236 



X. M. STKYKXS. 



very near the 45.9 per cent, out of the equatorial plates in 

 metaphase. The division of supernumeraries or their failure to 

 divide in the first maturation mitosis seems to be a matter of 

 chance, depending on their position in the spindle in the prophase 

 and on the attachment of spindle fibers from one or from both 

 poles of the spindle. In Fig. n both supernumeraries are con- 

 nected by fibers with both poles, in Fig. 12 the 5-chromosome is 

 connected with both poles and is about to divide, and in Fig. 13 

 one 5-chromosome is connected with both poles and will later 

 divide, while the other will go undivided to the upper pole of the 

 spindle and therefore to one second spermatocyte. The behavior 

 of the other chromosomes indicates a more or less definite attach- 

 ment point for the spindle fibers, near the middle of the chromo- 



11 



12 



FIG. ii. Spindle showing two supernumeraries (s), each attached to spindle 

 fibers from both poles. Mag. 2,000. 



FIG. 12. Anaphase showing supernumerary (s) about to divide. 



FIG. 13. Metaphase showing X, a supernumerary (5) attached to one spindle 

 fiber, and another (5) attached to two. m = mitochondria. 



some in both spermatogonial and spermatocyte mitosis (Figs, 

 ii and 13). The supernumeraries seem to be able to make 

 connections with both poles in most cases if they are in or near 

 the equatorial plate in late prophase stages. 



SEX RATIOS. 



The sex ratios in Diabrotica soror and Diabrotica 12-pitnctata 

 have shown very peculiar variations. In studying the male 

 germ cells of D. soror in 1907 I made no note of the number of 

 females found in random collections, but in dissecting D. 12- 



