2 cect Seen eee neetietliN “pet eM On H 82 Report of Dr. Avery with Drs. Stillman, Goebel, Dubos, Francis, Babers, Goodner, Terrell and Rogers. I. Type III pneumococcus pneumonia in monkeys, (Drs. Francis and Terrell.) 1. Experimental production of disease, 2. Treatment with snecific enzyne. 3. Pathology. II. Nature of lysis of Pneumococcus by autolysis, bacteriolytic enzyme and bile. (Dr. Dubos.) Ill, Study of the antigenic complex of type-specific pneumo- cocci. (Drs. Avery and Dubos.) IV. Methods of purification of enzyme decomposing the capsular polysaccnaride of Tyne III Pneumococcus. (Dr. Dubos.) ¥V. Isolation of pure cultures of microorganisms decomposing tne capsular polysaccharide of Type II Pneunococcus., (Drs. Dubos and Goodner.) VI. Pneumococcus dermal infection in rabbits. (Dr. Goodner.) 1. Therapeutic action of antipneumococcus sera. 2, Deter-:ination of active immunity by dermal infection. 3. Effect of pneumococcus autolysates upon pneumo- coccus dermal infection in rabbits. VII. Study of pneumococcus infection and immunity in rabbits by the inhalation method. 1. Pulmonary lesions produced by inhalation method. 2. Duration of deronstrahble antibodies in serum of rabbits following inhalation of Pneutiococcus. 3. Active and passive immunity in rabbits. VIII. Interconvertibility of specific tynes of Pneumococecus. (Dr. Rogers.) reece bros a IX. Study of the chemical nature of antibodies in antinneumo- coccus serum. (Drs. Goebel and Avery.) X. Derivatives of glucuronic acid and glucuron, (Drs. Goebel and Babers.) XI. Synthesis of p-amino phewol glucosides of lactose and gentiobiose. (Drs. Goedel and Babers. ) XII. sibliography. ---000--- I. The production and treatment of Tyne ITI wneunococcus pneumonia in monkeys. (Drs. Francis and Terrell.) In their first papers, describing the action of an enzyme of becterial origin upon the capsular nolysaccharide of Type III Pneumococcus, Avery and Dubos also presented convincing evidence tnat this enzyme is capable of decomposing the same substance wnen encountered in the animal body. Mice were infected with many lethal doses of Type III pneumococci. It Was. founda tnat as late as 12 to 18 hours after the infection was established intraperitoncally, the administration of the anzyme brought about recovery in most instances. Furthermore, when the enzyme was given 24 to 48 hours. preceding the injection of many lethal doses of type IID organisms into mice, it was found to have a “marked protective action resulting in recovery of most of tne animals. Tne mecnanisn of the therapeutic effect was shown to be due to its action on the cansular poly- saccharide, decomposing it, and thereby rendering the bacterial cells vulnerable to attack dy the phagocytic cells of the animal body. This study was limited to effect of tae enzyme H 84 on Tyne III Pneumococcus in an experimental infection primarily septicemic in nature. Later, Goodner, Dubos and Avery studied the action of the enzyire upon the localized infection of tne sin of rabbits (termed vy Goodner "dermal pneumonia") produced by a virulent strain of Tyve III Pneumococcuf, In this form of infection, the lesion is primarily localized and is accorpanied hy sevticemia of varying degree in different animals. Here again, a definite therapeutic effect of tne enz me Was denonstrable as shown by prompt subsiaence of tne cutaneous lesion, and sterilization of the blooc. However, with the a-ounts of enzyme used in the studies, tnere was found tt» »ve a degree of septicemia above which tne enzyme was unable ‘to control the infection, altnough the duration of lire of the rabbits might be prolonged over that of control animals. Since excellent results had been obtained in the treatment of septicemic and local infections of the smaller animals, it seemed advisable to study the effect of the enzyme under conditions in which the disease process more close- ly resembled that in the human patient suffering fron pneumo- coccus lobar pneumonia. in this disease, the organ involved is highly vital, the course 4s comparatively constant, the pathology typical and recovery problematic. Consequently, an atte’pt was made to nroduce in monkeys a disease resembling lobar pneumonia in man, which could then be subdjected to treat- ment with the specific enzyme under conditions more closely approaching taose encountered with vatients in the hospital wares. The Java monkey (CY, cynomolgas) was selected for the purpose, and tais cnoice has proved to be a fortunate one, since TOMY Not ae Kad H 85 the M. rhesus is much moreé resistant to the strain of Type III Pneumococcus employed. In the first part of the study the animals were given morphine and then under ether anesthesia a small catheter was inserted into tne trachea; the inoculum, consisting of organisms centrifuged from a given amount of