470 D. W. BRUNER AND P. R. EDWARDS 



the migration of the organisms, such agglutinins were removed 

 by absorption with appropriate bacilli. The serums were pre- 

 served by the addition of small amounts of chloroform. The 

 chloroform effectively sterilized the serums and did not affect 

 the growth or motility of the bacilli. 



RESULTS 



It should be emphasized that the changes described below in- 

 volved only the H or flagellar antigens; that is, the antigens which 

 are affected in the phase variation of normally diphasic types. 

 The heat-stable antigens remained unchanged throughout the 

 experiments. On the whole the cultures were quite stable and it 

 was more difficult to induce variation in S. paratyphi A than in 

 Eberthella typhosa, S. abortus-equi or the "totally and permanently" 

 non-specific types. In some instances it was necessary to trans- 

 fer the organisms repeatedly in the presence of agglutinating 

 serum before variation was observed. 



From cultures GV and W no induced phases were isolated. 

 Strains 17 and 37407 each yielded two induced phases (phases 3 

 and 4) neither of which was closely related to any of the naturally 

 occurring antigens of the genus. All the remaining cultures 

 yielded three induced phases. The normal phase of the bacilli 

 (antigen a of the Kauffmann-White classification) was designated 

 as phase 1. The induced antigens were denoted as phases 2, 3 

 and 4. Phase 2 is closely related to the non-specific phases of 

 the diphasic Salmonella types. Phases 3 and 4 are "artificial" 

 phases which closely resemble none of the natural antigens. The 

 relationships of the phases to each other and to the antigens of 

 other species are given in table 1. Only one culture of S. para- 

 typhi A is included in the table since the phases isolated from 

 other strains reacted similarly. Phase 2 is agglutinated in high 

 dilution by Kunzendorf and non-specific Sendai serums. Like- 

 wise serum derived from phase 2 agglutinates Kunzendorf and 

 the non-specific phase of Sendai to the titre of the serum. Phase 

 3 is slightly related to the normal phase 1, but has little relation- 

 ship to the other phases or to the natural antigens of the genus. 

 It is agglutinated in low dilution by all the serum of all types that 



107 



