208 Dr.J. E.Gray on <//e 



t. V. f. 117). Some of these unilateral spicules, instead of 

 having a hook, have a broad rounded disk at each end. 

 Bowerbank shows, by a series of figures of a spieule of this 

 kind, whieh he calls " bipocillated bihamate," from Ilalichon- 

 (Iria Hynchnani (B. S. t. v. f. 123-127), tiie various forms 

 which the spicule assumes before it is fully developed, and 

 the various states in wiiich it is to be seen in the same sponge ; 

 and the biclavate and bihamate spicules (B. S. t. v. f. 118-120) 

 appear to belong to the same series. 



The spicules which Bowerbank calls inequitrirotulate and 

 eccentric trirotulate (B. S. t. x. f. 131-134) are evidently allied 

 to those last described and t. v. f. 118-120. 



That these spicules are variations of the same spicule I 

 think is proved by their occurring mixed together in the same 

 sponge, and generally in conjunction with the bihamate spicules 

 (see Schmidt, Spongienf. t. v. f. 2, 3, 5, 6, 7, all different spe- 

 cies of Esperiadse) ; indeed they may be undeveloped or nidi- 

 mentary hamate spicules. 



The " tricurvato-acerate " spicule (figured by Dr. Bower- 

 bank, B. S. t. iv. f. 96-98, where the spicule is slightly sinuous 

 and the end only recurved in the most perfectly developed spi- 

 cule, f. 9(3), appears to belong to this type, or to be interme- 

 diate between it and the simple needle-like spicule. 



3. The quinqueracliate or "trifurcate " spicules are distin- 

 guished by always having only three equidistant lateral rays 

 from the sides of the more or less elongated axis. 



They do not very generally occur in a perfect state of de- 

 velopment, with an elongate axis having three equidistant rays 

 coming from the middle of its length ; but they are sometimes 

 found in this state mixed with the more usual form, Avhere the 

 upper part of the axis is very short or absent, and the rays 

 are very generally recurved (see B. S. t. ii. f. 58, where 

 the spines are directed towards the shorter end ; t. ii. f. 55, 

 56, where the spines are directed from the shorter end of the 

 axis, which is bent outAvards). 



Bowerbank calls the first spicule (fig. 58) " spiculated por- 

 rceto-ternate," and f. bb^ 56, 57 " spiculated recurvo-ternate," 

 keeping the tenu " recurvo-ternate " for the spicules in which 

 one part of the axis is entirely absent (f. 54). 



Most generally the axis does not reach beyond the rays, as 

 in the last referred to ; and the lateral rays are generally short 

 and recurved, as in B. S. t. ii. f. 48, 54, t. iv. f. 81 & 82 ; but 

 the rays are sometimes directed forwards (t. iv. f. 77-80, t. v. 

 f. 128, 129), and they are often expanded horizontally (t. ii. 

 f. 45, 47, t. iv. f. 95). Of course, all tliese spicules have 

 different names, though you may find them comlnned in the 

 same s])ecimen. 