culture and resuspended in 0.5 cc. of soluble starch Was allowed to flow {nto the lungs. During the procedure the animal was tipped to tse rigat with the hope of localizing tne matcrial in tne right lower lobe. Tais technic was, in the majority of Ainetances, successful. However, the location of the infection was not always predictable, since the exact site of the catheter at the time of injection was not known. At times, too, during the recovery from etner anesthesia, the monkey often went througn a stage of excitement, with hyperventilation and exaggerated physical activity, or the reverse condition with cessation of breathing which required artificial respiration Such incidents apparently resulted in pne monia scattered diffusely through the lungs and accompanied by a high septiceria. In spite of these variable factors, localized lobar pneumonia Was produced in a high proportion of the animals. | In the latter part of the study, a more accurate method has been employed using the technic described by Terrell and Robertson. With this method, the animal is given a dose of mornhine sufficient to make it auiet. The animal is placed in the supine position, and a sv ll radio-opagvue catheter 18, Anserted into the trachea. With the aid of the fluoroscope the catheter is tnen guided witn direct visual control into the eee ae wane Ls aXe oa soeeremipnrr—eannnnreinaien eee coast ic arenaaniaaiee meiremne vind pesrsvianamionignanianeiin nicer eee es ae om awae oer 7 =< Bieta mao | | ot A eee" 208 H 86 bronchioles of the lobe or LTobdules, in which the infection 4s desired. Wnen the catheter is so placed, the organisms sus- pended in 0.5 cc. of 5 per cent cornstarch emulsion are jnjected The metnod has several advantages: first, 4t eliminates general. anestnesia; second, the site of injection is known accura-el?; tnird, with better Localization the number of organisms required TERS ee m for infection is some what gmaller and the septicemia conseauent- ly tends to jncreaBe more gradually. eee: Throu snout the study, temperatures were recorded 3 oF 4 - times daily and the course of tne pneuxonia was followed by daily roentgenograms. Daily blood counts and blood cultures were made. In general, within 18 hours after infection, 4 aefinite well localized pnoumonic consolida*ion can be seen in the X-Tra?- Goncomitantly. a gradually rising fever 4s noted. The white blood cells may be increased in number, oT, if septicemia is present, they may have decreaset. The lesion continues to spread so as to Ainvoive the entire lobe and may extend to other Lobes. Tae temperature remains elevated, the senticemla increases, and tne Leuxocyte count tends to fall progressively. ‘Death may follow in 4 to 7 days in the typical case. On the other nana, in 4 certain number of instances associated at times with eitner a aif fuse infection of the lunes or with the well localized lesion, there 43 an immediate fall | in Leuxocytes, & heavy septicemia and a tendency to subnormal temperature with a ranidly snreading lesion and an early fatal termination. There is als0 a ground of monkeys in which recovery takes place spontaneously after 3 to 7 Gays: In tneseé, senticenia 204 H 87 o qnstances the ‘say be absent or slight in degrees although in tw as veen comparatively nich. There is 4 tendency of sg and a h Leuxocytost ad of the sacterenia p to resvond with cefinite fever, but with rel less spre ris grou atively g occurring in this neous recoverie The sponte 1 anizals in t yneu m e it aifficult to spear of contro of andividua 4 be eliminate he usual “zroup mak sense, since tne factor 1 variation, so far as He a or allowed for. peen able to a ally true whe canno 4scover, riment makes re expe nave n tae purvose of t. atal course This is especi gressive and f ively loné» but gradually pro the production of a comparat tae most gesiradle: From tne point of view of pneumonia, of course, 4% does not matter, but when an attenpt to erapeutic agent 4s made, the ideal 4g to approach untreated anin This has £ treated and evaluate a tn als. absolute comparisons © ys with een possible. Llts in g2 monke not 0 The gtudies to date ancelude the resu aefinite pneunmonic Lobar consolidation of various degrees. of these, 20 animals, or 24.4 per cent, fad no septicemia and all als, 28 per cent, naa rapidly nty-taree anim _ we er than 5,000 c 1 cc. of olonies per 1s died. The recovered. ticemia great mounting 88? ree days: n tne girst th were suvaivide Ail of these anime pups on the basis blood i h gepticemi® g ‘animals ‘a into two er af septicemia P whic ater than 250 co n whicn tne sept remainin resent: (1) those in Lonies ver 1 cc. of plood acemia ranged of the degree nt put not ere was preseé (2) those 4 e first 3 days; The first tn th 250 and 2,000 colonics in the first 3 days or 31-9 per cent, and {Tt is 4{n these petween