "RI' variants in I'S" cul- ture has been found to accompany the process of a$ng. Growth in homologous immune sora in vitro also converts the "S" ~011s into "R" forms. That variation, however, is more than in vitro or cul- tural degradation gains support from the fact that "R" forms hzvo been found in cultures taken directly from foci of infection in the animal body. In fact, in 17 cultures from different sources "R" forms wcrc found in 5 instances. Interestingly enough, the "R" variants were four,d only in chronic infections and alw,Ws in con- juxtion with t(Sfl forms. Further Observations on the Occurrence of Specific Types of Fried- Under's Bacillus in Disease. (Dr. Julianelle). In the preceding report, a summary was presented of the distribution and relative ' frequency of occurrence of specific types in infection8 associated with Frie&4nder's bacillus. Up to that time, observationa were made upon 39 etraina;`since then, the number 68. In the following table the distribution is summarirad to date. . . ' has been increased to of the specific types Total number of Type Type TTe Gr;up * Strains studied A B . 62 32 9 7 13 Type A - 32 strains. 24 strains from Pneumonia in man, 2 from adenoid tissue, 2 from throat cultures, 2 from fatal abscesses in guinea pits, 1 from liver abscess in man, 1 from cystitis. Ty-pc B - 9 Strains. 4 strains from fatal Pneumonia in &liilC3 pigs, 3 from Pneumonia in man, 2 from gcnito-urinary infection in horses. Type C - 7 Strains. 2 strains from Pneumonia in man,.1 from sputum of Pneumococcus pneumonia, 1 from subacute sinusitis of antrum, 1 from nose (sinus infection), 2 source unknoV!n. Group X - 13 Strains. 6 strains f ram Pneumonia in man, 2 from fat.21 infection in guinea pigs,1 from cystitis, 1 from feces (Pellagrs),l from liver abscess in man, 1 from lung ab- scess, 1 from throat culture. Although the total numbor of strains of FriedlYnder~s bacillus studied is still insufficient to furnish conclusive data on the relative frequency and distribution of types, nevertheless, it is interesting to obseyve that of 35 str.sins isolated from pneu- monia in pan, 29 or 83 per cent belonged to one or other of the three fixed types, The frequency of Typo A icfection in Friedl!knd- erls pneumonia in man is shown by the fact that organisms of this type were isolated from 24 (68 per cent) of 35 cases studied. It is remartible that these percentwes are the same now with a total of 62 strains, au they were upon the forrxr analysis b%sed on 39 s t rains. Since there seems to be great confusion as to That de- finitely constitutes a Friedl#nderls bacillus, nr,d v:hnt sugars nre fermented by organisms of this group, the reactions of 45 strains have beer, studied in media cont.aining tho differont carbohydrates, dextrose, lactose, sucrose, cnnnite ,arYd maltose. -yz 2s9' The results of t!le ferrcnt7tion studies rcvc4 th?t, in &oncrnl, there is c-rent variability a?.on< the different strains and that the adaptation of such reactions for a classification of this s>ecics offers more confusion th2n syste;r. The fermentation rext ioi:c, not only do not aid in the classification of types but offer no as- sistxxe in differentiating the Fricdltlndcr~s bacillus from closely allied orgxisms. The one striking fact brought out in this study is thzxt whereas the majority of strains in Typoe B and C and Group X ferrxnt lactose, 61 per cent of the strains of Type A produce neither acid nor gas from lactose. IV. Natural Rosistancc and Acquired Irzxnity to Pnemococcu~. (Dr. Tillett-: In three previous papers, experimental studies on natural and acquired immunity of rabbits against pncumococcus mucosus have been described. Since an interpretation of the results reported in the published experiment9 haa formed the basis of subsequent v*ork, a brief review of the earlier studies will be given.. From observations on the results of injection of Type III pneumococci into normal rabbits, two rain facts `have accrued which seer to be related. First, it was found that irxunization OF rabbits with heat killed Type III vaccine or living org&&risne failed to elicit specific antibody response; secondly, that injection of livid, eilCo.p.