14 PERCY SLADEN TRUST EXPEDITION 
supporting skeleton of the subdermal cavities. The latter is formed partly by small 
triradiates lying parallel to but beneath the surface, partly by inturned rays of some 
of the dermal triradiates, and partly by small triradiates which lie in planes more or 
less at right angles to the surface, with one ray pointing outwards and lying parallel 
to the inwardly-directed rays of the dermal triradiates in the trabecule between the 
subdermal cavities. 
The figure given by Poléjaeff of the skeleton of the subdermal cavities in 
Pericharax carteri, var. heteroraphis, appears to be highly diagrammatic. He does not 
seem to have recognized the nature of the inwardly turned rays of the dermal triradiates, 
and figures the deep triradiates of the ectosome as markedly sagittal, and the whole 
as far too regular in arrangement. I believe, however, that the deeper spicules of the 
ectosome are sometimes sagittal, like many of those in the dermal membrane. 
There is a thin gastral cortical skeleton, composed of quadriradiate spicules, which 
lines not only the central gastral cavity but also the larger exhalant canals, the surface 
of both of which bristles with the projecting apical rays. The spicules may be classified 
as follows :— 
(1) Large triradiates of the main skeleton (Plate 5, fig. 1, a); equiangular and 
typically equiradiate; rays long, straight and gradually sharp-pointed, measuring, say, 
about 1°55 mm. long by 0°1 mm. in diameter at the base. 
(2) Small triradiates of the main skeleton (Plate 5, fig. 1, b; fig. 2, a); equiangular 
and typically equiradiate ; rays straight, slender and gradually sharp-pointed, measuring, 
say, about 0°18 by 0°015 mm. Intermediate sizes (Plate 5, fig. 1, c) between (1) and 
(2) occur in comparatively small numbers. 
(3) Triradiates of the dermal skeleton (Plate 5, fig. 2, b); irregular, but probably 
really equiangular; with slender, irregularly curved rays, often’ unequal in length and 
often blunt at the apex. One of the rays (z) is frequently bent inwards out of the tangential 
plane, so as to take part in the formation of the skeleton of the subdermal cavities. Rays 
measuring, say, about 0°13 by 0°008 mm. 
(4) Quadriradiates of the gastral cortex (Plate 5, fig. 2, c); equiangular and 
typically equiradiate, with long, straight, or nearly straight, slender facial rays, fairly 
sharply pointed and measuring about 0°18 by 0°0125 mm.; with a sharp-pointed apical 
ray, irregularly hooked* at the extremity in a characteristic fashion and about 
0-1 mm. long. 
(5) Quadriradiates of the larger exhalant canals (Plate 5, fig. 2, d); similar to 
those of the gastral cortex, but with long, straight apical rays, measured up to 0°26 mm. in 
length ; as usual, intermediate forms occur. 
(6) Quadriradiates of the oscular collar (Plate 5, fig. 2, e); resembling those of the 
gastral cortex further in, but conspicuously sagittal owing to the sharp backward bending 
of the oral rays until they come to lie parallel to the oscular margin. These spicules are 
* This hooking does not occur in the “Challenger” specimen, but I cannot regard it as constituting 
a character of specific value. 
