MICHIGAN ACADEMY OF SCIENCE. 
101 
thread because of the many convolutions and their crowded condition in 
the nucleus. They are doubtless fewer in number than the autosomes 
in the spermatogonia, probably only one-half as many being present. 
It would appear from this that each loop represents two equivalent auto- 
somes of the spermatogonia. 
At a somewhat later stage the single spireme thread shows a double 
structure (Fig. 9) due to the splitting of the thread. This splitting has 
never been actually observed but the thread is distinctly double in places 
in the late growth stages. Each granule of one thread is opposite a 
similar granule of the other thread, making it appear as though each 
thread were made up of halves of the granules of a single spireme. The 
splitting is not distinct throughout any great length of the thread at one 
time, due to the convolutions of the thread and to the small size of the 
chromatin granules. No evidence of a side to side union of the spireme 
threads as described by Otte (1906) for Locusta was observed. Only a 
few cases were observed in which there was even an apparent side to side 
union of the threads, and these seem to be in all cases accidental ; two 
threads may lie side by side for some little distance but always diverge 
if followed far enough. The fact that there is a polar arrangement of 
the loops would necessitate the parallel position of some of them. It 
could not be determined whether the spireme was a continuous thread 
or was segmented. The complex course of the thread and the large 
amount of chromatin contained in it make it impossible to trace, with 
certainty, for any distance. 
At no time was there a massing of the loops at one side of the nucleus. 
Since there has been so much discussion and disagreement on the part 
of investigators concerning this point, it might be interesting to add that, 
in none of the properly fixed material did such a stage occur; however, 
in one slide prepared from a poorly fixed testis, this stage was very 
conspicuous, in fact, a large number of the nuclei which were in growth 
stages showed this condition. From these two facts it would seem that 
in Ccuthophilus latebricola at least, there is no synizesis stage. 
During the growth period the accessory chromosome appears as a 
deeply staining oval body against the nuclei’ wall. It can be distin- 
guished from the nucleolus throughout the growth period by its position 
and its oval form. In every case the accessory chromosome lies against 
the nuclear wall, whereas the nucleons may take up any position within 
the nucleus (Figs 6, 7, 8, 9). 
TETRAD FORMATION : 
DIVISION OF THE PRIMARY 
SPERMATOCYTES. 
At the close of the growth period tin 1 loops of the spireme begin to 
break up and become situated around the periphery of the nucleus. The 
early tetrad formative stages show the spireme broken up into segments 
of unequal length. When the thread begins to segment in this way it is 
possible to recognize the quadripartite nature of the future chromosomes 
(Fig. 10). Some of the broken segments may stretch entirely across the 
nucleus, whereas others are not more than one-fourth of this length. As 
development goes on the segments gradually condense but stiil retain 
the characteristic form. Throughout the entire period of tetrad forma- 
tion it is possible to trace the general shape and quadripartite character 
of the chromosomes. 
