20 Society for Experimental Biology and Medicine. 



happened in some instances that, after repeated injections with 

 cultures of one of the varieties of the dysentery bacilli, the animal's 

 blood agglutinated the variety of bacilli not used for immunization 

 in as great or even greater dilutions than it did the variety injected. 



It was found that a serum drawn from a horse which had been 

 immunized by repeated injections of the dysentery bacillus re- 

 ceived from Shiga, agglutinated this bacillus in dilutions of I in 

 500. The same serum, however, agglutinated the mannit-fer- 

 menting bacillus in dilutions of 1 in 1,000. The serum from 

 another horse which had received injections of the mannit-fermen- 

 ting bacillus, agglutinated this bacillus in dilutions of 1 in 1,000, 

 but also agglutinated the Shiga bacillus in dilutions of 1 in 500. 

 Judged by these reactions, these two varieties of bacilli would 

 appear to be much alike in their affinities. Indeed, Shiga and 

 Flexner seem to have come to this mistaken conclusion. Duval 

 and Bassett certainly fell into this error when they announced, in 

 the fall of 1902, that the mannit-fermenting type from the Balti- 

 more diarrhea cases and the Shiga type had identical agglutination 

 characteristics. 



The blood of young animals was found by the author to be 

 comparatively free from bacterial agglutinins. A goat 6 weeks 

 old was found to possess no appreciable bacteria-agglutinating 

 substance in its blood. After four injections of a culture of Shiga 

 dysentery bacilli, its blood in dilutions of 1 in 500 agglutinated 

 Shiga bacilli, but only in dilutions of 1 in 10 the mannit-ferment- 

 ing variety. A rabbit whose blood was negative before immuniza- 

 tion, after six injections of this bacillus agglutinated the mannit- 

 fermenting bacillus in dilutions down to 1 in 5,000. The Shiga 

 bacillus was agglutinated only in dilutions of 1 in 20. The slight 

 development of substances which agglutinated the mannit-ferment- 

 ing type during the process of immunizing to the Shiga bacillus, 

 cannot be considered as showing any affinity between the Shiga type 

 and the fermenting type, since about the same increase was found 

 in animals injected with nutrient bouillon and other substances. 



15. " Gastric secretion induced by a reflex from the intes- 

 tine " : HOLMES C. JACKSON. 



The main points in his results were summarized by the author 

 as follows : 



