500 



J, A. DAWSON 



A second experiment was carried out in a somewhat similar 

 manner. On January 1, 1918, from the three existing lines 

 (3A, 3C, 3D) of subculture 3 (see former history of this sub- 

 culture on pp 494, 497) were spread out to form three new sub- 

 cultures, 3A, 3C and 3D, of four lines each and, similarly, 

 from the four lines of subculture 4 (pp. 494, 497) four new sub- 

 cultures, 4A, 4B, 4C, and 4D, were begun. These subcultures 

 were carried on with daily isolations until January 22nd. The 

 division rates are shown graphically in figure 3. 



No cannibalism had occurred in any of the lines of subcultures 

 3 and 4 of which the subcultures now being considered are lineal 



P---S 



...... 



I 

 ....J 



Fig. 2 Comparison of cannibal (broken line) and non-cannibal (continuous 

 heavy and light lines) subcultures. December 12, 1917, to March 1, 1918. Con- 

 tinuous heavy line is a combination curve showing, in five-day periods, the 

 average division rate for non-cannibal subcultures (1 A, 2 A, and 3 A). Dotted 

 line shows a similar curve for cannibal subcultures (CI, C2, and C3) beginning 

 December 26, 1917. Light full line shows the division rate, averaged for five- 

 day periods, for the four lines of the non-cannibal subculture CNl which died on 

 February 4, 191.S. At A the four lines of subculture C3 died (February 19). 

 .\t B (February 23) the four lines of subculture C2 died. Methods of plotting 

 same as in figure 1. 



