J. W. H. Harrison and L. Doncastbr 241 



of 69 or 70 (56 + 13 or 14) which would be expected if no pairing of 

 chromosomes took place at all. This is further confirmed by the fact 

 that there are constantly 12 — ^14 large ones. Hirtaria spermatocytes 

 have eleven large and two small, and the largest of zonaria are hardly 

 big enough to be classed in the '■' large " group. If, however, the larger 

 of the zonaria chromosomes paired with the smaller of the hirtaria 

 chromosomes, these would make additional large ones in the hybi'id 

 spermatocyte equatorial plate. Several, in fact, often appear to be 

 double, and as the number in the best figures adds up to 65, it may be 

 concluded that about five zonaria chromosomes (probably the largest) 

 pair with five derived from hirtaria, while most of the rest remain 

 unpaired. An examination of the prophase figures of the first spermato- 

 cytes just before the nuclear membrane disappears, shows, in addition 

 to several clumps of three or more chromosomes together, a large 

 number of single chromosomes of various sizes, and among them a few 

 which are paired, either equally or unequally (fig. 24). Further con- 

 firmation of the evidence that some of the chromosomes are paired in 

 the first spermatocyte division and that the smaller number counted is 

 not due to error, is found in the tact that almost exactly the full 

 number can be counted in the spermatogonial divisions, which are 

 smaller and less easy to examine accurately than those of the spermato- 

 cytes. If it is easy to count very nearly 70 in the spermatogonia, it is 

 hardly possible that the smaller number in the spermatocytes can be 

 duo to error. 



In the first spermatocyte division, it appears that not all the 

 chromosomes divide. Most undoubtedly do, but some show no signs of 

 division in metaphase when seen from the side of the spindle (fig. 25), 

 and the appearance of the secondary spermatocyte equatorial plates 

 confirms this supposition. The chromosomes in these second division 

 figures are less easy to count with complete accuracy, but it is not 

 difficult to get a fairly close estimate of their number, and this is almost 

 constantly less than in the first division (fig. 26). Counts have given 

 9 large and 42 small, 9 large and 46 small, 12 large and 42 small. In 

 one case, where I counted 12 large and 52 small, it is almost certain 

 that several had already divided, and that the halves were counted as 

 separate chromosomes. 



{h) Hirtaria % x zonaria ^ . This cross gives a preponderance 

 of females, with some males. The ovaries of full-grown larvae are 

 extremely small and difficult to find, and fi-om the three female larvae 

 sent me by Mr Harrison I only succeeded in getting one ovary. 



