400 THE AMERICAN NATURALIST. [Vor. XXXIV. 
ical form, hard to distinguish from the freshly discharged ovum, 
almost all surface aspects of cleavage having disappeared. In 
this form it remains apparently quiescent for some hours, dur- 
ing which time, however, internal cell division goes on quite 
actively, as sections clearly show. With completion of this 
internal cleavage the embryo becomes a solid morula, with 
only the faintest indications of any differentiation into an ecto- 
derm (cf. Pl. IV, Fig. 6). Following this, however, the spe- 
cialization of an ectoderm soon takes place, and the embryo 
begins to assume the characteristic pyriform, or oval, shape of 
hydroid planule. Within from twelve to twenty hours cilia 
make their appearance over the ectoderm, and the free life of the 
larva is assumed. Up to this time, however, and for some time 
after, no definite endoderm has been formed; the entire mass 
of internal cells seem scarcely distinguishable from each other, 
except that near the central portion the remains of yolk débris 
are more or less apparent. It is not till after some hours of 
larval life that an endoderm is gradually specialized from the 
internal cell mass and takes on an appearance quite similar to 
that of the forming ectoderm. After the establishment of the 
diploblastic condition there still remain a mass of undifferen- 
tiated cells, intermingled with yolk granules, which seem gradu- 
ally to disintegrate and are consumed as food by the developing 
larva, which, up to the polyp stage, is wholly without mouth or 
other means of taking solid food, though in all probability 
absorption of water with soluble matter in small proportion 
takes place. 
The larval history of Pennaria seems considerably longer 
than the corresponding period of many other hydroids. In 
several cases specially noted the planule did not settle for 
attachment and transformation till some five days following 
the beginning of development, and only at the end of seven 
days were tentacles well marked, as shown in Pl. III, Fig. 10. 
The tentacles originate by a process of budding, the lower, or 
filamentous, series appearing first, those of the other series 
following somewhat later. 
Secretion of the perisarc begins almost at once after the 
attachment of the larva, preliminary to transformation. At 
