76 



THE VOYAGE OF H.M.S. CHALLENGER. 



tions, either between two neighbouring branches or on the connecting nodal points 

 of the network, and then its configuration becomes more irregular (PL II. fig. 7, h). 

 The meshes of the network are usually roundish or polygonal, sometimes more oblong ; 

 their diameter is very variable, but usually small (PI. II. figs. 5, 6, h). 



Perisarc. — The chitinous tube which surrounds the tubular branches of the reticular 

 hydrorhiza is of special physiological importance to the symbiotic Deep-sea Keratosa, 

 since it replaces the absent strong spongin-fibres. The network of the perisarcal tubes 

 forms the firm pseudo-skeleton of the soft sponges, and constitutes the solid framework 

 which supports their canal-system. It is very probable that the absence of the usual 

 strong spongin-fibres in these Deep-sea Keratosa is effected by the association with the 

 symbiotic Hydroids, the growth of which determines the form of the sponge. 



Coenosarc. — The soft and delicate epithelia of the coenosarc (ectoderm and entoderm) 

 hidden in the chitinous tubes of the perisarc were usually scarcely recognisable, and 

 more or less destroyed in the specimens examined ; they presented the same difficulties 

 in examination as the epithelia of the sponge itself, being much injured by the conditions 

 of capture and the sudden change of the physical conditions of existence. In a few 

 cases, however, they were tolerably well preserved, and I was able to convince myself 

 that the wall of the tubular coenosarc possesses the same structure as in the smaller 

 Hydroids. A striking character of these deep-sea Hydroids is the dark coloration of 

 the coenosarc produced by the accumulation of brown, greenish, or blackish pigment- 

 granules. These are very similar to the phseodella, or the peculiar pigment-granules, 

 which constitute the pha3odium, or the extra-capsular pigment-body of the Phseodaria, 

 described in my Report on the Challenger Radiolaria. 1 The striking similarity of these 

 dark pigment-bodies, and their general presence in the coenosarc and the hydranths 

 of the symbiotic Hydroids, caused some naturalists, who examined these Keratosa, to 

 declare them to be " large-sized Rhizopods with reticular tubes filled up by phseodia." 

 This mistake is the more conceivable, as usually the epithelia of the coenosarc are 

 destroyed, and their scattered pigment-granules fill up the cavity of the perisarcal tubes. 



Hydranths. — The nutritive zooids of the symbiotic deep-sea Hydroids are small, and 

 were in all the specimens examined highly contracted, usually more or less injured, so 

 that it was no easy task to recognise their true nature with certainty. This was possible, 

 however, in the case of Stylactis spongicola inhabiting Stannophyllum, and especially 

 in those species in which the pseudo-skeleton is composed of Globigerina ooze. After 

 having dissolved the calcareous matter by hydrochloric acid, I could observe hundreds of 

 hydranths arising from the superficial layer of the hydrorhiza, and prominent on the 

 dermal surface of the sponge. The hydranths were ovate or club-shaped, sessile in 

 Stylactis spongicola, shortly pedunculated in Stylactis abyssicola, and had a diameter 

 of 0'2 to 0'3 mm. in the former, 0'5 to 0"6 mm. in the latter. The tentacles were 



1 Zool. Chall. Exp., pt. xl. vol. xviii. p. 1532. 



