11 6W. 
CCELBNTBKATA. 
in the endodermal cells of Hydra and Clava ; the third part contains histo- 
logical observations on Clava squamata. 
The knowledge of the general intimate structure of Mednsw^ the cras- 
pedote especially, has been largely increased by the work of the Hert- 
AVIGS (14). All parts and tissues belong — with the partial exceptions 
referred to below — either to the ectoderm or the endoderm. The ectoderm 
is sometimes differentiated into the epithelium proper and the subepithelial 
(interstitial) tissue ; in such cases the epithelium is often connected with 
the basement membrane through delicate fibrillar elements ; the urticat- 
ing cells, the nervous system, and the sense organs (chiefly), the muscles 
and the generative organs all belong to the ectoderm ; the authors, at 
least, have not found muscular fibres inside the basement membrane, and 
another interpretation is given to a part of those described in this place 
by others. It is also evident that in most, at least, of the instances 
adduced by the authors, the organs of reproduction are truly portions of 
the ectoderm ; and it must be confessed that in cases where the evidence 
is less convincing, they are very naturally interpreted as properly belong- 
ing to the ectoderm. The circular muscles of the sub-umbella and velum, 
moving rhythmically, are transversely striated ; the radial and stomachal 
muscles are smooth ; those of the tentacula may belong to either kind ; 
but there are no dorsal muscles. The endoderm not only forms the 
epithelial investment of the gastro- vascular cavities, but also a delicate 
lamella, uniting the radial and circular canals and separating the jelly of 
the umbella from the hyaline membrane ; important morphological 
deductions are based upon the universal existence of this layer, the 
primitively double endodermal investment of the gastral cavity of the 
Medusa^ of which cavity the stomach and radial and circular canals 
of the developed Medusa are the sole remnants. To the endoderm 
also belongs i^he internal epithelium of the tentacular cavity, trans- 
formed into a single series of axial cells in Medusae with solid, not 
hollow tentacles, modified for feeling, useless for prehension. The 
jelly of the disk is regarded chiefly as a layer secreted by the endoderm. 
When containing cells (immigrated from the endoderm ?) and there- 
fore endowed with independent growth, as in Aurelia, it may assume 
the character of a mesoderm, and the same may be said of the axial 
tissue in the tentacles of those Medusae in which the original com- 
munication between the tentacle and the gastro-vascular system is 
finally interrupted ; and of the sub-umbellular sub-epithelial muscular 
layer in its highest development, when separated from the ecto- 
dermal epithelium through a secondary basement-membrane (in 
jEquorea). The general histological and morphological results attained 
by these observers agree well with those of Claus (7), but dis- 
accord with those of Bohm (4), who for instance regards the Medusae 
generally as “ Trihlasteria^' (while the Hydroids are Diblaster ia^^), refers 
the ovaria to the endoderm, the “ endodermal lamella ” of Hertwig (“ vas- 
cular plate ” of Claus) to the ectoderm, and denies the existence of a 
basement membrane in the tentacles, in the walls of the stomach, and in 
the sub-umbella. These statements are controverted by the Hertwigs, 
and explained by the imperfect method employed, the sectional method 
