SPERMATOGENESIS OF EUSCHISTUS 135 
smallest is the smaller idiochromosome (d). The autosome pairs 
present constant size and form differences, and at least five pairs 
(A, a, B, b, C, c, E, e, F, f) may be readily distinguished from the 
others. But of the three remaining chromosomes, contrary to 
my conclusions of 1906, I do not find it possible to decide which 
two are autosomes and which one is the larger idiochromosome, 
but that one of them is necessarily the larger idiochromosome we 
shall see is unquestionable. All of the twelve autosomes divide 
in every spermatogonial mitosis, for each first spermatocyte 
invariably comes to contain twelve, as detailed especially in my 
paper of 1910. ; 
Fig. 4 exhibits an early telophase of an ultimate spermatogo- 
nium. The chromosomes are here not fused, as the figure might 
imply, but merely so closely apposed that they can be distin- 
guished only with difficulty in a side view. 
The daughter cells resulting from this division are the first 
spermatocytes, and these are shown in their earliest stage in 
fig. 5. In the upper of the two cells there drawn, fourteen chro- 
mosomes can be counted, and in the lower, thirteen; in the latter 
it is probably the smaller idiochromosome that is hidden from 
sight. 
The next change is that the nucleus becomes larger while the 
autosomes grow more irregular in form, become less compact and 
swell in size, as shown in figs. 6 and 7, each of which exhibits 
fourteen chromosomes. A few delicate linin threads appear 
joining autosomes, but the connective fibers of figs. 4 and 5 
(derived from the linin constituent of the chromosomes) have 
disappeared without a trace, and so have the granules of the cell 
plate. Then the autosomes begin to lengthen (figs. 8 and 9), 
showing sometimes amoeboid contours (figs. 10 and 11), and 
soon become linear (figs. 12 and 13). The precise succession of 
the stages of figs. 6 to 12 is difficult to determine, but without 
doubt there transpires a gradual lengthening of the chromosomes 
at this time without either splitting or conjugation. Figs. 13- 
15 represent succeeding stages of this lengthening of the auto- 
somes, and there can be no doubt that the stage of fig. 12 passes 
