R. W. H. ROW—REPORT ON THE SPONGES : CALCAREA. 187 
Genus GRANTILLA, n. g. 
Grantillide with syconoid canal-system, consisting of radially disposed 
flagellated chambers, supported by an inarticulate tubar skeleton. 
The two specimens in the collection, each representing a distinct species, 
for which this new genus and family have been founded, possess certain 
distinctive features in their spiculation not hitherto described for any 
calcareous sponge. The most noteworthy of these is the occurrence of three 
distinct kinds of subdermal spicules, namely, quadriradiate, sagittal tri- 
radiate, and a modified form of triradiate, which from its similarity to the 
quadriradiates described by Jenkin (9) as chiactines, I have called a 
“* prochiact.”” 
It seems advisable, both for the complete appreciation of the meaning of 
this spicule, and also in order to avoid confusion as to the nomenclature 
employed to describe it, first of all to define the terms used in referring to 
triradiates generally, and then to state the views set forth in this paper on 
the evolution of this spicule. 
The most typical triradiate spicule is one in which all the rays and angles 
are equal and which lies tangentially in the body-wall of the sponge, all the rays 
being in the same plane. As a general rule, however, some of the rays are 
somewhat modified so that it is possible to distinguish two of the rays from 
the third either by form, or by the inequality of the angles separating them, 
or by both at once. To describe these “sagittal’’ spicules the following 
terms are used. The two rays still equal are called the paired or oral rays, 
and the remaining ray is called the basal or aboral ray. ‘The angle between 
the oral rays is called the oral angle, whether larger or smaller than the 
others, which are called the paired angles. The primitive plane in which all 
three rays lie may be called the facial plane. 
In the primitive Olynthus, the basal ray lies pointing away from the 
osculum and the paired rays toward it. In the radial chambers of the 
syconoid canal-system the paired rays point toward the gastral cavity, and 
the basal away from it. When the chamber is strongly curved the paired 
rays tend to encircle it. 
In describing a quadriradiate the three rays in one plane, regarded as the 
equivalents of the rays of a triradiate, are called the facial rays, and the 
fourth, or extra, ray the apical, or gastral ray. Moreover, all the modi- 
fications described above as possible in a triradiate are equally possible to 
the facial rays of a quadriradiate, and the same terms are equally applied 
to them. 
It is considered that the prochiact (fig. 1,b) has been derived from a 
tangentially lying cortical triradiate (fig. 1, a), by a change in position of the 
basal ray, which, instead of lying tangentially in the cortex, has been turned 
