CONDITIONS FOR CONJUGATION 365 



The epidemic in 25a was less intense, involving at any one time 

 a much smaller percentage of organisms, and lasting but a short 

 time (Mar. 18th to 25th). Isolation experiments yielded con- 

 jugants during a somewhat longer period, but in scarcely any 

 greater numbers. 



The sequence of events in these two cultures is seen to be very 

 similar to that which obtains in such races as 3a and h, 11a 

 and its derivatives. That is, there is a period of vegetative 

 growth lasting from two to three weeks, after which con- 

 jugation ensues. The duration of this period of conjugation 

 depends in large measure upon the susceptibility of the races 

 in question, beginning earlier and lasting longer in the more 

 susceptible ones — except in some few instances where, in small 

 cultures, the process is consummated in one or two days (e.g.. 



Ill)- • 



In order to determine whether these races {23a and 25a) 

 which had conjugated under uniform conditions had really 

 done so as a result of their greater natural susceptibility, and not, 

 as might be supposed, because of more favorable environmental 

 conditions, new strains were started from isolated exconjugants 

 of each. In this way the initial medium was greatly altered, 

 new strains of bacteria coming in, with the elimination per- 

 haps of many strains of bacteria present in the old cultures. 

 Of the two strains derived from exconjugants of 23a which were 

 reared successfully, conjugation occurred in each, approximately 

 three weeks later. In one of these, 23f, isolated March 21st, 

 conjugants were observed on April 14th (5 to 10 per cent), 

 April 15th and 16th (casual). In the other culture, 23i, isolated 

 March 21st, conjugation was only casual, one pair being seen 

 on April 12th in a mixed culture of two races, P. caudatum 17a 

 and P. aureha 23% (table 2). Two cultures of 25a exconjugants 

 {25e and 25h) started on the same day as those from 23a, and 

 similarly treated, showed no conjugants, even in isolation experi- 

 ments. The relative susceptibility of these two races, of the 

 second generation, to each other is thus shown to be the same 

 as for the two parental strains. That is, the race 23 seems more 

 susceptible than 25 in each case, although conjugation failed to 



