SEX-RATIOS IN GAMBUSIA HOLBROOKI. 179 



degree, for the very great disproportion in the numbers of the 

 sexes. 1 



It must therefore be concluded that the inequality of the sexes 

 of adult Gambusia is real, and cannot be accounted for on the 

 basis of faulty methods of collection or of inability to recognize 

 the sexes. This fact has led some investigators to hold that 

 sex in Gambusia is conditioned by obscure fluctuating environ- 

 mental or physiological factors, and that it is not unalterabh 

 determined at the time of fertilization. Conclusions similar to 

 this have been held by Woltereck ('08) in reference to various 

 Pceciliids, and by Okkelberg ('21) in reference to the brook- 

 lamprey. 



Sex-Ratios in Fishes. But little work has been done on the 

 sex-ratios of fishes generally, and nearly all of this only on pop- 

 ulations not under experimental conditions. In this work the 

 enumeration was only of adults, usually by superficial examina- 

 tion only, and hence the value of the findings is very uncertain 

 in the presence of so many variables. That is, the sex-ratios 

 reported for fishes have in nearly all cases been simply " sex 

 ratios of collection," and explicable upon a number of hypoth- 

 eses. Darwin ('75) anc l Fulton in various papers, have largely 

 given us what little data we have on the subject. 



Among the factors that might cause atypical ratios in adult 

 fish two alternative ones occur most readily to the mind: (a) a 

 possible differential death-rate of the sexes during the embry- 

 onic, juvenile, and adult period, coupled with a normal sex- 

 ratio at fertilization, and (&) an atypical primary sex- ratio, due 

 to an atypical distribution of sex-determining chromosomes to 

 the two daughter cells in the maturation divisions of the germ- 

 cells. As a result of this unequal distribution of the sex- 

 chromosomes, a preponderance of one sex over another could 

 conceivably be produced. The approximate proportions found 



1 Essenberg (BiOL. BULL., Vol. 45, pp. 46-97, 1923), in his very interesting 

 paper finds that in old populations of Xiphophorus the percentage of males is 

 often very high. He concludes on the basis of his observations that in this 

 species there may be large numbers of " retrogressive females " which later 

 transform into males. In my observations on Gambusia I have never found 

 intersexes or instances of sex-inversion. 