wprises 24 animals, groups ©0 of tneseé two 2 animals, or 14.5 per cent, the second tneluaes 1 two groups that the condition possivle etfect of al ciseasc. belo, cav paring tne aninals with tnat of siauiler the fyne IIT polysecensride-splitting enzyme. dividing tas Srouns on tne b all of tue aninols treated, 3 days arter inrection. Animals vith pneusont than 250 col. /ec. tne engy™e upon t The results obtained treatment wortality rate £ anivals receiving treatment with asis of tue first received tae onzre in the first s seem most favorable to study the he course of naoserics of untreated an and senticentia less in first 3 das. Gt YWO n &B~ the exnerizent- ore trbdulated The reason for 3 days is thet — Yo. of animals ; Mortelity Untreated 2 ented Recovered pied | ner cent 15 7 _9 55.3 __ 10 10 0 0 anianis with pneu-zonin 250 to 2000 col. /cc. ena senticenia from in first 3 days. No. of anizals | Vortalit? Untreated treated aecoveres | Died | per cent 8 2 6 75 J 4 4 0 0 Summar of both groups. No. of aninals . Yortality Untreated Treated Recovered Died | per cent 24 | 9 15 §3.1 _ 14 14 0 0 The results on the basis of conparative mortnalit: figures are striking. treatment tnere have been no 50 per cent of the unt aninals were treated on the first day after infection. In all of noted, in several tie In the animals of t fever ha fatalities, whereas, reated animels hanve succumbed. taese a hese groups wnien received @efinite consolidation a@ reached 105° to 106°, well over Five more and one on tne second day was and they eee ||" o gi 206, L H 89 geemed on a clinical basie to be suitable cases for the thera- peutic test; however, at the time of treatment tneir blood cultures were sterile and remained so. They all recovered. Six others with very hish initial senticemias were treatec and al- though they all died, one treated on the first da» with a septi- cemia of 18,000 colonies per cc. at that time did not oie until the fifth day, and then with 2,000 colonies of Type III Pneumno- coccus per cc. Another with 25,0C0 colonies per ec. on the second day of disease lived until the fourth dav, and at the time of death the septicemia hed been reduces. to 3,700 colonies per cc. This marzed reduction of the number of organisms in the circulating blood has veen noted in others of the treated group in which deatu occurred eerl:. In addition to the efiect of the enzyme on the senticeif, there is usually seen in the X-rays a definite prompt effect unon tae pulmonic consolidation. Tre spread of the lesion is limited and the advancing margin begins to clear. aAt the same time, the fever tends to subside and tne animal 4s alert and active, and takes interest in his food and surroundings. The immediate constitutional effect of the enzyre has varied with different preparations and with tne eeverity of tne disease. some preparations have caused a fall in terperature ané, white blood count and have made. the animals appear sicker. This is esvecially noticed when the monkey is already extremely sick. In animals moderately sick there is a tendency to a febrile rise following treatment. It mey be that the fall of the blood. count is partially due to an increased mobilization of tne leukocytes at tne site of iniection. "y poy we QQ H oO oO The treatment has usually consisted in the intravenous administration of 10 cc. of enz:me, var‘ing fron 2.5 to 20 units per cc. Following te original treatment, blood cultures and white counts are ade avd further treatment has been brsed on the results of these studies in conjunction with the general clinical aopearance and tie X-ray evidence. In some instances three treataents have been given in a period of 24 hours. In tne later treatments the intraperiton-al route has sometimes been employec with satisfactory results, and perhans some re- duction in the amount of general reaction. Altnougs the stud) must be carried further, the results so far appear nromising. Recently, it nas Seen our privilege to move tue monkey clinic to nuarters wnicn combine many features that tend to exveaite tne actual wor”. It is honed that with more experimental ani-uals, and with further improvement in technical procedures of concentration ard purification of the enzyme, more potent and highly purified prenarations may be available and that unequivocal results of its theraneutic effect may be obtained. | Pathology: As a side-product of the primar; clinical study of x ‘ experimental pneumonia in monkeys, an excellent opportunity has been prosented for tae study of the pathology of the experinent- al disease in anivals dving at different times with various degrees and stages of involvement. The microscopic preparations have been wade with the kind cooperation of Dr. Rhoads and rs. Whitehouse. Tne Gescription of the microsconic lesions is limited t» sections obtained from the