282 REVISION OF THE MONAXONID SPONGES, i., 



1'5 and 1*9 mm., (but only in rare specimens exceeding l*6/x) and 

 a maximum stoutness approximating to 30 /x. The spicules of 

 the second kind, which are typically of much smaller size than 

 the preceding, though connected with them by a perfect grada- 

 tion, contribute to form a "nuclear" skeleton surrounding the 

 centre from which the fibres radiate, and are found also in the 

 cortex in the penicillately outspread terminations of the fibres; 

 the smallest of them measure less than 275 by 10 /x. Those of 

 the third kind occur between the fibres, chiefly in the more peri- 

 pheral part of the choanosome, and they vary markedly in size 

 and abundance in different specimens. All three kinds are alike 

 styli, which gradually taper towards the basal end and usually 

 exhibit a faint constriction just immediately above that end: but 

 the first-mentioned, or chief fibral, spicules are fusiform, and 

 almost invariably have the apical end more or less rounded off so 

 as occasionally to approximate in form to strongyla; the second 

 are nearly cylindrical in shape, and are more or less abruptly 

 sharp-pointed; while those of the third kind taper gradually to 

 a usually very fine point. As already stated, the last-mentioned 

 spicules are subject to considerable variation in size and number. 

 Thus in one specimen (which is to be regarded as strictly typical 

 of the var. Icevis) these spicules are very few in number and rarely 

 exceed 600 by 6 jx in size; whereas in most of the specimens 

 labelled by Lendenfeld as the types of Tethya inflata and Tethya 

 corticata, they are, on the other hand, extremely abundant and 

 about equal in size to the spicules composing the fibres. Other 

 specimens which I have examined are less widely divergent in 

 these respects, and at present (although further investigation is 

 necessary in order to settle the point) I do not think that the 

 differences in question are varietal, more especially as they do 

 not appear to be associated with any constant differences in 

 respect of other characters. 



The spherasters are almost entirely confined to the more super- 

 ficial part of the cortex, and to the outermost region of the 

 choanosome adjoining the cortex; the largest have a maximum 

 diameter varying in different specimens from 60 to 85 /x ; the 

 rays, which in length are about equal to the diameter of the 