- sulated, mouse-virulent, Type III organisms in large doses into rabbits did not produce fatal infection. These two results led to the devolcp- Tent of the hypothesis that normal rabbits possess some mechanism which is capable of severely injuring the capsular coTonent of the Type III bacterial cell. On the basis of this conception the absence of type specific a&utinins in the serum of immunized rabbits is explained 8s being due to tho dissociation of the type specific coirrponcct of the anticcn; whea living pneuwococci are injected, injury of a sirilpr character is inflicted on that part of the ccl1 which is inticately associated with virulenoe, norely, the capsular fraction, and, in this instance, recovery of the infected animal is brought about. In the previous publiehed experiments, which concorncd acquired ro- sist,ance it has been shown that rabbits whes irzunized either with non-type specific, avirulent, R form of pneu?ococci or Rith the fixed types themselves, are resistant to infection with a strain of Tme III, which had been -?ade virulent for rabbits. It has also beer, subsequently found that i~ization with R forms of pneumococci induces active immunity in rabbits against infection wit3 Type I or Type II organisms as well as Type III. Since when R organismsare used for immunization the production of type specific antibodies is excluded, it has been necessary to seek elsewhere for a correct in- terpretation of this form of active immunity. It has seemed possible that the factors of natural resistance previously discussed might play a part in this non-type specific inn?unity end that the increas- ed resistance of rabbits immunized with R pneumococd might repres- ent an exaltation of the mechanism normally present. Consequently , the work of the past year has been devoted to a study of this form of specific immunity. It seemed of first irq?ortance to dotemine whether the acquired izunity was due to protective substances preseat in the circulating blood, in the fixed tissues, or both. Consequently, experincnt s have beer, carried out to determine tvbeth- er- this fom of resistance my be passively transferred to r-orml animals by means of the blood or serurr, of resistant rabbits. Prev- ious ettcr.ats to Trotcct --ice Rassively by the use of serum of rab- bits immunized with R Rneumococci have been uniformly negative. Repetitions of I)rotection tests of this kind using both xhole blood and serum in relatively large amounts and varying the time of serum injection rith respect to time of infection, have alaays resulted in failure. HoTever, in contrast to the failure Of passive transfer- ence in mice, it has been found that the blood and serum of ant i-R rabbits does passively protect normal rabbits to a marked degree against infection with any of the fixed types of pneumococci. The transference of protection, successfully accomplished by the use of whole titrated blood and by serum, demonstrates that the ;Wotectivo substance8 are present in the circulating blood. Following tho es- tablishment of this fact, it became necessary to determine how much blood or serum was required to furnish maximum protection; the de- gree of protection in terms of the infecting dose af culture; and the duration of the protection this afforded. It has been found that 15 to 20 cc. of blood or 10 to 15 cc. of serum afford great-' est protection muni ty against 100,000 lethal and that this amount constantly bestows a solid im- 10,000 lethal doses and in many instances against doses of virulent pneumococci of any type. It has 8180 been determined that the passive protection nffOre8d by whole blood endures for as long as three Reeks nftor transfusion - long- er periods tive not been to&xl. Experiments planned to define more clearly the nature of the nrotoctive substances.ln the blood of rabbits immunized with R organisms are at present in progress but tbo results are not 8s yet complete enough to warrant reporting. The nature of the investiga- tions at present under way include attempts to absorb from immune serum, by means of pneumococcal cells, the active ?rinciTlcs; to neutralize them with specific carbohydrates derived from pneumo- cocci; to determine their heat lability; to determine their influ- ence on growth and viability of the organisms in vitro. It is a striking fact that although immunization of rabbits with degraded, avirulent pneumococci stimulates no demonstrable type specific antibodies, these animals possess ahigh degree of active immunity to virulent pneumococci of all types. Furthermore, the blood of rabbits so immunized although unable to protect mice, does afford passive protection to normal rabbits, Work is being carried on to define more clearly the actual nature of this form of acquired rc- sistance which is effective in the absence of type specific anti- bodies and which appears to invoW principles other than those usually associated