138 TRANSACTIONS OF THE [FEB. 12, 
the spermatozoan are one and the same substance (see figures 6- 
11). Furthermore, similar preparations show that the archoplasm 
mass can be uninterruptedly followed from its position in the 
resting stage of the early germ cell to its position at the poles 
of the spindle during the prophase of karyokinesis (see figures 
1-4). Iam unable to state as yet, however, whether the archo- 
plasm mass disappears or not in the stages between the pro- 
phase, as in figure 4 and the anaphase of karyokinesis, as in 
figure 5. 
In the early stages of spermatogenesis the archoplasm mass 
apparently plays an important part in cell division where its 
actions are extremely suggestive, to say the least, of those of 
an attraction center. 
In its first phase of activity it appears as a homogeneous 
sphere at the extremity of the cell furthest from the center of 
the spermatosphere and rests against the nuclear membrane (figure 
1, A=archoplasm mass). It then divides into two equal parts, 
and in some cases a minute black dot is visible in the center of 
each part. I take this to be the true centrosome (figure 2). 
Both halves then pass around the nucleus through an angle of 
90° and the spindle is formed with its long axis tangential to the 
periphery of the ball (figure 3). When the spindle is nearly 
completed the archoplasm masses can be seen, in many cases, to 
extend part way up the spindle fibres (figure 4). In the later 
stages of karyokinesis the same masses can be seen collecting 
at the poles (figure 5). One is almost tempted to conclude 
from these phenomena that the achromatic spindle fibres come 
from and are retracted into the archoplasm masses at the two 
poles. After formation of the daughter cells the archoplasm 
masses are in the form of more or less flattened plates closely 
applied to the new nuclear membranes. (Figures 6,7 and 8, which 
represent the spermatid development, may apply equally well to 
the earlier germ cells, so far as the archoplasm masses are con- 
cerned.) These masses are next seen in this same shape at the 
farther extremity of the cell. Whether they move to these 
positions or are so placed by rotation of the nuclei, I am un- 
able to state. 
In the spermatid the archoplasm mass occupies a position in 
the cell as shown in figure 8. In form it is a concave dise 
lying closely applied to ‘the nuclear membrane. From its posi- 
tion here, it is a comparatively simple matter to trace the for- 
mation of the middle piece. 
The spermatozoan is formed from the spermatid by elonga- 
tion of the latter cell in the direction of its radial axis. The 
nucleus of the spermatid, which has the appearance of a compact 
