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



So far as measurements go Jennings believes that F 1 and A 1 

 may be the same, and F2 and A 2 the same. In the second se- 

 ries of measurements D and A 2 fall together, otherwise the groups 

 fall in the same order as in the first series and four lines or races 

 result, differing from one another by at least 4 microns in both 

 series of measurements. 



Now if similar positive differences are present in the progeny 

 of any one ex-conjugant of Paramecium caudatum cultivated 

 under identical conditions as nearly as possible, then the varia- 

 tions found by Jennings in progeny from different ex-con jugants 

 cannot be used as evidence of variations due to amphimixis and 

 the corollary that conjugation is for the purpose of bringing 

 about variations, cannot be deduced. 



Table 3 (page 484) shows that our lines measured March 10 fall 

 into three groups, line 20 at one extreme (183.15 microns) and line 

 21 (223.85 microns) at the other, each separated from the next 

 line by at least 10 microns. The other 14 lines form a group 

 separated from one another by not more than 3 microns. 



At the time of the second series of measurements all of the 

 lines were at least 50 generations older. All lines had increased 

 in size but in different degrees so that they do not fall in the 

 same order as in the first series. Line 20 however retains its posi- 

 tion as the smallest line while line 21 is removed by one from the 

 largest. With this exception we again have three groups and 

 the same groups as in th^ first measurement test. 



According to Jennings' method therefore, we have at least 

 three distinct races or lines differing in their means by not 4, 

 but from 6 to 10 microns and all deri\'ed from one ex-conjugant 

 (Jennings' lines vary from 176.9 to 206.36 and 187.03 to 220.56 

 differences of 29.47 and 33.53. Ours vary from 183.50 to 223.85 

 and 195.11 to 231.825 differences of 40.35 and 36.71 respectively). 

 We might choose five lines from these 16 and obtain five races 

 with differences between them (excepting one case where it is 

 3.75m) of four microns, at each measurement. 



Two of Jennings' lines (Fl and F2) might be compared with 

 the sum total of all lines of the J series, that is, the total progeny 

 of an ex-conjugant. If so the population in his cultures would 



