SEX-RATIOS IN GAMBUSIA HOLBROOK1. 189 



(d) Spermatozoa. The spermatids lose their cytoplasm; the 

 nucleus becomes hemispherical (Fig. 31) then elongate (Fig. 

 32-33). As the spermatozoa come to nearly attain their mature 

 form (Fig. 34) they migrate to the periphery of the cyst. The 

 heads are directed outwardly, and the tails, together with a col- 

 loidal substance, fill the central cavity, precisely as shown by 

 Philippi ('07) for Phalloceros caudomaculatus. These spermato- 

 zeugmata appear to retain their individuality until their discharge 

 from the genital pore of the male. The first spermatozoa are 

 formed in May and June, and the process contin es until October. 

 Their formation, then, is contemporaneous wrh the period of 

 warm weather. 



(c} Dimorphism of the Spermatozoa. Zeleny & Faust in 

 1912 made the discovery that in certain insects, the males of 

 which are heterozygotic, the ripe spermatozoa can be divided into 

 two groups which differ considerably from each other in size, 

 those in each group varying about a mode with but very little 

 overlapping. Later work by these investigators (Zeleny & Faust, 

 '150, '15^) extended their early results to several groups of in- 

 sects, and Wodsedalek ('13, '20) found the same dimorphism 

 in the spermatozoa of mammals. In Gambusia, the writer was 

 not able to find any such dimorphism. It is possible that when 

 spermatozoa from different spermatozeugmata are measured, 

 and the measurements grouped together, as was done in the 

 present case, a complicating factor enters in the fact that the 

 sperm-heads may thus be in different stages of maturity, and 

 hence of diverse volumes. In view of the facts, however, that 

 the sex-chromosome, if it occurs, is not very different in size from 

 the other chromosomes of the cell, and that the chromosome num- 

 ber is so large, thus making the relative increase in size of the 

 gamete possessing the sex-chromosome slight, it is doubtful if 

 such measurements of sperm heads would furnish any clue to the 

 existence of a sex-chromosome. Whether such a chromosome 

 exists or not may be much more easily ascertained by breeding 

 experiments involving sex-linked characters, as e.g., Schmidt's 

 work, already referred to. 



From the results obtained by the study of the spermatogenesis 

 of Gambusia, no light is cast on the origin of the anomalous sex- 



