198 S. W. GEISER. 



would have been a satisfactory criterion of sex in the three and 

 a half months old fish. However, to avoid any possible mistake, 

 through the existence of sterile males, the gonads themselves 

 were examined. 



In Litter a there Were 19 males, 17 females, and one whose 

 sex could not be ascertained because a mistake in technique 

 lost the gonads during imbedding. The individual whose sex 

 is uncertain, however, was probably a female, as such a notation 

 was made in the dissection notes. In Litters b, c, d, of the 58 

 surviving fish 27 were males and 31 were females. For the 

 total group of 94 young Gambusia whose sex was learned with 

 certainty, there were 46 males and 48 females, a very close ap- 

 proximation to a i : i ratio. 



It will be recalled that in Litters b, c, and d seven fish were 

 not recovered, and presumably died. If half of these dead were 

 males and half were females, the proportions of the sexes would 

 then show a still closer approximation to equality. Even if all 

 the dead were females, still the sex-ratio would not even begin 

 to approach the great disproportion found in adults. 



In these populations, taken all together, the percentage of 

 individuals whose sex was unaccounted for is very low only 

 8.3 per cent. so that the approximate equality of the sexes is 

 not accountable for on the basis of a differential death-rate in 

 favor of the males. On the other hand, there is evidence to 

 show that the converse condition is responsible for the slight 

 numerical inequality of the sexes in the older litters. 



Corroboration of the conclusion reached that the numbers of 

 males and females in Gambusia at birth are approximately equal 

 is found in the results obtained by the examination of populations 

 from two large pools kept under conditions as nearly ideal as 

 possible. /. Into one of these pools there were put in May, 

 5 pregnant females. About the middle of October 45 young and 

 three of the parents were recovered. Of these 45 young, 21 

 were males and 24 were females, again a very close approxima- 

 tion to equal numbers of the sexes. //. In a pond known 

 locally as the " Euglena Pond " all of the Gambusia had died 

 during the winter of 1920. This pond was stocked in May, 

 1921, with 48 gravid females. In the following October 284 



