"394 E. R. Downing, 
mine, as there have not yet been discovered any constant land- 
marks in the cell. In the flattened dodecahedral cells the nuclei 
partake of the change in shape of the cell measuring on an average 
72%xX6%X 54 u. The structure of the spermatogonia is like that 
of the ectoderm cells, except that the protoplasm is not vacuolate. 
Division of spermatogonia. Prophase. 
As division approaches the nucleolus fragments into two or 
more parts and disappears (Fig. 31). The microsomes at the inter- 
section of the yet indistinet karyoplasmic threads become quite 
prominent, giving to the nucleus a very granular appearance This 
phase lasts for a comparatively long time and at the base of the 
spermary the majority of the spermatogonia of the first generation 
will there be found in this condition. Then the thread connecting 
the microsomes become more apparent and constantly coarser, while 
the microsomes themselves are reducing in number (Fig. 32, 33). 
This process of reduction continues until forty-eight are to be noted, 
joined by coarse strands of the karyoplasm (Fig. 34). The chroma- 
tin seems to be almost entirely collected in these forty-eight micro- 
somes and since they persist to form the chromosomes we may 
cease speaking of them as microsomes and designate them the chro- 
momeres. All but one of the connecting linin bands between any 
two chromomeres dissapear giving thus a continuous thread on which 
the chromomeres appear bead-like. During the changes mentioned 
above the nucleus seems to hold no constant shape. It now however 
usually becomes an ellipsoid of revolution with the long axis at right 
angles to the line connecting the poles at which the centrosomes 
will appear. 
Spireme. 
The thread of chromomeres now becomes a spiral on the nuclear 
wall, the axis of the spire coinciding with the major axis of the 
nucleus (Fig. 35). It is diffieult to count so large a number of 
chromomeres with absolute accuracy, and I cannot be absolutely 
sure of their number or the arrangement during this prophase of 
the spermatogonic division. There is no doubt, however, that during 
the division of the spermatocytes there are twenty-four chromomeres 
usually, and that the spiral thread makes three complete turns of 
the nucleus. So far as it has been possible to count accurately, 
there have been found forty-eight chromomeres in the spermatogonia, 
and the thread makes six complete turns of the nucleus The nu- 
