GERM CELLS IN THE ARTHROPODA 133 
Montgomery found in one cell a mass of granules 
from which the rod may have developed (Fig. 41, A, 
X), and von Winiwarter noted that the rod had a 
granular appearance in the earliest stages he ex- 
amined. It is also perfectly distinct from the io- 
zome (see Fig. 41, B, I) and is apparently not 
directly derived from the nucleus. Von Winiwarter 
is not so certain as Montgomery regarding the history 
of the spermatogonia, the“ cristalloide de Lubarsche,”’ 
and the ‘‘ batonnets accessoires,” as he calls the 
rodlets. He was unable to decide regarding the 
number of spermatogonial divisions and believes it 
to be indeterminate. He finds, contrary to Mont- 
gomery, the rod persisting in fully developed Sertoli 
cells, and considers the fragmentation or ‘fission of 
the rod to form the primary rodlets as doubtful. 
Further investigations with more favorable material 
are very desirable, but notwithstanding certain 
differences of opinion between the two writers whose 
results have been briefly stated above, it seems cer- 
tain that Sertoli cells and germ cells are both derived 
from primordial germ cells, and that the Sertoli 
cells differ from the ultimate spermatogonia in the 
possession of a peculiar rod probably of cytoplasmic 
origin. Montgomery considers this a sort of secon- 
dary somatic differentiation (the Sertoli cells repre- 
senting the soma of the testis) ; the first somatic dif- 
ferentiation occurring when the tissue cells become 
differentiated from the germ cells in the embryo. 
Amitosis. Wilson (1900) defines amitosis as 
“‘mass-division of the nuclear substance without 
