Species of Echinonematous Sponge. 49 



the intermediate acuates becoming thinner and the dermal 

 acuates thicker, and so approaching the one to the other. Hence 

 I see no difficulty in believing that the spicule-cell of the larger 

 acuate is a direct descendant of that of the dermal spicule. 

 Again, the dermal spicule sometimes becomes rounded off at 

 both ends, and in one instance was observed with an inflation 

 at each end, each inflation being sparsely microspined 

 (PL VII. fig. 20); on the other hand, the dumbbell-like 

 axial spicule loses one of its inflations, retaining the other, so 

 that at one end it possesses a spherical head, and at the other 

 is merely rounded off (PI. VII. fig. 26) , or terminated some- 

 times by a blunt mucrone (PL VII. fig. 25), which, however, 

 still retains its microtuberculation, though the tubercles ex- 

 hibit a tendency to acquire the characters of small spines ; 

 again, the axial spicule varies much in proportional length, 

 being sometimes very short and stumpy, like PL VII. fig. 21, 

 and at others elongating and narrowing till one has the attenu- 

 ated form, PL VII. fig. 22, very similar in proportion to one 

 of the dermal spicules. Thus the passage of a dermal spicule 

 into one of the dumbbell-like forms is by no means incon- 

 ceivable. As regards the spined echinating spicules, these 

 exhibit as many variations as the others ; the extent to which 

 the pointed end is spined is very inconstant, some forms re- 

 maining almost smooth (PL VII. fig. 13, for instance), in 

 which case they closely resemble the young spinose-headed 

 forms (PL VII. fig. 12) of the intermediate acuates, from 

 which they might easily have been derived. Our theory of 

 the relationship between these various spicules may be ex- 

 pressed in the following diagram : — ■ 



Adult smooth acuate 

 (tiff. 3). 



\ 



Young form of -=*— Dermal spicule —**- Axial spicule 

 acuate (fig. 28). (fig. 4). 



(fig. 12). 



/ 



Adult spiuose acuate 

 (fig. 5). 



Of other variations displayed by the spicules one may 

 notice the tendency, very frequently displayed by the large 

 acuates, to exchange the pointed for a rounded extremity, 

 thus passing into the cylindrical spicule, PL VII. fig. 9; and 

 this may be combined with a general stunting in linear 

 growth, which leads to such forms as figs. 10, 11, PL VII., 

 and might, if carried far enough, give us a truly globular 



Ann. & Mag. N. Hist. Ser. 5. Vol. iv. 4 