lungs of monte:s infected by allowing te {noculu: to flow by gravity to a given site. In the animals dying early in the disease there has been an acco-panying heavy septicenia. As carly as 24 hours an entire lobe may be involved. Its color at this time is a dark bluish- red, It is large, firm and airless, except, perhaps, at the extreme margins. There is usually a rogéerate degree of fibrinous pleurisy, sometimes a small amount of serous fluid. The septur between it and adjacent lobes is obliterated »y adhesions. On cut section, the involved lobe is comparatively dry, but a viscous fluid, typical of Tyne III pneumococcus pneumonia, may be scraped off. Microscopically, the most advanced mrocess May be seen in tne portion nearest the root of the lobe, while extending peripherally different stages may e ercountered. [In the region of the main vessels 2nd bronchi ‘there may already be seen a marked cellular exudate primarily composed of leukocytes. Fibrin has appeared in aifferent parts. Tne alveolar walls are moderately swollen and their canillaries are engorged. Ordinar- ily, the number of red blood cells in the alveolar exudate is not striking. The lymphatics of tre larger bronchi and bronchioles, although dilated, show Little evidence of infection. Tae walls of tae vroncnioles show little infiltration, although much exudate is present in their lumina. The walls of the blood vessels do ‘not avnear to be involved. In this area comparatively few organieng are seen. | uxtending peripherally in a gradual transition from the preceding phase, there is seen marked engorgement of the alveolar canillaries witn gwelling and desauamation of the alveolar epithelium. The alveoli are filled with exudate coT- posed of edema fluid, moderate numbers of polymorphonuclear leukocytes and erythrocytes usually in small numbers. There way be a few deposits of fibrin spicules and organisms are numerous in the exudate, but not notable in the walls of tne alveoli. At the margins of tae lesion tne predominant features are the filling of tne respiratory lovules with edema fluid conteining great numoers of becteria and co-naratively few cells. There may be extreme congestion of the alveolar capillaries so that tne walls are outlined »y erjthrocytes filling the capillaries. tk The alveolar walls are swollen, but the outstanding feature is toat, in spite of tne number of organisms in the material wram eTCTT filling tne alveoli, they are infrequent in the walls of the blood vessels, the sronchioles and alveoli. Lae SPN ST In animals dying on the fourth to seventh days, one may find several lobes completely involved, and the woole lung covered with fibrous pleurisy ana adherent to the thoracic wall. The lobe which was the site of the original nneumonic consolidation is a slate grey in appearance, oF bluish grey. It is large, firm ana airless. At times a local area, red in appearance, may be seen. Tm lung on cross section is reddish grey and ver. ary with rough, granular surface. Microscopically, most of the lobe will be seen to be in an advanced stage of the pneumonic process, The alveolar walls are poorly distinguished because of the dense character of the exudate and the large amounts of fibrin. The alveolar canillaries contain erythrocytes but are not prominent. In many phases there is a shrinkage of the exudate away from the walls of the alreoli, but strands of fibrin can be seen extending from one alveolus to another. In the terminal bronchioles the process is quite marved. In the areas of dense involvexzent, especially where fibrin is abundant, f, ere ge H ae few organisms are seen ond the number of leucocytes is Icss. Paue fibrin eppears to be ost marked in areas adjacent to the blooG vesse¢ls. There is practically no evidence to su.-gest vascular thrombi, and tne walls of tne vessels are survrisingly free of evidences of infection. In tae larger hronchi, much exudate may be seen in tie lumen, but there is little infiltra- tion of the valls even adjacent to arens of marzxed nneuxonia. The lymphatics do not anpear to be infected. In certain areas, peraaps most frequentl; in the peripheral portions of the lobe go involved, one notes the apnenarance of large yellowish cells in tne alveoli. Where they arc present the fi>vrin is disanpear- ing and tne polymorphonuclear leucocytes are mucn dirinisned. These cells are tac tynical fors seen in resolution, and ere apparently actively phagocitic. In the other lobes, depending upon the duration of involve- ment, the type of lesion appears to vary with the age. In some instances there seems to be an area of direct extension from lobe to lobe, but usually. the oldest lesion is seen about the hilus, waile in tne lobe. in waieh the process is most recent the peripheral port ions may show enzorgement of the alveolar capillaries with much edewa fluid