150 J. A. DAWSON 



necting strand was always relatively shorter and thicker (fig. 

 21). Such a 'miscible' condition was present in the animals 

 (part I) which simulated conjugants and in the cannibals, but 

 the typical physical condition was in these cases always quickly 

 resumed if the pairs were separated or the cannibals isolated and 

 allowed to divide, the daughter cells in both cases being always 

 constricted off sharply from each other with no pulling out of a 

 cytoplasmic strand. 



A more striking contrast was seen when twins divided by fis- 

 sion to give both twins and single animals, as the latter invariably 

 divided in the manner usual in the typical single animal. It is 

 therefore evident that the environment, although it was probably 

 the principal factor in bringing about the pseudoplastogamic 

 union, was not responsible for the continuance of the state of 

 protoplasm necessary in such union. These results lead to the 

 conclusion that, since the 'miscible' state of protoplasm contin- 

 ued only so long as the twin state persisted, it was the intimate 

 union of the protoplasm of each component with the consequent 

 possibility of protoplasmic interchange which was responsible for 

 the permanency of the twin condition. 



In other words, once a twin animal is formed the condition of 

 protoplasm necessary for the maintenance of this union is trans- 

 mitted to its progeny. This 'miscible' quality of the protoplasm 

 of the double animal apparently acquired by the cells just previous 

 to their union is clearly transmitted from generation to genera- 

 tion. Thus the inheritance of the 'miscible' condition in twin 

 animals conforms with the conditions postulated by Jennings 

 ('08, pp. 628, 629) as necessary for the inheritance of an acquired 

 characteristic. The selection experiments showed unmistak- 

 ably that the ability to remain united was possessed by some in- 

 dividuals in a greater degree than in others. A comparison of the 

 pedigrees of subcultures T4 and T44 (fig. 8) as well as a study of 

 figure 7 leaves little ground for doubt that there was a strong prob- 

 ability of obtaining a strain of double animals in which all the 

 progeny would breed true for an indefinite number of generations, 

 that is, no single animals would be thrown off by the double 

 strain. 



