654 REVISIOK of TttE AX1NELLID.E, iii., 



metrical, and rather variable in degree; when most considerable, 

 it slightly exceeds that of two-fifths of the circumference of a 

 circle. 



Genus Holoxea Topsent. 



Definition. — Axinellid{e(?) of massive or encrusting habit; with 

 a more or less irregular, halichondroid, main skeleton, and with 

 or without a dermal skeleton of horizontally-disposed megascleres. 

 The megascleres are oxea of one or two kinds, and the charac- 

 teristic microscleres are minute microxea, somewhat resembling 

 sanidasters; in addition, trichodragmata ai-e typically present. 



Type-species, H. furtiva Topsent(45). 



The considerable agreement which exists between this genus 

 and Desmoxya in the matter of spiculation appears to me to 

 justify the supposition of a relationship between them. It is 

 true that, in H. furtiva, the type-species, so Topsent informs us in 

 his second description thereof(51), — the ectosome is charged with 

 more or less horizontally-disposed oxeote megascleres (differing 

 from those of the main skeleton only in size), and is differentiated 

 to form a cortex "peu epaisse, mais assez resistante"; but the 

 importance to be attached to this feature is minimised by the 

 fact that the other two species which have been ascribed to the 

 genus, — ^iz., //. collectrix and H. valida Thiele(39), — a specially 

 characterised cortex is, apparently, wanting; moreover, it is to 

 be noted that, in Desmoxya lunata, the dermal layer is particu- 

 larly well-developed, and if provided with a megascleric skeleton 

 would probably constitute what might be termed a cortex. Tlie 

 information at our disposal regarding the structure of the main 

 skeleton in the several species of Holoxea is extremely scanty: 

 Topsent merely mentions, in the case of J/, furtiva, that when 

 the sponge is confined to narrow crevices the megascleres are 

 constrained, owing to their great length, "s'orienter dans un 

 sens determinee, par faisceaux sur des longueurs variables", and 

 that "il en resulte souvent un faux-semblant de charj^ente 

 fibreuse"; while Thiele goes no farther than to state that the 

 megascleres (of //. collectrix) "lassen keine bestimmte Anord- 