and many organisms in the , alveoli typifying the advancing margin of tne spreading “infection, In animals sacrificed after spontaneous recovery has begun, there {s frequently a persistence of fibrous pleural adhesions. The involved lobe 4s smaller taan normal, has a rubvery consist- ency, and its color 4s a ratner pale greenish yellow. The surface may be wrintled. On cutting the lung, it 1s rather firm, SCL H 94 dry, but flabby. Very little air is present. In the micro- scopic section the striking feature is tnze predominance of the large yellowish phagocytic cells. Certain alveoli contain large numbers of the’ with few polymorphonuclear leukocytes and less fibrin. Elsewnere the alveoli are clear but for 2 snall number of these cells; the alveolar walls are becoming less swollen; the epithelial cells are being replaced. The capillaries are per- haps more dilated than in the stage of dense consolidation. Scattered throughout the section are portions in which fibrous tissue is mating ite anpearance and beginning organization. At times tne plugs of exudate are coated with a layer of epithelium and almost suggest the organization of a thrombus with small new vascular channels making their avpearance. Tne conplete clearing of tne infection is comparatively rapid, although just as the laying down of fibrin is marked in the acute process, 80 is the amount of fibrous tissue in the reparative process. ~--000--- II. Nature of Lysis of Pneumococcus by Autolysis, Bacteriolytic Enzyme and Bile. (Dr. Dubos.) Among tne interesting properties of the Pneumococcus cell is the ease with which it undergoes autolysis and the readiness with which it is dissolved by vile and bile salts. These two properties serve to differentiate pneumococci from other Gramnm- positive cocci; they are also of sone interest in the study of bacterial variation in the Pneusococcus group, since the R variants are more resistant both to autolysis and bile lysis 302 H 95 than the § forns. Finally, these two processes result in the loss of the ch racteristic antigenic property of the S cell, i.e., its power to stimulate the production of tne tyve-snecific capsular antibodies. What is tne nature of the phenomena waich cause such striki- modifications of the cell? It has long been realized that auto- lysis is the result of the breakaown of some cell constituents by the so-called autolytic enzymes. Cell-free preparations of these enzymes have the propert, of changing heat-killed pneumo- cocci into a mass of Gram-negative devris; this action is fairly limited to tre Pneumococcus group, but, within this group, both R and S forms, irrespective of type, are eoually affected. A very satisfactory method of preparation of the bacterio- lytic enzyme is as follows: A chilled suspension of pneumo- coccus cells is thrown into 15 volumes of crilled acetone, the precipitate separated by centrifugation and dried as a grey powder. When resuspended in a medium in which the cellular enzymes cannot function, this material may be snown to consist of avparently normal Gram-positive cells. If, however, the cells are resuspended in saline they undergo lysis instantan- eously, leaving only a mass of Gram-negative debris. Outside of the interest attached to tnis phenomenon of flash lysis, the technique provides a rapid method of obtaining a solution which in turn is very active in causing lysis of heat-killed pneumococci. Although autolysis Was soon recognized as an enzymatic phenomenon, most investigators have felt that bile lysis was of a more physical nature, particularly because of its amazing Tt was also felt that rapidity, even at ice box temperature. the property of undergoing lysis in bile was in some way associated with tae life and death of the cell, since any procedure which killed tne organism would also render it bile in- soluble. However, the following experiments, very briefly out- lined, seem to indicate that bile lysis is only an activation of the autolytic enzymes, and that it does not occur when these enzymes are inhibited or destroyed. When pneumococcus cells are suspended in citric acid, for instance, they slowly die but remain Gram-positive and retain their morvhology. If these dead cells are then suspended ina buffer solution at pH 7.0, they undergo lysis at a normal rate, they are found to be bile soluble, and when subjected to freezing and thawing give an active preparntion of bacteriolytic enzyme. This experiment snows that dead cells remain bile soluble when their enzymes have not been destroyed, but have been only placed under conditions where they cannot function. When, however, the colls are killed ‘Sy agents such as heat, mineral acids,salts of heavy metals, whicn not only inhibit but also destroy the bnacteriolytic enzymes, they are no longer bile soluble. A direct evidence of the activating