MUNSON: SPERMATOGENESIS OF PAPILIO. 
85 
As the cell is about to constrict, however, it assumes a more spherical 
shape (pi. 16, figs. 112, 113). 
The spermatocytes second order thus formed have a more spherical 
form (pi. 16, fig. 117), and do not remain so definitely related to the 
surrounding cyst wall. This is especially true of the small forms of 
spermatocytes, which are by far the most common (pi. 13, fig. 31). 
The above account of the maturation divisions is based on the 
giant spermatocytes (pi. 13, fig. 35-38). But I have the best of reasons 
for assuming that the processes in these large ones do not differ from 
those in the ordinary spermatocytes. They, too, show an eccentric 
nucleus and a similar orientation of the spindle; they show the spireme 
thread, the early development of the asters, and a similar disposition 
of the chromosomes (pi. 13, figs. 29, 30, 31). I have given most 
attention to the giant spermatocytes for the obvious reason that the 
more difficult details are there more easily made out. The small 
spermatocytes are considerably more crowded. 
Summary and Conclusions on Maturation. 
1. The centrosome of the growing spermatocyte of Papilio rutulus 
appears as a clear sphere, lying close to the nucleus, causing an inden¬ 
tation in it. In Grapta silenus it is more variable in position being 
sometimes close to the nucleus, sometimes halfway between it and 
the cell periphery. There are two little granules lying side by side. 
2. These granules separate and move to opposite poles of the 
nucleus. 
3. Asters are then developed out of the general cytoreticulum, 
the astral rays extending in all directions to the very periphery of the 
cell. 
4. At the equator, between the two asters, the astral fibers cross 
one another, suggesting that they are fibers instead of compressed 
alveoli. 
5. There are reasons for suspecting that not only do the centro- 
somes divide and separate, but that there is also a division of the 
fibers of the reticulum and cytomicrosomes, the division of the centro- 
somes being merely the first of a long series of divisions in both cyto¬ 
plasm and nucleus (fiber and granule) which finally results in the 
formation of two cells from one. 
