392 E. R. Downıng, 
elongated bodies of which extend through the mass of interstitial 
cells to the mesogloea, thus forming supporting strands of protoplasm. 
Migration of interstitial cells. 
There might be another means conceived for the multiplication 
of the interstitial cells at the point of formation of the spermary. 
i. e. migration. This method of formation of genital organs is quite 
common in other of the coelenterates. Thus Harpy (29) finds in 
Myriothela phrygia that sex cells are modified interstitial cells which 
migrate to the point where the gonophore develops. In the medusae 
of Millepora murrayi he finds that the „sperm cells originate in the 
ectoderm of the coenosarc, wander into the ectoderm of the zooids 
where they fuse intoaggregations to form a spermarium”. IsHIKAWA (42) 
says sex cells in Æudendrium racemosum originate in the ectoderm 
and migrate to the endoderm. Examples might be multiplied. The 
sex cells are usually recognized by their amoeboid-like form and 
undifferentiated structure. WertsMANN (89) figures such cells in actual 
transit from.ectoderm to endoderm. No such cells have been observed 
in Hydra. The interstitial cells never appear amoeboid. THALL- 
wırz (80) says the sex cells may be recognized in transit in Sertu- 
larella polyzonias by the fact that they frequently display mitotic 
figures. Such a criterion will not apply in Hydra as other cells 
besides the interstitial display mitotic figures, and even if this were 
not so we could not tell that a certain sex cell in mitosis was 
migrating to an already formed spermary rather than initiating in 
situ a new spermary. 
To determine whether migration of the interstitial cells does 
occur in Hydra to form the accumulation that marks the spermary 
both longitudinal and cross sections of the Hydra with spermary 
were taken, and the number of interstitial cells noted in every 
36.5 u (20 divisions of the eye-piece micrometer) of the body wall. 
If there be a migration of the cells toward the spermary from other 
points on the hydra one ought to find the interstitial cells more 
numerous as the spermary is approached and less so the greater the 
distance from the spermary. This is not so. Taken then with the 
fact that no cells with amoeboid appearance are found, this seems 
sufficiently conclusive that the spermary is formed by multiplication 
of the interstitial cells in situ rather than their migration to the 
spermary from distant parts of the Hydra. 
It is to be noted that the basal portion of the Hydra is free 
