RELATIONS OF SAGITTA BIPUNCTATA. 17 



memoir, but is admitted bj him in his subsequent observa- 

 tions. When the animal is placed in spirits of wine, the 

 surface presents numerous distinct, whitish, well-defined spots, 

 which closer examination shows to be rounded eminences 

 belonging- to the cellular epidermis, and from which project 

 minute bundles of excessively delicate, rigid filaments or 

 setce. These spicules, as they may be termed, were first 

 pointed out by Wilms,* and the species on that account was 

 termed by J. Miiller, S. setosa ; but from Wilms' description 

 it appears to differ in no important particular from S. hipunc- 

 tata. Wilms describes them as constituting a single series 

 on each side, whilst Busch,t in speaking of a form termed by 

 him >S^. ceplialoptera, notices that they are disposed, in that 

 species, in a double series on each side. Krohn also remarks 

 that he has seen these spicular bundles, not only in S. bipunc- 

 tata, but in several other species also ; their existence, there- 

 fore, would seem to be general throughout the genus, and 

 careful observation may, perhaps, educe from their disposition 

 specific characters of some importance. In S. bipunctata, the 

 spicules project on all parts of the body, but they appear to 

 be more numerous on the anterior portion than elsewhere. 

 So far as I have observed, they seem to be scattered irregularly 

 over the surface, although Krohn states that they are appa- 

 rently arranged in symmetrical longitudinal tracts on the two 

 sides. He says also, that they occur on the caudal fin where 

 they are disposed in a curved line across its width. In some 

 species he remarks that they exist also on the posterior 

 lateral fins. 



As has been said before, all these bundles of spicules are 

 placed upon rounded eminences, and in most cases they 

 appear to radiate on all sides from the centre of the eminence ; 

 but closer examination will sometimes show that they are 

 disposed in a simple line, and in close contiguity. This is the 

 case, at any rate, according to Krohn, in S. bipunctata. 



Notwithstanding their rigidity, the filaments, of which 

 these spicular bundles are constituted, have nothing in 

 common either with spines (aculei) with which Wilms com- 

 pares them, nor with set(B, as they are termed by Busch. 

 According to Krohn they are epidermic processes. And this 

 notion he remarks is supported by the circumstance that the 

 spicules, like the epidermis itself, are detached with extreme 

 readiness, and consequently are only to be observed in per- 

 fectly fresh specimens in a good state of preservation. 



Some analogy may, perhaps, be conceived to exist between 

 the filaments of which these epidermic spicules are consti- 

 * L. c, p. 11, fig. 1, 16. t L. c, p. 93. 



VOL. IV. C 



