812 AY. J. SOLL.VS ON THE STBUCTUKE AND 



the elongation of the spicules seems to be at the expense of the 

 material of their articular processes. The shafts of the spicules arc 

 not seen, and may bo inferred to penetrate the stem at right angles 

 to its surface As, then, the long rays all lie in the same direction 

 (that is, with the length of the stem), and as the terminations of the 

 spicules overlap one another by passing above and below those 

 alongside them, and as the shafts appear to penetrate amongst the 

 spicules of the interior, we may consider that the arrangement we 

 have here is of the nature of a " plait," serving to keep the spicules 

 in place, and yet not binding them together with the same rigid 

 union which we find in the network of the sponge-body. From this 

 results flexibility combined with security, the value of which will be 

 understood when we remember that the Siphon ice lived in a some- 

 what shallow sea (75 to 375 fathoms), and were exposed to 

 currents which a flexible stalk would be better able than a rigid 

 one to sustain. 



Longitudinal Section of the Stem. — This docs not expose any very 

 clearly defined structure ; but what there is to be seen agrees com- 

 pletely with the foregoing description. 



Development. — Amongst the specimens from Haldon is a very 

 small one (PI. XXY. fig. 3), which I believe to be a young form, and 

 which agrees in every particular with similar ones figured by 

 Sowerby as the young forms of his S. pyriformis from Blackdown, 

 near Cullompton. The body is somewhat fusiform, 1 of an inch 

 long and nearly |- of an inch broad ; it is supported on a straight 

 slender stem, the proximal end of which is broken oif. At the apex 

 of the body is a small conical depression produced by the oscular 

 openings of some four or five excurrent canals, and very insignificant 

 in size when compared with the rest of the body. 



From this youngest known form the adults arise by successive 

 coatings on tiie exterior, the coatings on the body being thicker 

 than those on the stem. Each coating possesses all the characters 

 which belong to Siphonia; and thus we have produced successive 

 groups of longitudinal canals opening in a series of tiers vertically 

 over one another in the cloaca or central cavity left in the axis of 

 the body of the sponge. Thus, also, the longitudinal canals are 

 axial in direction below the cloaca, and become more parallel with 

 the curve of the existing exterior surface as they lie nearer to it, 

 the successive groups of canals indicating, indeed, the successive 

 surfaces of the sponge. Thus also arise the concentric rings of 

 skeletal network seen in transverse section around the cloaca ; and 

 by the opening in different places of fresh radial canals for every 

 fresh exogenous layer, the discontinuity of the radial canals when 

 traced for any distauce results. Fig. 1 b (PL XXY.) is a drawing of a 

 section taken longitudinally from an adult S. pyriformis, Sow. : 

 it is easy to understand how the series of changes just described 

 would result in producing such a structure as this from the young 

 form represented in fig. 3. 



The differences presented in the surface of many fossil S!pho)ii<r, 

 differences which have been held of specific importance by many 



