REPORT ON THE CALCAREA. 17 



disposition of their spicules. In one case — Hseckel's " articulated" (" gegliedertes ") tubar 

 skeleton — the spicules, usually triradiate, lie in the walls of the radial tubes in many rows, 

 one row behind the other, their basal ray being directed towards the dermal extremity of 

 the tube, whilst the lateral rays diverge towards its gastric ostium ; in the second case — 

 Haeckel's " non-articulated " (" uugegliedertes ") tubar skeleton — there are no transverse 

 palisades, formed by the lateral rays of the triradiate or quadriradiate spicules (Sycandra 

 arctica, H.) ; moreover, there are no special tubar spicules, but only tubar rays of the 

 dermal or subdermal and subgastric triradiate or quadriradiate spicules. For instance, in 

 S-yciUa cyathiscus (Kalkschwamme, Bd. iii., pi. xliii. fig. 9) the subgastric quadriradiate 

 spicules lying with their facial rays in the gastric wall, parallel to its inner surface, send 

 their apical rays centrifugally towards the dermal surface, and these apical rays joining 

 the centripetally directed apical rays of the corresponding subdermal quadriradiate 

 spicules form the supporting skeleton for the radial tubes. If the question be asked, 

 which Sycones must be considered to be more primitive, these with articulated tubar 

 skeleton or those with unarticulated one, there can be but one answer : — these with the 

 articulated. 



The embryology of the Sycones of the type of Sycilla cyathiscus remains certainly 

 still to be explored, but there are in the development of the Sycones of the second type 

 no phases which would lead us to assume that the non-articulated form of tubar skeleton 

 might be a primary one. On the contrary, the ontogeny of Sycandra raphanus shows 

 obviously that Prof. Haeckel's conjecture that Sycetta primitive/,, Syccdtis conifera, 

 and, generally speaking, the Sycones with free radial tubes are the most primitive forms 

 of the family, is quite in harmony with the facts of the case. Now, the tubar skeleton 

 in the forms with free radial tubes is always articulated, and, as it is evident that a 

 non-articulated tubar skeleton could not have been developed before the formation of 

 a cortex which was produced by the fusion of the distal ends of the radial tubes, there 

 can be no doubt that the articulated tubar skeleton has brought about the possibility 

 of the formation of the non-articulated, and not vice versa. For the rest, this can be 

 proved by means of anatomical comparison, there being amongst the Sycones hitherto 

 described many intermediate stages connecting the extreme forms of these two different 

 types. From Sycetta primitive/,, through Sycandra coronata and Sycandra raphanus, 

 we come to Sycandra arctica, the radial tubes of which, grown together in their superior 

 part, give rise to a structure closely resembling a cortex, and it is only the form and 

 disposition of the spicules of its skeleton which force us to regard the outer surface of 

 this sponge as still lacking an independent cortical layer. In Sycandra compressa, 

 Sycortis laevigata, Sycetta strobilus, &c, we already find specimens with a perfectly 

 distinct cortical layer, provided with a auite independent skeleton. This latter consists 

 certainly of spicules homologous with those in the distal end of the radial tubes of, 

 for instance, Sycandra arctica, but so much modified, and having adopted such a 



(zool. CHALL. EXP. — part xxxv. — 1883.) Aa 3 



