126 | THE VOYAGE OF H.M.S. pee 
undermost glands is figured. These sara project into the interior of the intestine, and. 
are, as far as I know, the only true glandular bodies which stand in connection with the 
alimentary canal. They are invested by the same membrana intima as the wall of that 
part of the cesophagus, at the end of which they are found. The form of the cells which 
compose them is nearly the same. The whole of the gland must be considered as having 
taken its origin from an excrescence and bending towards the wall of the intestine, of the 
- hinder part of that cesophagus. . 
In regard to the structure of the remaining part of the intestine I do not wish to 
- enter into any details. I only observed that the structure of the wall of that part which 
follows immediately after the cesophagus, and of the caeca, which penetrate as a rule 
as far as the sixth joint of the leg, is nearly the same. We find this wall everywhere beset 
with extremely numerous tubes or villi, which in some genera (Nymphon) are of a 
shorter and more rounded form, and in others (Colossendevs) are very slender and almost 
cylindrical. While the outermost part of the wall is formed of a single row of large 
distinct nucleated cells, these villi show a multi-cellular structure also. Each of these 
cells contains numerous globules, which for the most part seem to be of a fatty 
nature. The form of the cells is different, but they are commonly rounded. I call 
them cells, because each of the bodies has a distinct oval nucleus with a small nucleolus. 
As has been observed by almost all authors writing on the structure of the wall of the 
intestine, these cells often become detached from the wall, and are found lying loose in 
the contents of the alimentary canal. The nuclei in the cells of these villi were by no 
means easily observed in all the sections I studied. They were very distinct in the cells 
of the villi of Colossendeis proboscidea, Sab. (sp.). 
It seems to be characteristic of the genus Colossendeis that the ceca destined 
for the different legs should branch off from the main duct, which runs straight from the 
proboscis to the abdomen, very close to one another, and close also to the place where the 
cesophagus communicates with the intestine; at least I observed that in the three species 
of Colossendeis I studied (Pl. XVII. fig. 1). The number of lateral cxca given 
off at both sides of the main duct is six-in Colossendeis. Of these the first two are 
rudimentary, one being the rudiment of those destined for the mandibles, the other (the 
first lateral pair) being that for the proboscis. Each of the four remaining ceca, of. 
which the hindermost pass through a much longer part of the body before penetrating the 
leg than do the more anterior ones, shows a considerable swelling in the lateral processes, 
at the ends of which the legs are inserted.?. The anal aperture of Colossendezs (Pl. XVII. 
.fig. 1d) is an oval-shaped slit. It is not placed terminally or in the median line of the 
abdomen, but laterally. 
1 On a transverse section of the body of a Colossendeis between the lateral processes for the second and for the 
third pair of legs, five round sections of the intestine are seen placed close to one another. This furnishes a good 
opportunity of comparing their structure, but no difference is observed. Compare fig. 14 of Plate XXI. 
