Further Studies on the Ovogenesis of Sagitta. 245 
second accessory cell (Fig. 9). At this: time both accessory cells 
show signs of degeneration and separation takes place between the 
accessory cells and the wall of the ovary (Fig. 9), or sometimes 
between the two cells (Fig. 10a and b), or between the second 
accessory cell and the egg. The epithelial cells on either side stain 
more deeply and appear to be denser than usual (Fig. 9 and 10a). 
The egg now pushes its may between the ovary and the sperm- 
duct, crowding the latter against the body-wall (Fig. 11). The 
epithelium of the ovary peels off from the wall of the sperm-duct 
leaving such smooth surfaces (Fig. 1, 11, 12) that one can hardly 
believe that the duct in which the ripe eggs lie is a temporary 
one, and that a short time before it was formed, sperm tubes had 
extended from the sperm-duct through the accessory cells to the 
oocytes; but there is abundant proof that this is true. In many 
cases eggs are-found at intervals either passing from the ovary into 
the oviduct or lying wholly in it, while between such eggs are 
portions of the ovary which show no sign of the oviduct. The 
opening from the ovary to the oviduct seems to close immediately 
behind the egg, and in most cases the eggs certainly make their 
way into the oviduct independently, and the opening bears a direct 
relation to the point where the egg breaks away. A few eggs were 
found lying free in the ovary with no opening formed, and it is 
possible that they would follow eggs with which they were in 
contact into the oviduct through openings already made by the 
latter. The appearance of the eggs in the sections indicates that 
each individual egg makes its way into the oviduct by means of 
its own activity, but to be sure of this point one must observe this 
phenomenon in the living animal. 
The number of eggs ripening in the two ovaries for one laying 
varies greatly. In the specimen of Sagitta bipunctata examined, there 
were three in each oviduct; in Sagitta elegans there were from 2 to 
9 and as great a difference between the two ovaries of one individual 
as 4 and 8. 
The Chromosomes. — In a few specimens of Sagitta bi- 
punctata, the chromosomes in the stage shown in fig. 13e and 15 of 
tab. 20 (1903) were split longitudinally, but the preservation of the 
material was poor and it was feared that the appearance was not 
normal. In Sagitta elegans, however, I find this splitting to be a 
normal condition at a certain stage of the reduction in the size of 
the chromosomes. Two such chromosomes are shown in Fig. 13. 
