THE DEVELOPMENT OF THE SPERMATOZOA, 421 
In places the degenerated blastophoral corpuscles are seen 
projecting radially towards the lumen. 
The next stage for description is that of the sperm-poly- 
plasts, which is reached about the end of July or the begin- 
ning of August. The testis at this stage is considerably 
swollen and very vascular, showing that growth is rapidly 
going on. ‘The minute blood-vessels are so full of blood 
that if the surface be examined under a simple lens the ter- 
minations of the crypts are marked out as hexagonal are 
by the blood-vessels. In a section (fig. 22) it is seen that 
the crypts contain large multinuclear masses which almost 
obliterate the lumen of the tabe. They are in various states 
of progress, the more mature forms projecting as pyriform 
masses into the crypt, and they are more or less held toge- 
ther by the interstitial cells, so that in making a teased 
preparation it is exceedingly difficult to free them from one 
another without breaking them up. 
Another important fact about them is that they are hollow 
vesicles formed of a single layer of cells, and it is towards 
the centre of the cavity that the spermatoblasts will elongate 
to form the spermatozoa. The elongation in this case is 
different to that in Helix, it’ is‘ centripetal, while that in 
Helix is centrifugal. At first it seemed a puzzle how 
from these hollow vesicles the spermatozoa could come 
to be arranged in bundles, each bundle connected with 
a nucleus and cell, for at first there was no evidence of a 
nucleus. . 
But on examining the sperm-polyplasts in fresh salt solu- 
tion, and staining with magenta or picrocarmine, keeping 
the mass under observation while the staining is going on, 
certain nuclei were seen to take up the staining fluid more 
rapidly than the rest, two or three to each polyplast, before 
the others showed more than a faint tinge, while the inter- 
mediate mass of plasma swelled up with the water of the 
solution, and became mapped into arez corresponding to the 
spermatoblasts (fig. 23). On more careful examination it 
was found that these nuclei were superficial to the rest, and 
were surrounded by a granular protoplasm; and in their 
general character and colouring seemed to show their con- 
nection with the blastophoral nucleus, which is described 
with the next stage. 
In sections stained with hematoxylin an interesting dif- 
ference of the nuclei of different polyplasts is observed, 
represented in figs. 27, 28, 29. This appears to be due to 
changes in the nuclear network, which seems to break up 
into an irregular mass before the nucleus forms the head of 
