376 Gates and Rees. — A Cytological Study of 
between them, because that would mean that the body is composed of 
four consecutive segments which remained attached and in which the end 
segments must have then folded back on the others. This procedure is 
unlikely, and it seems more probable that the end to end coalescence took 
place after the chromosomes of two loops had come to lie side by side. 
In certain other cases (see Fig. 48) where only eight bivalent 
chromosomes were present in diakinesis, probably two of the shorter 
bivalents had fused end to end, but this could not be determined with 
certainty because two short chromosomes arranged end to end would be 
indistinguishable from a single bivalent of intermediate length. These 
fusions will be compared later with others observed on the heterotypic | 
spindle. We have no positive evidence whether the fusions at this stage 
are temporary or would persist and affect the later chromatin distribution, 
but probably the latter would be the case. 
Heterotypic Mitosis. 
■ 
Between the stage of diakinesis and the arrangement of the chromosomes 
on the heterotypic spindle, the chromosomes become much more condensed. 
In the heterotypic metaphase they are condensed to such a degree that 
there remains comparatively little observable difference in length between 
them, and many of them appear globular or but slightly elongated. The 
differences in length during diakinesis are not, however, merely a result of 
different rates of condensation, for the longer chromosomes undoubtedly 
contain much more chromatin than the shorter. That the differences in 
length of the chromosomes are constant features is also shown by the 
somatic mitoses, in which the same graded series of lengths appear 
(Figs. 63 and 64). 
A striking feature of the heterotypic metaphase is the frequent 
coalescence (more or less complete) of two or four of the chromosome 
bivalents, so that only eight or seven bodies appear on the heterotypic 
spindle. Probably over 50 per cent, show fewer than nine bivalents, seven 
or eight bodies being the most common number, but sometimes (Fig. 57) 
no more than five bodies are present. This is in cells in which the figure 
has not been cut, and must be due to a further coalescence of chromosomes. 
Such cases were the first to be observed, and it was only later that the 
cause of this apparent diminution in chromosome number was discovered, 
through the observation of cases in which the nine bivalents were clearly 
distinct, and also stages in their fusion. Fig. 50 shows an equatorial view 
with only seven bodies, and Figs. 53 and 54 side views with eight. In 
Fig. 50 the two bodies {a) formed by the end to end fusion of two bivalents 
can be determined. But in Fig. 54 particularly the coalescence is so 
complete that there is not the slightest indication as to how it has taken 
