92 
CHARLES L. BOULENGER. 
are numerous small, interstitial cells and stinging-cells in all 
stages of development. The fate of these stinging-cells is 
not very obvious ; there are no functional nematocyst-batteries 
on the manubrium, and the cells do not seem in process of 
migration to other parts of the body; as far as I can make 
out the stinging-cells are for the supply of the medusa-buds 
in much the same way as in Moerisia, where the developing 
medusa-buds receive a supply of nematocysts from the ecto- 
derm of the parent hydroid (5). 
(iii) The Proximal Region. 
The proximal region or stomach proper consists of much 
lower cells than the other two, and therefore appears thin in 
sections. The ectoderm resembles that of the oral region, 
although interstitial cells occur less sparingly; the structure- 
less lamella is distinct, although not of great thickness. 
The endodermal lining resembles that described by Lan- 
kester (8) and Gunther (7) as occurring in a similar position 
in Limnocodium. It consists of a single layer of large, 
almost cubical cells (fig. 6), which are of two kinds — (a) large 
vacuolated cells, and ( b ) gland-cells. The former contain 
conspicuous food-balls (/. b.) and foreign particles in the 
vacuoles, and it is evidently in this region that the greater 
part of the digestion takes place. The gland-cells ( gl . c.) con- 
tain coarge granular particles, the contents being very unlike 
those of the goblet-cells in the oral region, from which they 
also differ in being broader and in not tapering at the base. 
The endodermal covering of the mass of jelly which pro- 
jects into the manubrial cavity from the dorsal side consists 
of a simple epithelium of smallish cells. 
5. The Method of Nutrition. 
The problem of the nutrition of Limnocnida has given 
rise to some speculation. Gunther, in his account of the 
anatomy of the medusa, described no gland-cells of any kind 
