354 
HERBERT H. BROWN. 
ternal should become a supporting cell, while the others all 
develop into spermatozoa. 
3. This improbability is rendered still more glaring by a 
comparison with the corresponding elements in the testes of the 
hedgehog and other animals, for in the hedgehog the nuclei of 
the supporting cells and of the young spermatozoa are remark- 
ably dissimilar. 
4. Nuclei of supporting cells cannot be seen in the lumen of 
the tubule, or in the semen from the vas deferens or epidy- 
dvmis, and the nuclei which have migrated inwards show no 
signs of disintegration. 
5. In the stage of development immediately succeeding this 
(vide fig. 2) nuclei of supporting cells are seen, which are 
quite as large as those in the present stage. 
On all these grounds, then, it is impossible to adopt this view 
of the origin of the connection between the supporting cells 
and the spermatozoa. Probably the migration of the nuclei 
into the midst of the young spermatozoa has something to do 
with bringing about this connection, and having accomplished 
this the nuclei return to the outer layer. 
The grouping of the young spermatozoa becomes more evi- 
dent as soon as the preceding generation of spermatozoa with 
the seminal granules has left the tubule (this stage is repre- 
sented by fig. 2). Each group contains about ten or twelve 
spermatozoa. At this time a curious appearance is presented 
by the supporting cells themselves; large globules, which stain 
black with osmic acid, are seen in the protoplasm near the 
nuclei; these globules are not simply fat-globules, since they 
stain very darkly with gentian violet, and in chloride of gold 
preparations become quite black from the great affinity which 
they have for the metal (vide fig. 13). Iu sections stained 
with hcematoxylin vacuoles are seen in the protoplasm of the 
supporting cells, each of which contains a spherical body, which 
is slightly stained by the reagent; so that they would appear 
to consist of a mixture of fatty and albuminoid material. In 
many cases these bodies are found indenting the nucleus of the 
supporting cell, and sometimes look as though they were 