with type-specific immunity. V. Specific Anaphylaxis with Pneumococcus Polysaccharide. (Dr. Tillott) Previous studies on'bacterial anaphylaxis with Pneumo- COCCUS have dealt with the action of cell solutions and autolytic extracts containing unavoidable `admixtures of protein and other soluble products which are often primarily toxic! As the result of recent chemical etudiea, there are now available in highly purified form specific carbohydrate derivatds of b3UmOCOCCUS which are devoid of toxic properties and free from protein de,-rad- ation poducts. In view of the increasing significance of these specific polysaccharides in the proccsscs of immunity, it seemed import`ant to determine whether these chemically purified and pro- tein-free substances isolated from type strains of pneumococci 273 actively psrticiyate ic the rcactiox of bacterial ana#ylaxis. The results may be briefly summarized as follows: Pnemococcus polysaccharldee isolated in Trotein-free form from Fkeumococcus mea I, II, and III are devoid of the fumtion of inducing active sna?hylactic sensitization in guinea gigs. This lack of sensitizing power adds further evidence ir, support of the view that these Bubstances, when cheaically purified, are devoid of true antigenic function. The fact, however,, that in the dis - sociated form they retain, as haptens, the property of combining specifically with antibacterial antibodies led to attcqts to test their capacity to indtice ana?hylactic shock in passively sensitiz- ed animals. Guinea pigs my be rendered Taseively anaphylactic to the ppecif ic carbokgdratee with the Freciyit?,ting sera of ixzune rabbit 8. &nixils mssively sensitized in this tamer die in three to .four ninutes with acute shock followirig an iajection of mounts as small as 0.. 005 erg. of grofoin-free polysaccharides of the hoigol- ogous type.. "no .symptomaloQI and DathologJ of the animle are iden- tical `in avery way .wlth those of true protein axxqhylaxis. The re- actions are type-specific, VI. Are the Specific Frccipitabls Bubstmces of Pnewococcus, Hagtens? (Dr. Julianelle) . . This is the title of a Schiemann and Caspsr of the Koch paper recently publish& by Institute in which, on the basis of their experimental results they answer this question in the neg- ative. They isolated czrbohydrnte substmccs from the specific types of pneumococci and purified ther? by chmic4 rlethods until they no longer gave the Trotein color,teets. In ace icst mce, by . repeated purification the xtorizl ::`a~ rendcreh nit rogcn-f rec. They found that mice injected with theso yrel>srntions in repeated sr:.all d9se8, acquired a considerable degree of active imunity xaicst in- fection with virulent pneurococci of the horologous ty?e. By reason of the irrrr-unity induced, they conclude that the specific yrecipitokle subst-aces are true ,antigens and not haptens. These results were amareatly BO definite, 3nd at the s?m.e tine so contrary to our own concelAion of the knxnological nnturc of the specific ?olysaccharidca, that confirmation of one or the other point of vie3 seemod easontial. Fortunately, we had at our disposal highly purified pre- parations of the specific carbohydrates which by chemical test were known to be free of protein and in the caee of the Type II and Type III substance to be aleo nitrogen-free. The Type I substance differs from the other two in containing nitrogen as an apparently essential Component of a nitrogenout? sugar Compound; despite the Occurrence of nitrogen in this preparation tt fails to give any of the protein color tests. To test the antigenic and immunizing power of these sub- Stancee, mice were given five injections lntrap8ritoneally at four day intemr`ale of a solution containing weighed amounts of carbohy- hydrate of each of the three types. The total amounts of substance given the various groups of mice cowrising 96 animals in all, rang- ed from 5 mgs. to 0.5 mg.; which in terms of bacterial equivalent represents a large amount and a wide range of dosage. Wine days after the last inJection the animals Rere infected Fith graded doses of virulent pnsumococci of the corresponding typos. The results of these experiments were decisive. Under the conditions employed, repeated injections of the purified, pro- tein-free polysaccharides failed to induce in mice any moasurable immunity against infection with virulent pncumococci; the animals died of an overwhelming septicemia following an infecting dose as small as O.OOOQOl cc. of virulent culture. These results, together with the previous