cxliv 
THE VOYAGE OE H.M.S. CHALLENGER. 
Gray’s modification of this classification in 1872 is by no means an improvement, but 
notwithstanding numerous alterations the distinctness of the Placospongiadse from the 
Geodiadse is maintained.^ 
Carter in 1875 {loc. cit.) regarded the Geodiidse as a subfamily group of equal value to 
the Stellettidae and Tethyina (our Tetillidae), these three groups together constituting the 
family Pachytragidse, Carter. The genus Placospongia was removed from all association 
with the Geodiidse and placed in the subfamily Donatiua, Carter, belonging to the 
Suberitidse. In 1880^ I adopted Carter’s group Geodina, but included in it Placo- 
spongia, a Sponge which I probably had not then seen, and was clearly not well 
acquainted with. In the same year Carter® justified the association of Placospongia with 
the Suberitidae on the ground that it possesses tylostyles and spirasters ; subsequently 
Carter* proposed a new group, Placospongina = Placospongiadae, Gray, adding that “in 
spiculation” it unites the two groups “Suberites” and Geodina. Vosmaer does not 
include Placospongia in the Geodiidse, nor indeed does it find a place in his system. 
Definition . — Sterrastrosa possessing triaene megascleres. 
The characters of the cortex, which is the most distinctive feature of the Sterr- 
astrosa, have been most closely studied in the Geodiidae. The sterrasters are united 
together by fusiform fibrillated cells, probably inocytes, which are attached to the 
projecting ends of the actines, and these are frequently furnished with recurved spines 
to afford a surface of attachment : the connecting cells extend directly in a straight line 
from the surface of one sterraster to the opposed faces of its surrounding neighbours, 
and thus a strong, tough, composite sclerose and fibrous layer results, which we shall 
term the “ sterrastral layer.” It is a unique feature in the Sponges, nothing quite 
similar to it occurring in any of the other groups ; in some cases, as in Disyringa for 
example, oxeas tangentially arranged are bound together by fibrous tissue in the cortex, 
but the union is produced by the fibres wrapping round the united spicules ; a somewhat 
similar spicular layer also occurs in Dragmastra, which is characterised by a dragmastral 
layer of the cortex, but the dragmas in this case lie in clusters in a collenchymatous 
layer, and are not in organic connection with it ; in another Stellettid genus, Aurora, a 
sclerose layer, in this case spherastral, occurs, but so far as one can conclude from an 
examination of dried specimens, there is just as little organic union of the scleres here 
as in Dragmastra ; in Craniella and CinacJiyra a composite layer of oxeas and fibrous 
tissue occurs, but without presenting any. such union as occurs in the Geodiidse ; finally, 
in the Monaxonid Tethya the cortex is characterised by spherasters embedded in fibrous 
tissue, but though I have sought carefully for some signs of a direct connection between 
the scleres and the fibres, I have never been able to find any. 
^ Gray, Ann. and Mag. Nat. Hist., ser. 4, vol. ix. p. 460, 1872. 
2 Sollas, Ann. and Mag. Nat. Hist, ser. 5, vol. v. p. 241, 1880. 
® Carter, Ann. and Mag. Nat. Hist, ser. 5, vol. vi. p. 55, 1880. 
^ Carter, Ann. and Mag. Nat. Hist., ser. 5, vol. ix. p. 357, 1882. 
