250 BARBER. 



doses given alone, the above series show more decisive results. 

 While in no case was immunization to the test dose effected 

 by less than one 24-hour agar-slant-culture of avirulent pest 

 bacilli alone, the 22 which received only 1/10,000 to 3/10,000 

 of a loop mixed with agar gave 8 recoveries. This result, taken 

 with the proof obtained of the long survival in animals of aviru- 

 lent pest when inoculated with agar, gives some encouragement 

 of success with a series inoculated with a slightly more virulent 

 strain of plague in agar. Experiments with agar and organisms 

 other than plague are now in progress. 



Summarized briefly, the results obtained demonstrate: 1, The 

 long persistence of ordinary nutrient agar mixed with avirulent 

 plague bacilli and inoculated subcutaneously in guinea pigs 

 (twenty-nine days) ; 2, the long survival of avirulent plague 

 bacilli in pure culture in such inoculation (twenty-five days and 

 twenty-nine days) ; 3, the possibility of immunizing a propor- 

 tion of animals with very small initial doses of avirulent plague 

 bacilli in agar (1/10,000 loop). 



The results indicate that as good or better success may be 

 obtained in immunizing with living avirulent plague bacilli alone 

 as with doses mixed with agar. This is true of the strain of 

 avirulent plague used. The method might give better results 

 with another strain of avirulent plague or in immunization with 

 some other organisms. 



ADDENDUM. 



It was noted that even with highly immunized animals which had sur- 

 vived a severe infection with virulent plag-ue some local reaction followed 

 the test doses of avirulent plague. This was especially true of the animals 

 in Table I receiving a test dose of three-fourths million. It was thought 

 worth while to ascertain if a smaller dose would give in highly immunized 

 animals such local reaction. Ten animals of the series described in Table 

 I, which had survived an infection following a dose of three-fourths mil- 

 lion avirulent bacilli, were inoculated subcutaneously with doses of from 

 45 to 80 virulent plague bacilli. A similar dose was fatal to 7 out of 10 

 nonimmunized controls. None of the three noninfected controls or of the 

 10 immunized animals showed a local reaction greater than that which 

 could be accounted for by the prick of the very fine capillary pipette 

 used in inoculating. 



