364 HYALONEMA (PHIALONEMA) PATERIFERUM. 
of the specimens allows one to judge, this is also the case in the forms EF and F. 
It is certainly true in some forms, probably in all, that from the margins which 
mark the boundary between the dermal and the gastral faces there arise centro- 
tyle amphioxes, diactine pinules, and spicules transitional between these forms. 
Besides the lateral rays of the pinules, amphidises, paratangentially extending 
centrotyle amphioxes, and the lateral rays of pentactine megascleres are found 
in the dermal and gastral membranes. According to Wilson (loc. cit.) the amphi- 
dises of the superficial membranes are large macramphidiscs, and he says that 
these spicules are very abundant in these membranes of the forms (A and B) 
studied by him. In the specimens of forms D and EF, examined by me, where 
the superficial membranes are more or less preserved, I found them occupied 
by micramphidises in places very abundant, but nearly destitute of macram- 
phidises. Below the superficial membranes, the apical (proximal rays of the 
pentactines, centrotyle amphioxes, and a few transitions between them.and diac- 
tine pinules, occur. All these spicules (spicule-rays) are situated radially. 
Hexactine megascleres, canalar pinules, microhexactines, transitions between 
these and the pinules, and amphidises are met in the interior of the body rhabds, 
which are, for the most part, centrotyle amphioxes. In the vicinity of the point 
of origin of the stalk stout-rayed acanthophores occur. Many of the rhabds 
of the interior form bundles which traverse the choanosome. The hexactine 
megascleres appear to increase in size toward the central part of the sponge. 
In the forms C, D, HE, and F the canalar pinules are scarce, and irregularly 
and sparsely scattered over the walls of some of the canals only, the walls of 
other canals appearing to be destitute of these spicules. The microhexactines 
vary considerably in respect to their size, their spinulation, and the curvature 
of their rays. Wilson (loc. cit.) considers the large, straight-rayed, and strongly 
spined ones (in the forms A and B) as canalaria. In the forms D, EH, and F 
these spicules do not appear to be restricted to the canal-walls. In form C I 
failed to find any of the large, straighter-rayed microhexactines. The hexactine 
and pentactine transitions between the microhexactines and the pinules are, 
in the forms D, E, and F true canalaria. The amphidises, among which four 
forms can be distinguished, are exceedingly abundant. Micramphidiscs, chiefly 
large ones, clothe the walls of the efferent canals of form D in dense masses. 
In the forms C, D, EL, and F macramphidiscs are scattered in very large numbers 
through the choanosome. According to Wilson (loc. cit.) only few macramphi- 
discs occur in the interior of forms A and B. The large macramphidises are 
much more numerous than the small ones. The stout-rayed acanthophores 