effect of bile-on lysis by the bacteriolytic enzyme is supplied by the fact that the rate of breakdown of heat-killed cells by the enzyme is very much accelerated in the presence of bile or bile salts. The enzymatic nature of bile lysis is also indicated by the influence of temperature on tne phenomenon. The rate of bile lysis was measured at 0° c., 5° C., 20° C., 37° C., 45° c., eee een enn Z 204 BR 97 53° C. and 60° CC. Betweon 0° C. and 45° CO. tne phenomenon was found to have a temperature coefficient of 2 to 3. At 53° C. lysis began very rapidly but soon stopped and never was com- pleted, suggesting that the enzyme itself hnd become inactivated. At 60° C., no lytic action could be detected. All these facts seen to indicate the enzymatic nature of bile lysis. The following observations also tend t» prove that nutolysis, bile lysis and lysis of hext-killed cells hy sterile preparntions of bacteriolytic enzymes are fundamentally the same phenomenon; all 3 processes have a temperature optimum at 45° ¢C. and are inhibited at tevperatures greater than 52° C.; the range of pH lies between 5.5 and 8.2 with an optimum near the neutral point; the rate of all three processes becomes exceedingly slow in distilled water, but increases with salt concentration, at least up tn 5 per cent NaCl; finally, in all three cases, lysis does not occur in the presence of agents ‘kenown to inhibdit enzymes It seems that the foundations have now veen placed on which to build a study of the mechanism of lysis. Four observations are being studied as possible leads towards an understanding of tnis mechanism. 1. The fact already referred to that. cells precipitated with acetone @éxhibit flash lysis. 2. Activation of lysis by bile and bile salts. 4. The rate of autolysis and of lysis of heat-killed cells by bactertolytic enzymes is very much activated by small amounts of iodine. This activation, however, can be recognized only within a narrow range of todine concentration. Above a certain concentration the enzymes are completely inhibited by fodine, the cells retain permanently Gram-positive and are no longer soluble in bile. It may be mentioned here that cells thus treated with an excess and then washed free of iodine are still susceptible to lysis by the bneteriolytic enzymes. It has been mentioned early in tais discussion that all neat-killed pneumococci, irresnective of types, ore suscentivle to lysis by tne bacteriolytic enzyme. It should be added hore, however, that from tnis point of view the Tyne I cells are by far tne most resistant, ana the Tyne III the least resistant. It is perhaps more tian a coincidence that the Tyne I contains also the most stadle type-snecific antigenic comolex, and T:rve IIT the least stable. ~~-000--- Ill. Study of the Antigenic Complex of Tiyne-specific Pneuno- cocci. (Dr. Avery and Dr. Duos.) When virulent, encapsulated pneumococci are subjected to any form of lysis, tney lose their property of stinulating tyve- specific antigenic response in rabbits. Such prevarations of lysed cells, however, are still capable of immunizing mice agains: infection with virulent pneumococci of the hovologous type. This observation suggests the existence in the intact virulent Pneumococeus cell of two distinct antigenic functions, both of which are type-specific. 1. Capsular polysaccharide antizen, This antigen is con- cerned with the production in the immunized aninal of a type- specific antibody directed against the polysaccharide constit- uent of the Pneumococcus capsule. Practicall: all rabbits immunized with suspensions of intact heat-killed Type I pneuno- cocci respond with the production of type-specific precipitins, H 99 agglutinins and protective antibodies. If, however, these sane cells are first allowed to undergo autolysis, or are dissolved by vile, or their hesat-killod bodios are dizested by the addi- tion of bacteriol-tic enzynes, and these preparations are in- jected into rabbits, tuc animals fail to oxhibit tzne-specific response. As shown carlicr in this report, these three forns of lysis are only different expressions of the activity of the same lytic enzywes. We may say, therefore, that the capaular pdlysaccharide antigen is rendered ineffective by the action of the cells' own onzy7es.— When rabdbvits are inmunized with a auspension of heat-killed Type III organisrs, only a few animals respond with the produc- tion of type-specific antibodies. When, however, fodized cells are used for immunization, practically all aninmels exhibit the fornation of type-specific precipitins, agglutinins and protect- ive antibodies. As has been stated earlier, iodized cells are. easily attacked by tre vacteriolytic enzymes and thus changed to a mass of Gram-negative debris: Now it has been found that treatment with the bactérfolytic enzyne does not seem ’té modify the