THE SPERMATOGENESIS OF DOMESTIC MAMMALS, 379 
irregular (Fig. 8). When the cells have thus reached, or are about to reach 
the maximum size (about twice as large as the spermatogonia in the earliest 
generation), the nuclear wall begins to constrict at one side where a cytoplasmic 
mass is always situated (Figs. 9, 10). This cytoplasmic mass is apparently 
similar to that described by vom Rarx (’93) in the salamander. Soon a cleft 
appears at this constricted point (Fig. 10) which continues to deepen, until 
a complete division of the cell is effected (Figs. 11—14). The same type 
of amitotic division was reported by McGrecor (’99) in the spermatogonia 
of Amphiuma, but Parrerson (08) has established two types of amitotic 
division in the pigeon’s egg; in one type the nucleus elongates, after which 
a constriction in the nuclear membrane appears on the entire circumferenec 
of the nucleus; in the other type a nuclear plate is laid across the nucleus 
and the division consists in a splitting of this plate. He added that besides 
these the modified form of the first type occurs. In this type the constriction 
of the nuclear wall proceeds from one side, and the nucleus usually does not 
elongate previous to the appearance of the constriction. 
The method of amitotic division in cattle is quite similar to this 
modified fori. 
At the early stage of this division, the cytoplasmic granules are gathered 
together and appear as a granulated mass, which comes to lie on the concave 
side of the nucleus (Figs. 9, 10). Sometimes, however, the mass surrounds 
the nucleus more or less completely. When a trace of fission has appeared 
on the nuclear wall, this granular mass draws together to form a rounded 
sphere in which no other substance is to be seen (Figs. 10, 12). The 
centrosphere of ring form such as that described by Mrves (91) in salamander, 
can not be found, but the form and behavior of the mass is quite com- 
parable to the sphere described by vom Rara (795). After the nucleus has 
divided into two, the sphere again begins to disintegrate into the granules 
and is placed between the two daughter nuclei (Figs. 14, 16). From the 
behavior and the position of the sphere, it seems more probable that the 
sphere might play an important réle in the amitotic division, independent of 
centrosomes. A view similar to this was put forward by Meves (’91) in the 
amitotic division of the spermatogonial cells of Salamander, where he says: 
“ Betrachtet man dagegen ihre Beteiligung bei demselben Vorgang in den 
