12 ROBERT W. HEGNER 



regained in every case, although in one branch of hne 150.2b 

 thirty-three succeeding generations of uninucleates were pro- 

 duced before an empty shell was formed and nuclear doubling 

 occurred, and in other branches doubling did not take place 

 until after thirty-two, thirty-one and twenty-three uninu- 

 cleate generations were passed through. Sometimes the first 

 offspring produced by a uninucleate would be empty and the 

 second offspring binucleate, but in other cases 6, 7, 9, 11, and 12 

 uninucleate progeny were formed by single specimens before 

 nuclear doubling occurred. During the time these lines were 

 being reared three binucleate specimens threw off emptj^ shells. 

 What caused them to do this is unknown, and the process was 

 not accompanied by any nuclear changes within the parents nor 

 followed by any marked modification of the later offspring. 



Why the organisms do not remain in a uninucleate condition 

 indefinitely, why an empty shell is formed at the time of nuclear 

 doubling, and how this doubling actually takes place, are ques- 

 tions still to be answered. However, it seems probable that 

 when the factor or factors that initiate this phenomenon become 

 operative the uninucleate specimen undergoes nuclear division 

 and forms a shell, but the shell is cast off empty and the parent 

 retains the two nuclei. From a genetic standpoint it is of par- 

 ticular importance to emphasize the fact that it requires three 

 or four generations before the full diameter and spine number 

 of the line are regained after nuclear doubling, since here a very 

 great single change in the internal condition of the organism is 

 revealed b}- the body as a series of small changes, and hence 

 the small and gradual heritable changes that have been noted 

 by various investigators who have selected body characters 

 in their breeding experiments, may really have been due to 

 large changes in the germ plasm that were only slowly exhibited 

 by the soma. 



5. Correlations. Are diameter and spine number correlated 

 in binucleate specimens and in uninucleate specimens? The 

 correlation between the diameter and spine number in both 

 uninucleate specimens and binucleate specimens within lines 

 150. 2a and 150.2b is marked as shown in tables 4 and 5. The 



