398 TRANSACTIONS OF THE CANADIAN INSTITUTE. [VoL. VI. 
It is clear then that in Z. soc¢atus there is neither a histological con- 
tinuity nor a histological identity of the upper part of the median streak 
of the filament with the lower part or glandular streak proper. The up- 
per part is merely a continuation downwards of the stomatodzal ectoderm, 
and this gives place to a low epithelium destitute of gland cells and of a 
generally indifferent character, below which the characteristic epithelium 
of the glandular streak comes into view. It seems to me from these re- 
sults that one is not justified in assuming, as has so frequently been 
done, that the glandular streak epithelium is a prolongation of the 
stomatodeal ectoderm. What I have just described, taken in conjunc- 
tion with the observations of E. B. Wilson (1884) on the development of 
the filaments in the Alcyonaria and with what | have found (1891) as to 
their development in the Hexactiniw, seems to me rather to point toa 
complete distinction between the two kinds of epithelium, and I regard 
the structure of the adult filament of Z. soctatus as confirmatory of the 
conclusions obtained from embryological studies, that the ciliated bands 
of the filaments are ectodermal in origin, while the glandular streak 
proper is an endodermal structure. 
II]l.—THE DEVELOPMENT OF THE FILAMENTS IN EGG EMBRYOS. 
The material at my disposal for the study of the embryonic develop- 
ment of the filaments was not sufficient for an exhaustive study of the 
subject. The youngest larva already possessed twelve mesenteries ar- 
ranged in the manner described by van Beneden (1890) and myself 
(1891a.). On none of the mesenteries were there any indications of the 
ciliated bands, but, on the other hand 
the glandular streaks were plainly 
indicated on the perfect mesenteries 
as an epithelium occupying the free 
edge of the mesentery and composed 
of cells with closely set, elongated 
and deeply staining nuclei, very 
different from those of the general 
endoderm of the mesenteries. But 
what is more interesting, on the 
lower part of the free edge of each of 
the imperfect mesenteries a similar, 
but smaller, patch of epithelium was 
plainly visible. In Fig. fo is given 
i Ss _ a representation of a part of a-sec- 
Fic. 10.—Transverse section through a portion. 
phithe colmn of al yout zoauimdilacve: tion through the lower portion of 
