82 
TEE VOYAGE OF H.M.S. CHALLENGEE. 
batory poucli the cells are fusiform, and are arranged with their long axes placed along 
the length of the neck. They are very close together, and form a nearly compact tissue 
of longitudinally placed rows of cells. These fusiform cells have their ends in many 
cases greatly drawn out so as to resemble fibres. Cells of the normal spindle-shape 
without the terminal fibres occur also on the lobes at the margins of the apertures. 
Here also, as on the neck of the pouch, they form a rather compact layer. 
The muscle fibres are non-striped, very long, almost colourless, stain yellow with 
picrocarmine, and exhibit no traces of a nuclei. The different parts of the mantle vary 
greatly in the amount and disposition of the musculature. In the thin part covering the 
abdomen, and adhering closely to the subjacent organs, muscular fibres are usually not 
visible at all, though occasionally in the upper part, next the thorax, a few delicate 
stragglers may be seen. The thorax, on the other hand, has fibres running in almost 
every direction. They are arranged in bands, generally two or three fibres thick, which 
branch and unite again (PL YI. fig. 5) ; single fibres being alone for short distances only, 
as when they stretch from one band to an adjacent one. A number of bands are more or 
less transversely placed, encircling the middle and lower part of the thorax as a series of 
equidistant parallels (PI. YI. fig. 4) ; often, however, they are more irregular. This is 
the rule on the anterior part of the thorax where the bundles cross at all angles and seem 
at first to have no definite arrangement. On account of the anterior curvature the 
circular bands, as they pass over the dorsal and ventral margins, are directed posteriorly, 
this gives them an oblique course and causes a decussation of those from the two 
margins (PL YI. fig. 4). A few longitudinal fibres are also present running down the 
sides of the thorax from the sphincter muscles of the apertures. These cross the densest 
part of the decussation of the circular fibres, which is in the middle of each side just 
posterior to the branchial aperture, and thus form a patch from which fibres seem to radiate 
in all directions. On each side, near the ventral margin, a long and rather close bundle of 
fibres may be seen running in a curved direction from the branchial siphon to the 
posterior end of the sac where the fibres diverge, many of them sweeping round to the 
dorsal edge. 
The ventral margin over the endostyle (PL YI. fig. 4) seems to be more muscular 
than the corresponding region of the dorsal edge, the bands also cross, branch, and 
anastomose more frequently ; possibly this may be accounted for by the more anterior 
position of the ventral margin, the atrial aperture being slightly posterior to the 
branchial. Circular muscle bands act as sphincters at the two apertures (PL YI. 
figs. 2, 3, 4). They are placed around each siphon and the fibres are much coarser and 
closer than on any other part of the thorax. The branchial sphincter is stronger than 
tlie atrial. 
Over the incubatory pouch the musculature, though stronger than over the 
abdomen, is feebly developed. In the young pouch not a fibre is visible, but in those which 
