The spermatogenesis of Hydra. 403 
migrate to a position at one end of the nucleus as figured in Fig. 61 
and 49. But it seems more likely that it retains its place, the cell 
elongating in the polar axis. For the customary division of the 
centrosome would then bring the daughter centrosomes to the poles 
of the new cell. (Compare Fig. 49 and 61.) 
Formation of the spermatozoon. 
The spermatozoon is formed out of the nucleus of the spermatid 
by the absorption of the cell body by the nucleus and formation of 
the various parts of the spermatozoon by a process of growth. 
Immediately about the nucleus appears a transparent film which 
increases in breadth. Apparently the cell body next to the nucleus 
is becoming liquified. At one end of the nucleus, the future tail 
end of the spermatozoon, there appears a tiny droplet of clear sub- 
stance, the nucleus becomes cylindrical with the growing droplet at 
the posterior end and a conical point at the anterior end. When 
about one half the cell body has been liquified, the droplet has 
nearly attained its maximum size and has a diameter about one- 
third that of the minor axis of the cell. The nucleus has now in- 
creased in length until it is as long as the cell, the pointed tip of 
the nucleus is the anterior end of the spermatozoon. The posterior 
edge of the dome-shaped droplet which is the middle piece, touches 
the cell wall. The unliquified cell body remains now about the 
forming spermatozoon as a cylinder, the longitudinal section of 
which will give on either side a concavo-convex section, the con- 
vexity (exterior) corresponding to the surface of the cell (spermatid). 
During this process the nucleus lying free in a fluid often shifts 
its position. Thus in Fig. 66 its long axis is at right angles to the 
position it first held in the polar diameter of the cell. 
At about this time the tail rudiment appears as a growing 
point at the posterior end of the middle piece. At the outset the 
tip of the tail is somewhat enlarged, but it soon tapers to a 
point. I am unable to determine whether the cell wall is carried 
out aS a membrane over the growing tip or not, but I have no 
evidence that it is and believe that at the point where the limiting 
membrane of the middle piece and the cell wall touch there is a 
fusion. It is from this point that the tail outgrowth proceeds. As 
the tail grows the middle piece decreases in size. The liquefaction 
and absorption of the cell body and wall continues until all has 
