98 
CHARLES L. BOULENGER. 
dermal cup is distinctly two-layered and (b) the entocodon 
is approximately square in section, and that this causes the 
walls of the double cup of endoderm to be brought into con- 
tact with one another opposite the four corners of this organ, 
thus leaving four distinct pouches in a perradial position 
opposite the sides of the square mass of hollow ectoderm 
(Text-fig. 2, 2a). A longitudinal section through an inter- 
radius would give the erroneous impression that the endoderm 
was solid. 
The two layers of the endodermal cup fuse interradially to 
form the endoderm lamella, the four perradial pouches 
persisting and giving rise to the radial canals of the older 
medusa-bud. 
The next stage in the process of bud-formation consists in 
the proliferation of the superficial layer of ectoderm to form 
a solid plug, as described by Gunther ; I could see no signs 
of a second invagination of ectoderm such as Moore des- 
cribed. The cells of this solid mass very soon arrange them- 
selves in a single layer round a cavity which appears in their 
rpidst; those on the proximal side of this v cavity form the 
exumbrellar ectoderm of the future velum; those on the distal 
side persist as a superficial layer of smallish cells, which cover 
over the mouth of the umbrella until just before the liberation 
of the young medusa (Text-fig. 3, and PI. 15, fig. 8). The 
ingrowth of the above-described mass of apical cells causes 
the distal wall of the entocodon to become somewhat flattened, 
but I have never seen one wall pushed into the other so as to 
cause the organ to assume a cup-shape, with the consequent 
obliteration of the cavity; I cannot help thinking that 
Gunther must have arrived at this conclusion from the exami- 
nation of tangential instead of radial sections of the medusa 
buds. 
The four perradial pouches, as mentioned above, form the 
rudiments of the radial canals; these grow out centrifugally 
and form the endodermal axes of the first four tentacles ; the 
latter appear in the cavity on the distal side of the velum, 
and their axes are, of course, surrounded by the ectoderm 
