76 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
cysts, it has already been pointed out that a possible connection may 
exist between these giant spermatocytes and the stalked cells (compare 
the series, pi. 15, figs. 68, 69, 77, 83, 84, 85 and figs. 86, 87, 88, 89, 78, 
79, 80, the former being free cells, possibly originally stalked like fig. 
68, the latter, giant spermatocytes in process of maturation). 
Aside from their origin, the giant spermatocyte and the ordinary 
smaller ones differ in no particular except size. The end product, 
too, turns out to be identical. I have discovered no difference in the 
spermatozoa, but the spermatids differ greatly as to size. 
Because of their size the giant spermatocytes are especially favorable 
for the study of the phases of mitosis and reduction, as well as for the 
various problems connected with the transformation of the spermatid 
into a mature spermatozoon. 
Spermatocyte First Order. 
When the spermatocyte begins to grow, its relation to the other cells 
of the cytocyst changes. For, whereas during the multiplication 
period, the spermatogones form a solid cyst like the morula of a seg¬ 
menting egg, the spermatocytes form a hollow cyst resembling a blas- 
tula. Under the heading “Spermatocysts ” a physical explanation 
of this has already been suggested. All the spermatocytes assume 
a definite relation to the surrounding cyst cells. Each spermatocyte 
now has one pole in contact with the cyst wall, the opposite pole free and 
next to the center of the cyst. On all other sides they are subjected 
to pressure from neighboring cells. Consequently they are elongated 
radially. Transversely to the long axis the outline of the cells is 
similar to that of the spermatogone in some phases of its development 
(pi. 15, fig. 70). In the radial axis, they show a flattened base next 
to the cyst wall, and a free rounded side next to the lumen of the 
cyst (pi. 13, figs. 35, 36). At this smaller, free pole the spherical 
nucleus always lies (pi. 13, figs. 35, 36, 37). 
The chromatin network resembles that of the spermatogone at 
rest, but the arrangement of the chromosomes is not so definitely peri¬ 
pheral. An irregular, deeply staining body is also present here. It 
has a somewhat variable position near the center of the nucleus. 
Relatively, the cytoplasm increases more than the nucleus though 
the latter also exceeds the nucleus of the spermatogone (compare 
