S04 W. J. SOLLAS ON XIIE STRUCTURE AND 



fossils of the Chalk and the living forms of the Atlantic, seems here 

 to have been led astray : the tubes of Cortosphwra arc external, and 

 those of Qhoanites internal, to the sponge-body ; and the peculiarity 

 in the junction of these tubes with the body of the sponge in Coelo- 

 splicera is not alleged to have been observed in the case of Choanites. 

 For my own part I certainly have never seen it. But it would be use- 

 less to argue the matter further on these grounds, since Ccelosphcera 

 is, in spiculation, closely allied to Halichondria incrustans, Johnst., 

 while Choanites, on the other hand, possesses the genuine Lithistid 

 skeleton, and belongs without doubt to the genus Siphonia. 



3. Description. 



General Form and Structure. — The outward form of Siphonia is 

 exceedingly variable ; and it is by not making clue allowance for the 

 extreme polymorphism of the genus that its species have been so 

 extensively multiplied. Its principal part consists of a head or body, 

 which is usually, but not always, supported on a distinct stem. 



The stem, when present, is more or less cylindrical, straight near 

 the head, but generally irregularly undulating lower down ; it ex- 

 hibits great variation both in length and breadth, sometimes becoming 

 so short as to render the sponge-body almost sessile (S. curta, cylin- 

 drica, etc., Court.), at others attaining a length many times that of 

 the sponge-body — e. g. in S. Webster I, Sow., which presents us with 

 a slender stalk some four or five feet long ; between these two 

 extremes eveiw intermediate gradation may occur. It is usually 

 simple, but sometimes becomes branched (S. arbuscida). 



At its proximal extremity it breaks up into a number of diver- 

 ging irregular ramifications, by which it appears to be attached to 

 the surface of some foreign bod3 r . 



It rarely is found entire, having in most specimens been broken 

 off at a greater or less distance from the head. When absent, as it 

 sometimes is in S. {Choanites) Konigii, it is replaced by a number 

 of rooting fibres which ramify in all directions through the sur- 

 rounding matrix of the fossil. With this substitution of anchoring 

 filaments for a process of attachment may be correlated the fact that 

 S. Kbnigii is the chalk or deep-sea form of its genus, while the 

 species provided with stalks are characteristic of greensand deposits, 

 and consequently flourished in a somewhat shallow sea. 



The same kind of adaptation is exemplified in the case of Euplec- 

 tella, which is an anchoring sponge when it floats half immersed in 

 the chalk-ooze of the Atlantic, but becomes fixed and adherent when 

 it enters the shallower waters near the coast. 



The sponge-body presents almost every possible variety of form. 

 Commencing with S. pyriformis, Goldf. (tab. vi. f. 7 a), we have a 

 head nearly spherical in shape ; this by elongating vertically gives 

 rise to a series of more or less prolate ellipsoids, S. ovata, Court., 

 nuciformis, Mich., pyriformis, Sow. (loc. cit. f. 3 & 9) ; by becoming 

 flattened at the extremities these assume a cylindrical shape, S. cy- 

 lindrica, Court. On the other hand, a shortening of the globular 



