J. W. H. Harrison and L. Doncaster 237 



usually shows no sign of doublcness, and on the spindle when seen side- 

 ways it sometimes {ippears to be attached to one half only of the large 

 one, as if it were going entire to one pole of the spindle. This, howevei*, 

 appears to be exceptional ; the compound chromosome is usually recog- 

 nisable only with difficulty in side view (fig. 9, a, b, c). If it were the 

 case that the small chromosome went undivided to one pole like a 

 heterochromosome, the second spermatocyte equatorial plates should 

 be of two kinds of equal frequency; half of them should have thirteen, 

 and half either fourteen, or thirteen of which one should be compound. 

 All these conditions in fact are found, but the great majority have 

 thirteen of which one may be more or less clearly composed of a large 

 and small member attached to each other with varying degrees of close- 

 ness. In some, perhaps the majority, only thirteen apparently simple 

 chromosomes are visible ; in others a large one clearly has a small one 

 attached to it, and in a few fourteen separate chromosomes may be 

 counted (figs. 11, 12, 13). In one of the latter class of figures, the 

 largest chromosome has clear indications of being compound although 

 there are thirteen separate chromosomes in addition to it, an arrange- 

 ment which is undoubtedly abnormal (fig. 14). 



Usually there are two chromosomes, in addition to the small member 

 of the compound one, which are conspicuously smaller than the rest, 

 but, probably according to the depth of staining and the position of the 

 chromosomes on the spindle, there is some variation, so that only one, 

 or sometimes three small ones, may be visible. It is thus not easy to 

 identify any one chromosome with confidence, but when fourteen are 

 visible, one very small one is usually, if not always, in close proximity 

 to one of the largest. It is clear from this account that equatorial 

 plates of the second spermatocyte divisions do not make it quite certain 

 whether one chromosome goes over undivided in the first division or 

 not. Unfortunately I have found only one case of a first division 

 anaphase in which the chromosomes of both groups can be counted 

 (fig. 10), and this is not entirely unequivocal. In each group there are 

 thirteen chromosomes, one of which shows signs of being composed of a 

 larger and smaller unit in each group. This would indicate that the 

 double chromosome divides equally, a large and a small portion going 

 to each pole. In one group, however, there is a body outside the group, 

 and at a different level in the section, which might possibly be a 

 chromosome. It is nearer the pole than the group of thirteen, and I 

 am fairly confident that it is not a chromosome but an extra-nuclear 

 body. Such occur regularly in the spermatocyte cells. 



