ON SPERMATOGENESIS IN THE RAT. 
353 
cells and young spermatozoa at this stage (fig. 1) are difficult 
to explain. 
At this stage some supporting cells are seen which contain 
a more or less conical nucleus, situated in the outer layer 
upon the wall of the tubule, from which a protoplasmic strand 
extends inwards through the third layer of cells as far as the 
granules among which the heads of the spermatozoa are 
embedded ; but the nuclei of other supporting cells appear to 
be pushing their way towards the lumen of the tubule ; they 
are elongated in a radial direction, and may be seen between 
the second and third layers, and even in the middle of the third 
layer of cells, but farther inwards than this I have never seen 
them. In many cases the nuclei are seen to be connected to 
the outer layer by protoplasm, so that probably this is a 
migration of the nucleus and not of the entire cell (figs. 1 and 
1, a). At the same time some of the young spermatozoa 
appear to move in the opposite direction towards the wall of 
the tubule, so as to occupy the position in the outer layer 
vacated by the supporting nucleus (fig. 1, a). These appear- 
ances at first appeared to me to justify the supposition that 
the supporting cells which have finished their work, having 
served for the support of the crop of spermatozoa which has 
just been discharged, are now in their turn being cast off into 
the lumen of the tubule, there to undergo disintegration, and 
that the cells which are passing outwards towards the wall of 
the tubule are destined to become the new supporting cells, 
retaining their connection with their brother cells, which 
develop into spermatozoa j so that according to this view the 
supporting cell and the group of spermatozoa which it sup- 
ports would result from the division of a single cell. 
But there are numerous objections to such a view as 
this : 
1. The young spermatozoa at this stage are apparently free, 
and not connected together in groups. 
2. It is very difficult to believe that, of a group of cells which 
are all exactly alike, and are produced by the karyokinetic 
division of a single cell, that which happens to be most ex- 
