CYTOPLASMIC INCLUSIONS OF THE GERM-CELLS. 419 
ship to the centrosome it is impossible to say. It is certain 
tli at either the mitochondrial granules are not of the same 
size, or they have among them other larger granules of a 
different nature ; but it is always possible to detect larger 
masses here and there among the smaller granules. It is 
these larger masses which make it so difficult to be certain 
of the identity of any given body in a primary germ-cell of 
either sex, and though the granule marked M in PI. 23, fig. 1, 
is probably the micromitosome, it might quite possibly be of 
another nature. When the primary germ-cell is dividing, the 
mitochondria become disposed on one side of the spindle, as 
in PI. 23, figs. 2 and 3, and in PI. 24, fig. 21. At this stage, 
and for some time afterwards, especially in the female, the 
mitochondria resemble in shape rods rather than spheres or 
grains. 
Viewed at the metaphase, the mitochondrial cloud is found 
to form a halo surrounding from one-third to two-thirds of 
the surface of the equator of the spindle ; it never, or very 
rarely, forms a complete circle around the amphiaster, as the 
mitochondria often do, in the spermatocyte divisions (PI. 23, 
fig. 11). It is impossible to say with certainty whether the 
amphiastral rays are concerned with the division of the mito- 
chondria between the daughter cells, but I am inclined to 
think that they are not. In fact, mitochondria always seem 
to. clear a path for the astral rays instead of being directly 
caught up in them, and my observations seem to favour the 
view that mitochondria are partly distributed by cytoplasmic 
currents. 
PI. 23, fig. 3, shows a later stage in division; the distribu- 
tion of the mitochondrial matter seems to have been equally 
carried out, and I do not remember having seen any stage 
of division in which one cell appeared to be receiving more 
than its share. In all the spermatogonial divisions the cells 
act in the same way, and after these mitoses are finished the 
secondary spermatogonium about to become a spermatocyte 
possesses a cloud of mitochondrial granules at present dis- 
tributed towards one side of the cell. This side is generally 
