420 
THE YOYAGE OF H.M.S. CHALLENGER. 
chamber-system of Epallax is scarcely further advanced than that of Astropeplus, and 
it is of great interest to find that the same is true of Phacellia as described by Mr. 
Dendy (Eidley and Dendy, loc. cit., pi. xlix. fig. 3). The Axinellidse are also nearly 
related to the Tetractinellida, and the occurrence of microcalthrops described by Carter 
in Microciona quadriradiata ^ and in Dictyocylindrus vickersii ^ becomes intelligible 
as a case of survival. The spheraster of Dictyocylindrus stuposus, Bowerbank, and 
the aster with dichocladal actines of Dictyocylindrus fasicularis, Bowerbank, are readily 
derived from the simple asters of Epcdlax, so too are the small styles which occur in 
conjunction with the larger rhabdi of so many Axinellidse, though the passage of these 
can perhaps be more exactly traced through the microcalthrops of Tricentrium ; I have 
described this spicule as presenting various forms of di-, tri-, and tetractinose forms in 
Tricentrium muricatum, Ehlers,^ and Carter has described a whole series of species in 
which the selection of one or other of these forms has taken place ; thus in Tricentrium 
[Dictyocylindrus) vickersii (Carter), we have asters with from three to five actines 
present and serving as “ echinating ” spicules, in Tricentrium [Microciona) quin- 
queradiata (Carter), pentactinose asters are present, in Tricentrium [Microciona) 
quadriradiata (Carter), tetractinose forms (microcalthrops), in Tricentrium muricatum, 
Ehlers, chiefly triactinose forms, in Tricentrium [Microciona) curvispiculifera (Carter), 
diactinose forms, and in Microciona (?) hulboretorta, Carter, we reach the style ; 
the smaller “ echinating ” styles of the Axinellidse are traceable ultimately to calthrops 
or asters, and so far afibrd a justification to Vosmaer’s view, by which Tricentrium is 
assigned to the Tetractinellida, but when the whole assemblage of characters is taken 
into account this view becomes untenable, for the genus Tricentrium is evidently an 
Axinellid and cannot be removed to the Tetractinellida without taking the Axinellidae 
along with it. 
The Dorypleridse are readily understood as resulting, so far as the spicules are 
concerned, from an overgrowth of the oxeas and asters of Astropeplus. 
The Tethyidse must be traced backwards towards a Placinid ancestor in order to 
explain the radiate arrangement of the skeleton, which evidently depends on their mode 
of growth. If in the case of Epallax we regard the vasiform sponge as resulting from 
the widening of the oscule of the original Ehagon, and a subsequent outward and up- 
ward growth of its margin, we shall perceive how the arrangement of spicules which it 
presents may be brought about, for we have two chief directions of growth, — one from 
the base along the sides of the cup or radial, this will lead to the over-development of 
such spicules as happen to be directed along radial lines ; the other at right angles to 
this in the direction of the axes of the chief folds or evaginations of the gastral wall. 
^ Ann. and Mag. Nat. Hist., ser. 5, vol. vi. p. 42, pi. iv. fig. 4, 1880. 
2 Op. cit., vol. iii. p. 285, pi. xxvi. figs. 1-4, 1879. 
^ Ann. and Mag. Nat. Hist, ser. 5, vol. iii. p. 19, pi. vi., 1879. 
