The Spermatogenesis of tlie Opossum (Didelphys virginiana) etc. 
47 
that this granule is really the centrosome now visible by reason of its 
conspicuous location. Around this granule the future sphere may again 
form and in it the centrosome again appear to vanish. 
I 
d) Late Postsynapsis. Origin of Chromidia (mitochondria; 
chondriosomes). 
ln late postsynapsis (or early prophase) the above described pale, 
mossy, bivalent cliromosomes pass into a lightly-staining semireticular 
phase (figs. 22 to 24). This immediately precedes the typical prophase 
as indicated by the division of the centrosphere (figs. 23 and 24) and 
coincides with the time of origin of chromidial elements (mitochondria). 
These bodies appear to be true chromidia in the sense defined by Hert- 
wig (1902), Goldschmidt (1904) and Popoff (1907) since they are nuclear 
products (probably transforming chromatin) passing from nucleus to 
cvtoplasm. Tlieir appearance is coincident witli the loss in staining 
capacity of the prophase chromosomes — a significant fact from the 
standpoint of the question of origin. They are distinctly of a lipoid nature 
as revealed by their black color in unstained preparations of Flemming — 
fixed material. The degree of color is intensified by the iron-haematoxvlin 
stain. 
After passing out of the nucleus, the chromidia are grouped in single 
file for a while on the nuclear membrane (fig. 24). They appear compact 
at this stage, have a sharp contour and a spheric or short dumb-bell shape. 
It needs to be emphasized that dnring this process the chromosome- 
nucleolus (accessory chromosome) does not loose its intense staining 
quality, nor decrease in size, nor show any evidence of a loss of its material. 
Furthermore, the chromidia — eontrary to the behavior of typical mito- 
chondria described for various forms e. g. Euchistus (Montgomery 1910); 
guinea-pig (Duesberg 1910) — are not gathered about the centro- 
sphere any more frequently than elsewhere in the cytoplasm (figs. 25 
to 55). Chromidia are present in all succeeding stages up to the time 
of formation of the spiral filament of the middle-piece of the Sperma- 
tozoon. Their fate in this connection will be discussed in detail in a 
later section. 
Chromidia are produced anew in small amounts during the resting 
stage of the secondary spermatocyte and in the early spermatid stage. 
But whether these last formed (and their descendants), or persisting 
representatives of those earlier formed; or whether representatives of 
both sets (the more probable condition judging from analogous examples) 
