Relationships of the Porifera. - 251 



this effect, though very sparingly. I refer here to Stylocordyla 

 borealis (LoY.), Wyv. Thomson, in which Loven has actually 

 found the rudiments of the axial cross *. And do not most 

 of the spicules of this sponge indicate that spicules which 

 have such a swelling are originally descended from six- 

 rayed forms? I might further point to the peculiar spicules 

 of Suberites lobice^ys, O. S. f The so-called " anchor-spicules " 

 [M. ta. <^<90] are, as is well known, very abundant in Hex- 

 actinellids ; and the question arises, how far are the similar 

 structures in the Tetraxonina related to them ? The same 

 holds true of the little chelate spicules f of the Desmacidonidse. 



The possibility that all originally descended from Hexac- 

 tinellid-like ancestors, and that the occasional isolated appear- 

 ance of spicules with remnants of a triaxonid form may thus 

 be attributed to atavism, is to me personally not unlikely. 



This occasional appearance is much more common than 

 people think ; I have found many sponges in vihich there are a 

 few rudimentary six-rayed spicules lying among the normal 

 ones. This is somewhat analogous to the case of men with 

 rudimentary tails or with extraordinary hairiness. And so it 

 is with those Halichondrine forms which, in the arrangement 

 and form of the parts of the skeleton, recall certain Suberitidse 

 {Clavulina). The common characters of all the three orders 

 named are more numerous and more important than those of 

 the two classes, and thus they are more intimately connected. 



Coming now to consider the closer relationship of the forms 

 within each order, no one will take exception to the Hyalo- 

 spongi^ as a natural group. The mutual connexion of the 

 Spiculispongige rests upon the following grounds : — The 

 examination of the difierent suborders of the Spiculispongige 

 appears to me to show unmistakable signs of degeneration. 

 Leaving out of account the Lithistina, which, owing to their 

 peculiar knotty structures, stand somewhat on a separate 

 footing (although the condition of the canal-system and 

 ground -substance, as well as the often well-marked tetraxonid 

 skeletal elements, distinctly show their affinities), we may 

 perhaps assume that the Tetraxonina represent the older 

 forms. The presence of distinct tetraxonid spicules, a more 

 or less distinctly radiate arrangement of the skeleton, a more 

 or less distinctly pronounced cortex, the granular character of 

 the ground-substance, and a rather highly developed canal- 

 system are their characteristics. We see all this most 



* I may in the meanwliile refer to my work " Sponges of the ' WiUem 

 Barents ' Expedition," 1880 and 1881, pp. 10-12. 

 t Spoug. atlaut. Gebiet. p. 47, pi. v. tig. 5. 

 i Viatic? and anc. ana. 



17* 



