146 J. A. DAWSON 



show a close agreement except at the beginning and near the close 

 of the series. The curve for the cannibal line (S5C) shows a di- 

 vision rate very nearly the same as that for the non-cannibal line 

 (S5CN), except that the division rate during the initial five-day 

 period for the cannibal subculture is slightly higher than that of 

 the non-cannibal — a result consistent with that obtained in section 

 4, part II, of this work. In general, the division rates are so 

 similar that they might well represent four cultures carried for 

 four entirely similar normal animals. The twin culture and all 

 the subcultures derived from it died on March 15, 1918, the twin 

 culture (S5C) being then in the sixty-second generation. No 

 striking differences could be noted in these twins from beginning 

 to end of the history of the culture either in appearance and be- 

 havior when living or in details of structure studied in stained 

 preparations. In respect to cannibalism, it was found that 

 it invariably took place in stock slides after a short period of 

 multiplication. 



Series VI. The twin beginning this series was isolated Jan- 

 uary 22, 1918, from a stock culture in a depression slide dating 

 back to January 17th, made from stock of subculture 4B (p. 500, 

 pt. 1). This twin gave the following: 

 2d generation. 2 twins 

 3d generation. 3 twins, 2 single 



From the next generation the cultures (S6T and S6TS respec- 

 tively) of four lines each were made from twin and single animals. 

 Records of these cultures were carried in the case of the twins to 

 the thirty-third generation and of the normals to the thirty- 

 seventh. These lines at time of writing (March 26th) are still 

 living and dividing as usual. In the curves for this series as 

 well as those of subsequent series it will be noted that there is 

 no striking difference in the division rates. 



Series VII. The twin forming the starting-point of this 

 series was isolated on January 25, 1918, from a ten-day-old de- 

 pression- lide stock cu ture of subculture 4B (p. 600, pt. 1). In 

 the second generation it divided to give two twins and in the third 

 three twins and two single animals, and from these, as in the 

 previous series, cultures of twin (S7T) and single (S7TS) lines 



