118 CLAUDE W. MITCHELL 



was this exchange of fluid followed by mutation of the saccate 

 to the humped type, and in this case the transition did not occur 

 until after five days. This instance at least is thus not com- 

 parable to those transitions caused by qualitative food change, 

 for these latter almost always follow within one to two days after 

 the new stimulus. * It is more probably to be compared to the 

 occasional mutation incidental in mass cultures throughout the 

 course of experimentation. These instances will be further dis- 

 cussed as a whole later in the paper. 



Return to Paramecium feeding 



After about two weeks of feeding with Euglena the supply of 

 this organism had so decreased that its use had to be discontinued 

 in all cases except the pedigree series D. Lines D 2, E, F and G, 

 were returned to the Paramecium diet. Within one to four gen- 

 erations all had returned to the saccate type. In some cases 

 transitional individuals had intervened; in others the change was 

 immediate (tables 8, 9, 10, and 11). But all continued to pro- 

 duce only saccates for several generations of Paramecium feeding, 

 or until they were discontinued. In line D, in which Euglena 

 feeding was continued, the humped type was reproduced without 

 exception during this period. Nevertheless, after another ten 

 days, the supply of Euglena quite gave out. At this period, 

 however, an effort was made to mingle other and varied Protozoa 

 with the regular diet of Paramecium in the hope that this might 

 prevent the backward mutation to the saccate type. This was 

 successful only in part, for although in the linear series the species 

 remained constant at least as far as the ninty-sixth generation, 

 yet many of the sisters not included in the pedigree lines pro- 

 duced the saccate type. These latter were usually the later born 

 individuals of the older parents. Thus for twenty-five genera- 

 tions following the mutation which occurred as the result of 

 Euglena feeding Series D continued in the humped form. A 

 single temporary return to the saccate type occurred in the ninety- 

 seventh generation but this, as table 11 shows, again reverted 

 to the humped type in the following generation. Table 11 



