STUDIES ON INSECT SPERMATOGENESIS. 407 



ance being necessary, however, for the expanded basal end of the 

 head, which is probalily to be viewed as a "centrosonial middle- 

 piece" (see Bowen '22o). 



Unlike Murgantia, in Evschistus the length of the sperms as a whole 

 also differs markedly, the actual length of the large ones being about 

 1 mm. These large sperms are not easy to find entire, as they are 

 usually broken or tangled up with others in making the smear, but one 

 can always be certain of an entire sperm by the presence of the thread- 

 like head and the characteristic swelling near the tip of the tail. Sniear 

 material from Arvclius was unfortunately not available, so that I have 

 been unable to determine the length of the large sperms. If they vary 

 in size as in Euschistus, (and sections at hand seem to indicate that 

 they may), the large sperms would be one of the largest animal sperms 

 recorded.^ 



These results have an unexpected bearing on the attempts made 

 some years ago by Zeleny and his pupils to establish quantitative 

 differences in the length of the sperm head referable to differences in 

 the volume of the sex.(X and Y) chromosomes. As a matter of fact, 

 their plotted measurements of sperm heads did give a liimodal curve 

 in many cases, and this was held to be proof of the original contention. 

 Without undertaking a critique of this whole conception, it may be 

 pointed out that differences arising from polymegaly, such as I have 

 here described, were not taken into account; yet they would obviously 

 introduce a fatal source of error. Nor is it sufficient to say that be- 

 cause visible differences in the spermatocytes are not readily to be 

 detected, therefore, none occur in the resulting sperms; for in Mur- 

 gantia, to cite an instance, the spermatocytes are not always easily 

 separable on the basis of size, and yet the spermatic differences are 

 very considerat)le. It is easily conceivable that a pentatomid might 

 be found in which polymegaly of the sperm would be distinguishable, 

 and yet size differences in the spermatocytes might be to all intents 

 and purposes lacking. Some of the forms classed in Table I as failing 

 to show polymegaly may, indeed, belong to just such a tA'pe. Poly- 

 megalous differences of a minor type may possibly he much more 

 common than is suspected. As a matter of fact Zeleny and Senay 

 ('15) did measure the sperm of Euschistus variolarius, obtaining a 

 strikingly bimodal curve, in which, however, the ratio of difference 

 was verv much more than it should have been according to the 



7 The largest animal sperm on record seems to be that of Notonecla, which 

 also happens to be an hcmipteran. Pantel and de Sinety ('06, page 89) give 

 its length as 12 mms. and over. 



