376 C. W. MITCHELL AND J. H. POWERS 



It can not be other than surprising to find this all but uniform 

 maintenance of high and low potential respectively in these 

 series of rotifers, carried not only through long series of partheno- 

 genetic reproduction, but through sexual reproduction as well, 

 and this in spite of the fact that sexual reproduction involves the 

 temporary return to the apparently uniform saccate type. 



To test the matter still further, through more cycles of sexual 

 reproduction, and also to make up for the paucity of numbers in 

 the series on the saccate side in our last experiment, further experi- 

 ments were begun as follows: Mass cultures were founded from 

 the tenth and fourteenth generations of the saccate series I 2. 

 These were carefully fed so as to induce male production without 

 the customary accompaniment of the transition to the humped 

 type. This end was successfully attained in both cultures. A 

 young female from the mass culture derived from the tenth genera- 

 tion was fertilized, and by March 20 had produced resting eggs. 

 Two of these hatched on March 27, one of which became the 

 parent of a pedigree series termed 1 oil 3. From the mass culture 

 derived from the fourteenth generation oi I 2 more resting eggs 

 were secured. A number were dropped by four individuals on 

 March 12 and 22. These were isolated and from a large number 

 which hatched on March 28 and 29 fourteen were chosen to 

 become the parents of pedigree series, which make up the second 

 part of the group of series termed / 3. This group of series thus 

 compose the third group of inbred saccate lines derived by sexual 

 reproduction, each group from the preceding group, and all 

 from the original parthenogenetic individual. 



That these series oi I 3 might be paralleled by at least a few 

 series derived from our humped lines two new series were started 

 from J. It was perhaps unfortunate that these series could not 

 have been started from some one of the ten series oi J 2, but these 

 series had been discarded after the end of the tenth generation, 

 as we had not at that time considered the possibility of the pro- 

 longed continuence of our experiments. It thus became necessary 

 to have recourse to the original series J. On March 15 the tenth 

 individual of the fiftieth generation of J was fertilized by a male 

 produced by a sister individual of the same generation. Resting 

 eggs were formed which hatched on March 26. One only was 



