ON SPERMATOGENESIS IN THE RAT. 
349 
ment from the spore-cell to the fully-formed spermatozoa 
would occupy four cycles, during which time four fresh gene- 
rations of cells would have been produced ; and so the process 
goes on. 
The first part of this process, i. e. the origin of the new crop 
of growing cells, which makes its appearance in the outer layer, 
is the most difficult to make out. These cells are first seen at 
the stage of fig. 9, and are produced by the karyokinetic division 
of larger cells, which are also found in the outer layer (this 
division by karyokinesis is represented in fig. 8, a"). The 
parent cells appear to be derived from the spore-cells by a 
process of division by budding, and not karyokinesis, but it is 
difficult to feel certain about this. The spore-cells apparently 
increase in size, and divide by a process of budding, for this 
seems to be the explanation of an appearance such as is repre- 
sented in fig. 7, a', as if a small outgrowth made its appearance 
from one part of the nucleus, increased until the two parts 
were about equal in size, and then separated off. Of the two 
cells which result from this division one divides by karyokinesis 
and produces growing cells, but the other remains in the 
resting condition as a spore-cell, destined to repeat the process 
in the following cycle, and thus perpetuate the production of 
spermatozoa in the tubule. 
During the time in which a new crop of small-growing cells 
has been produced, the growing cells of the previous generation 
have been increasing in size, and now, since the new cells have 
appeared between them and the wall of the tubule, come to 
form the second layer. The manner in which this growth takes 
place will be understood if they are followed through the series 
represented by PI. XXII, 1 figs. 1 — 10, b, b ' , and fig. 1, b". In 
fig. 1 these cells are all in the resting condition, and their nuclei 
present the ordinary appearance of resting nuclei, but very 
soon they begin to pass into the kinetic condition, the nuclear 
membrane disappearing and the chromatin becoming converted 
1 Figs. 3, 8, 9, and 10 are drawn on a somewhat smaller scale than the 
others. 
