2OO S. W. GEISER. 



mens regarded as females may really have been immature males. 

 It will thus be seen that the sex-ratio, when an adequate col- 

 lection is at hand, does not materially differ from that found in 

 other fishes." He applies this conclusion to the Pceciliida gen- 

 erally. Schmidt's ('20) experimental litters of Lebistes had a 

 ratio of males to females of 100:116.6 (total of 78 young) and 

 his other data also show an approximate equality of the sexes. 

 Lebistes possesses very many physiological and cytological char- 

 acteristics in common with its close relative, Gambusia. It 

 would be a singular thing if certain closely-related species of a 

 compact family such as the Pceciliidas possessed an anomalous 

 ratio of the sexes at birth, as has been assumed for Gambusia. 



In a very recent paper Aida has given data which demonstrate 

 a practically even sex-ratio in a Japanese fresh-water pceciliid 

 fish, Aplocheilus latipes. On assembling his data for His " Ex- 

 periments 7-10, loA and B, nA and R, 12, 14, 17-21 " (all data 

 given) we find that he had a total in his experimental litters of 

 2,438 females and 2,400 males, a sex-ratio of 100 males to 103.9 

 females. This is very close to unity. 



The Chromosomal Constitution of the Pceciliidcc and its Bear- 

 ing ,on the Sex-Ratio. It is unfortunate that the chromosomes 

 of teleosts are so unsuited to investigation because of their small 

 size, lack of individuality, and tendency to clump on the equatorial 

 plate. It may be proper however, to consider in detail in this 

 connection some genetical investigations whose results give 

 strong evidence of the type of genetical constitution possessed by 

 Pceciliids. Schmidt ('20) in breeding-studies was able to isolate 

 a color-marking in a race of Lebistes which was transmitted 

 through the Y-chromosomes. He undertoook crossing experi- 

 ments with two types of males on one type of females. The 

 " new " male type, which we will call " B" possessed a brilliant 

 dorsal fin-spot which was entirely lacking in the " old " type, 

 "A." Other color characters made the two types instantly dis- 

 tinguishable. Crosses were made of Type A<$ Type B? and 

 Type Bg Type A%, and back-crosses were made with Fj popula- 

 tions. Breeding records were kept, with copious notes (and fre- 

 quently water-color drawings) of a large number of individuals 

 (e.g., the registered males of F 2 -F 5 total 998.) It was found 