failure to de- monstrate the presence of antibodies in the serum of rabbits re- peatedly injected with larger amounte of these substmces, and our recent failure to induce active sensitization in guinea pigs after the administration of relatively enormous amounts of those scsme preparations, convince us of the validity of the views that pneu- mOCOCcuB polysaccharides when sufficiently purified are non-anti- genie BUb8taIZe8 which conform in all respects to the immunologi- cal principles governing haptens in general. VII. Inmrune &3sponss of Rabbits to intracutaneous Vaccination. (Dr. Julianelle). RsbbitB haV8 beeninj8cted intraCUta,tX3OUsly at weekly in- tervale with auspenaion.:of heat killed pneumococci and the charact- er of the local reaction, the antibody response and the development Of ELCtiV8 i.nXUUnity h&W8 b88n folloW8d over a considerable period in a lEU'g8 series of an@E%ls. The injection' of the dead bacterial bod- ies into the skin gives rise to a local reaction which increases in size and intensity for the first four or five injections, then grad- ually diminishes but never completely disappears. Accompanying these reactions the rabbits acquire a marked resistance to intra- venous injection of virulent, living organisms; the active immunity . ? q) , 276 acquired in this manner is sufficiently solid and broad to protect the animal against infection 71th hetcrologous types of pnoumococci. This form of actively acquired imm.nity, mo+eover, is built 3p fr, the absence of demonstrable type-specific -mtibodics; for rsrely do rabbits vaccinated by the intracutanoous route, even after prolong- ed treatment, develop detectable traces of type-specific immune sub- st.oncos in their blood. This is the more striking since in the cnsc of pneumococcus, Type I, at least, coqarable amounts of heated cells injected intravenously, invariably stimulates the formation of type SpeCif iC antibodies in readily demonstrable quantities. During cu- taneous vaccination, but often only late in the Drocess, the so-cal'i- ed secondary antiprotein ar,tibodies appear in considerable titre in the blood. However, antibodies of this variety are unrelated to the type of pneumococci used in immunixation and reproso;lt the respoi;se to bacterial protein without references to type specificity. Although animals cutaneously vaccinated are themselves actively immune to in- fection, their sorun~, in the absence of type-specific antibodies, fails to Coafer passive protection on mice against iIIf8CtiOn with pnoumococci . The relation of these results to prohpylactic vaccination, to bacterial allergy and acquired resistance offers M intoreuting field for further investigation. VIII. Studies on Oxidation and Reduction by Pneurnococcus. (Dr. Dubosl 1) Bscterinl Antagonism. The antagonistic action of Pneu- r;iococcus uqor- tho grovth of Staphylococcus mrcus was first r.otcd by Aliv%s?.t os T:ho rccordod his obscrvntions 5:ithout atte!?TtiP! to CC- plnin tho -:echmis!-. , si:lco lion8 of the reasons usually ascribed ade- quately eqlnil;-ed the ~he;:o~;eno:: he observed. Whenever Pneumococci and staphylococci are seeded together on the surface of an sscitic agnr plate, the latter organisms fail to develop and are often com- pletely sqpressed in the areas Theress gronth of pneumococcus is most abundant. It is known that during aerobic gronth, pneur~ococcus cells combine with oxygen and that as a result of this union tkcro is form- ed a substance, having the reactions of hydrogen perolaido, nhich ac- cumulates in considerable concentration in the medium. The presence of this agent in aerobic cultures hns in earlier experiments been shozn to account for many of the phenomena associated with the lift ald death of the bncterinl cell. The formation of methemoglobin, the destruction of the endoc&llulsr hecolysin, the inactivation of the cell enzymes, are in each instance reactions dependent upon ox- idation processes, brought about through the action of pncw~o~occu3 peroxide. It has nom been demonstrated that the mechanism of the bacterial antagonism between pneumococci and staphylococci is direct- ly related to the presence and sccumulstion of this same peroxide, to the toxic action of which staphylococcus is particularly suscept- ible. Proof that the mechznism is one of cellular oxidation lies in t-ho fact, th?t these aame tno species, between which there exists such marked antagonism on aerobic cultivation, grow together in per- fect symbiosis when air is excluded f ro;r: the culture or nhen a cat- alyst is added to the -