400 PROCEEDINGS OF THE CANADIAN INSTITUTE. 
take a brownish-black tinge in osmic acid, by which also the nucleus 
is rendered indistinct. 
The cells in the base of the gland are nearly oval, not provided 
with a process, coarsely granuled, and the large nucleus situated in 
the centre of the cell. The granules are to be found equally in all 
parts of the cell, which, on the whole, takes a slightly darker 
stain than those of the body of the gland, which are never found to 
bulge outside the general limit. The former when not in a resting 
state give an irregular appearance to the base. This was best seen 
in young specimens of cat-fish which are always feeding. Macerat- 
ing the mucous membrane of such specimens in Ranyvier’s alcohol, 
Miiller’s fluid or in’a mixture of the latter and serum, appearances, 
such as Fig. 5 gives, were obtained. There the cells of the body of 
the gland are rhomboidal in outline and form a pretty regular inner 
border. Those at the base, however, cause a bulging out of the 
membrane, some being situated in wedge-shaped niches between the 
other cells. During activity they preserve this form, shrinking to a 
certain extent when resting. 
Between the cells of the body of the gland and those of the base, 
staining reagents show not the slightest difference, carmine, hema- 
toxylon, aniline blue, stain all alike in intensity. The slight differ- 
ence obtainable in osmic acid hardly merits mention. The granules 
in all are equally coarse, and four or five hours after the introduction 
of food into the stomach are arranged about the lumen, which in 
these glands is more or less indistinct. The cells are unprovided 
with a membrane, and in serum are all spherical, the processes being 
retracted. They, however, preserve their original forms in Miiller’s. 
fluid and Ranvier’s alcohol. ; 
F. E. Schulze! describes in Silurus glanis large spherical cells. 
lying in niche-like swellings of the basement membrane, and he evi- 
dently intended a comparison of these with similarly situated cells 
in higher vertebrates. As Amurus and Silurus belong to the same 
family, it is quite probable that these structures are alike in both 
and that they have no more morphological value than what I have 
attributed to them. 
Edinger? discovered in Perca fluviatilis differences in these cells. 
which, however, he does not describe. Still he believes that a dis- 
1 Joc. cit. 2 Loc. cit. 
