54 HELEN DEAN KING. 



which had developed under natural conditions. The number of 

 females is greatest in the lot of tadpoles fed on the yolk of egg, 

 being 8.1 per cent, above the normal; and it is least in Lot C 

 where it falls 5.9 per cent, below the normal. These figures are, 

 however, well within the limits of possible normal variation for 

 the frog as determined by the investigations of Pfliiger and Gries- 

 heim, and presumably, therefore they are also within the limits 

 of normal variation in Bufo. 



It has been suggested by Born, and emphasized by other 

 investigators (Cuenot, Morgan (8) ) that the results obtained in 

 feeding experiments may possibly be influenced by the mortality 

 that occurs during the course of the experiments, individuals of 

 one sex dying more readily than those of the other. During the 

 course of my experiments from 30-150 individuals in each lot 

 died before metamorphosis. These individuals, as I have stated, 

 were preserved and the sex ascertained when possible by means 

 of sections. From the records that were made it appears that 

 tadpoles of one sex did not die in greater numbers than those of 

 the other. In the entire number of individuals that were exam- 

 ined the proportion of the sexes was practically the same ; in 

 some lots the females died in greater numbers than the males, 

 while in other lots the reverse was the case. These results con- 

 firm Pfluger's contention that there is no relation whatever be- 

 tween mortality and sex among tadpoles reared under artificial 

 conditions. 



Taking into consideration the entire number of individuals used 

 in the experiments, it is found that in the total of 1,900 tadpoles, 

 823 or 43.31 per cent, developed into females; 713 or 37.52 

 per cent, became males ; leaving 364 or 1 9. 1 5 per cent, in which 

 the sex of the individuals was not ascertained. If we assume, 

 for the moment, that all of the individuals belonging to this 19.15 

 per cent, would have developed into females (although the inves- 

 tigation of the sex of the individuals that died during the course 

 of the experiments does not warrant such an assumption), the 

 number of females would then be increased to 1,187 or 62.47 

 per cent, of the whole number of individuals ; on the other hand, 

 if all of the in dividuals in which the sex was not ascertained had 

 developed into males, then the number of males would be 1,077 



