IQI5. | J. RrrcewiE: Aydroids of the Indian Museum. 561 
indicates that only one form, long and bean-shaped, is present in 
Gonionemus. It is possible that these resemblances points to the 
true relationship of the problematical Haleremita: that it is the 
metagenetic hydroid phase of a hydrozoon medusa, a larva which 
in due course will assume medusa form. It is interesting to find 
some confirmation of this view in the simplicity of structure (to 
be expected in a larval form) which has led to the relegation of 
Haleremita to the primitive family Hydridae; yet the bud-forma- 
tion in the two is by no means identical. 
Perkins describes some interesting features in the develop- 
ment of the Gonionemus hydroid buds. The buds, which occurred 
singly on the hydroids, arise about halfway between the base 
of the polyp and the ring of tentacles. During their early growth 
the endoderm is solid, and in this condition becomes isolated from 
the endoderm of the polyp by the gradual constriction of the 
ectoderm at the junction of the two. Finally the bud comes to 
be attached simply by a long thin neck of transparent ectodermal 
protoplasm. The release of the bud, in the only case followed 
throughout its complete development, was accomplished by the 
gradual stretching and final rupture of the ectodermal neck. The 
released bud settled down upon its former free or distal end, and 
at the other pole, formerly attached, a mouth and tentacles de- 
veloped. This bud became attached near the parent polyp, but in 
most cases an escaped bud was discovered after a few days some 
distance from the parental form. During the interval ‘‘it seems 
probable that it is a creeping unciliated form, although my first 
conjecture that it was a ciliated planula has not been proved 
erroneous”’ (Perkins, 1903, p. 771). A general idea of the develop- 
mental period of such planula-like buds can be gathered from 
Perkins’ observations. The development of a bud from its first 
appearance as a simple knob to the completion of the formation 
of the coelenteron and the appearance of tentacles, lasted from 
ten to fourteen days, distributed as follows:—‘‘ (a) the first period 
including as far as the detachment of the bud, 5 days; (b) motile 
form, 2 to 5 days; (c) from attachment to appearance of tentacles, 
3 to 5 days (Perkins /oc. czt.). Schaudinn found that the develop- 
ment of Haleremita buds, up to the point of escape from the 
parent. varied from 5 hours to 6 days (Schaudinn 1894, p. 
230). 
In all the cases above mentioned, as well as in that of our 
Indian form, the buds arose equally from ectodermal and endo- 
dermal elements, confirming the observations of Braem (1894) and 
contrasting with Lang’s (1892) description of the purely ectodermal 
origin of Hydroid buds. 
So far as can be determined from my examination of the 
comparatively few buds available in the Indian species, they 
agree most closely with those of Haleremita cumulans. In both 
species, in contradistinction to Gonionemus, the bud possesses a 
hollow structure from the beginning, and the internal cavity 
remains in connection with the coelenteron till the time of escape. 
