Knowledge of the Spongida. 257 
developed, especially towards the surface, with no dendriform 
fibre-skeleton, but with an abundance of spicules, might also 
claim a place among the Carnosa—although I cannot speak 
with such certainty of the other species of Pachastrella that 
have come before me, which, as Schmidt has stated, are very . 
ill-supplied in this way (‘sehr arm an Weichtheilen,” No. 31, 
p- 69), while the habit of Dercitus niger of extending itself into 
the cracks and crevices, however minute, of the rock on which 
it may be growing, inclines me to the view that it would 
do this with other objects, such as shells and corals, under 
similar circumstances, if growing upon them: hence, on one 
occasion, 1 found its spicules, together with those of Cliona 
mucronata, Sollas, in the excavated multilocular cavities of 
a branch of stony coral from the island of Cuba. 
And this opens another question, viz. how many of the 
Keccelonida (No. 29, p. 496) or excavating sponges may 
belong to the Carnosa ; for, on the one hand, the tetractinellid 
spiculation of Samus is in form very evidently allied to that 
of Pachastrella, ex. gr. Dercitus niger &e., and on the other 
to that of Corticiwm plicatum, Sdt., C. abyss’, Carter, and 
C. versatile. 
So far as Cliona alata, when growing within its excavated 
multilocular cavities in shells or rocks (calcareous), and 
when free in the form of Rhaphyrus Griffithsii, Bk., and 
Cliona corallinoides (‘ Annals,’ 1871, vol. viii. p. 14, pl. 1. 
figs. 33-36) go, there is nothing but the absence of an “ evi- 
dent skeleton,” as in Halichondria suberea, that would induce 
me to place them among the Carnosa; and of the rest I can 
state nothing in this respect; but as regards the genera Samus 
(No. 30, p. 59), Alectona (olim Gummina) Wallichit and 
Millari (No. 29, p. 494), if not of Thoosa socialis, Dotona 
pulchella, and Alectona Higgint (No. 30, pp. 56-58), whose 
almost microscopic dimensions render this evidence respecting 
them presumptive only, the gum-like consistence of the sar- 
code, together with the presence of the elastic-tissue filaments 
and the absence of a fibre-skeleton, seems to claim for them 
all a place among the Gumminida; indeed Schmidt’s Corti- 
cium versatile appears to be my Samus simplex (No. 30, p. 60, 
pl. v. figs. 26 a-c). 
All these are “ border-questions,” as I have said before, in 
which the transition of one kind of structure into another, as 
exemplified in different species of sponges, becomes perplexing 
to’ the classifier, who, after all, can only divide them at the 
confines of his grouping by an empiric distinction, chiefly 
based upon ‘‘ degree,’ which arrangements, under the best of 
circumstances, must be conventional, as there is no line of 
Ann. & Mag. N. Hist. Ser. 5. Vol. viii. 18 
