Ixviii 
THE VOYAGE OF H.M.S. CHALLENGER. 
as of independent origin. Nothing in the characters of the spicules of Placina seems at 
first sight to suggest for them a simpler origin, yet a general consideration of the subject 
leads me to suspect that they also may be traced back to a sigmaspire, but not through 
a genuine spiraster. Bowerbank, represents as the spicules of an undescribed sponge 
several forms of sigma with one or more rudimentary actines proceeding from the 
centre ; these spicules, which Bowerbank ^ terms exter-umbonate, inter-umbonate, and 
bi-umbonate bihamates, we may incorporate into our system of nomenclature as 
centractinate sigmas. The particular form, which is of interest as bearing on the present 
enquiry, is that in which the actine is directed outwardly from the convex side of the 
sigma, and to this alone we shall refer in using the term centractinate sigma in the 
following remarks. The first point of interest is that such spicules are by no means 
confined to the Sponges, but occur also in the Echinodermata, and in the Nudibranch 
Molluscs, and that not at all uncommonly. Next in both these groups, the centractinate 
sigma passes into a centractinate form in which the actine is of considerable size in 
relation to the rest of the sigma; so large is it that the spicule presents a close 
resemblance to a triod, into which form indeed it passes by the straightening of the two 
arms of the sigma, one on each side of the actine ; finally, by the appearance of an 
additional actine the triod passes into a microcalthrops. There is thus suggested for 
our consideration the possibility of a similar origin in the case of the microcalthrops 
of Placina. In this sponge microtriods and microxeas are almost as plentiful as micro- 
calthrops, and they are always distinguished either by a curved inflexion at the’ centre 
(PI. XLIII. figs. 14, 14a), or by two of the actines forming together a regular arciform 
curve. Thus, what little evidence we can adduce is in favour of at least a double 
origin for the aster, and we may suppose that in some cases it has been developed from 
the sigmaspire through the spiraster, and in the others from the same spicule through 
the sigma and its centractinate modification. 
In both . Placina and most species of Thenea the mesoderm is not largely 
developed, and it is possible that this is connected with the comparatively few actines 
which the asters in these sponges possess ; with a larger development of mesoderm or 
perhaps from other causes the aster acquires additional actines ; with increase of size, 
and sometimes without, the aster becomes reduced and furnishes the microrabd or even 
a microstyle, and indeed in some sponges a globule, so that this last-named spicule may 
arise in two ways either as a primitive form or as a reduced aster. The megarabd or 
rhabdus is in all probability an overgrown microrabd. The style may have been 
derived from a microstyle, but transitions from the rhabdus to the style are of such 
frequent occurrence that it may with equal and perhaps greater probability have 
descended from the rhabdus. The same is true of the tylostyle, and the shortening of 
the rhabdus may proceed so far as to produce the sphere. These modifications of the 
1 Bowerbank, Mon. Brit. Spong., vol. i. figs. 115-11'7. 
