474 G. N. CALKINS AND L. H. GREGORY 



is absolutely necessary to have representative lines from each of 

 the first four cells. The J series was the 16th set, starting as 

 an ex-conjugant July 22, and dividing for the first time on the 

 24th. The second division occurred on the 25th, thus estab- 

 lishing the four quadrants. These, and the following divisions 

 are shown graphically in diagram 3 and the origin of the differ- 

 ent pure lines given. These were numbered in the order of 

 their appearance and without regard to sequence in the different 

 quadrants (see p. 487). 



The five cell divisions, necessary to start the 32 lines, although 

 lagging in quadrant marked B, were finally completed on July 

 27 with only one mishap. This was a pathological division of 

 no. 19 which resulted in two deformed specimens no, 19 and no. 

 32, and both of which finally died on August 3 without further 

 di\dsion. Two of the 32 lines were thus lost at the outset. The 

 single individuals representing these 30 lines were allowed to di- 

 vide without further isolations after the 5th division and the 

 progeny of each were transferred to vials for the permanent cul- 

 tures. These were examined every four or five days, the physi- 

 cal Gondition recorded, and the culture medium changed each 

 time. The record shows that the organisms in certain lines were 

 weak from the start, thus lines no. 3, no. 22, and no. 23 were 

 weak at each period of inspection and finally died, no. 3 and no. 

 22 on August 22, no. 23 on August 26. Two other lines, no. 

 5 and no. 6, also were dead on the 22nd, although they had 

 been strong and normal up to that time. Line no. 4 died out 

 on October 7 after a period of weakness. The death of these 

 eight lines left 24 pure lines distributed as follows : Four from quad- 

 rant A, five from quadrant B, eight from quadrant C, and seven 

 from quadrant D. One other line, no. 14, from quadrant C, has 

 died since October aRer the conjugation tests had been running 

 for three months, so that only 23 lines are now alive (diagram 3, 

 p. 487). 



