sroxGiiDA. 419 



DESMAC1I)LXII)J<: (Schmidt, 1870). 



If all those sponges which contain hooked or bow-like flesh- 

 spicules were, in accordance with Vosmaer's views, as expressed 

 in his very useful Revision (Notes Roy. Mns. Netherl. ii. p. 99), 

 included in this family, it would not only be the largest, in all pro- 

 bability, of the families of Siliceous Sponges, but it would leave 

 some of the remaining ones mere skeletons. Judged by the facts 

 now known, the boundary region between the Desmacidinidae and 

 Chalinidae is now narrow, but not in reality so narrow as it would 

 be if the above definition is insisted on. ' Whatever may be the 

 affinities of Homuodictya, with its anchorate flesh-spicules (referred 

 by Mr. Carter to the Chalinidae), those of To.rochaUna, mihi (see 

 Chalinidae, supra), are undoubtedly with that group; yet it has 

 a bow-like flesh-spicule in conjunction with a Chalinid acerate 

 skeleton-spicule, horny fibre, and digitate habit. Until the homo- 

 logies of the flesh-spicules are better understood than they are at 

 present, I believe that cases such as those just mentioned will 

 have to be considered separately on their individual merits as they 

 arise, having special regard to the direction in which the greater 

 assemblage of affinities point. It seems probable that this family 

 will only prove a fresh illustration of the maxim " Xatura non 

 facit saltum." Besides Toxochalina I here exclude from the family 

 those genera (e. g. Claihria, Acarnus, Echinonema) in which anv 

 of the spicules project laterally from the fibre ; such forms as these 

 seem to pass by gradations (Echinodictyum, Raspailia) almost into 

 A.riiuJJa and Phacellia, by losing, in the first case, the flesh- 

 spicules, and in the second [Axinella &c.) the spined echinating 

 cylindrical. Rhizochalina, on the other hand, seems linked to the 

 family by its occasionally horny fibres, and by its ally Oceanapia 

 with its bihamate flesh-spicule ; and I have ranged it (although only 

 provisionally) here as a degraded Desmacidine. It probably owes 

 its peculiar form to its mud-loving habits. Two new generic types, 

 Gelliodes and Iotrochota, are described below. 



RHIZOCHALINA. 



Schmidt, Atl. Geb. p. 35. 



Phlceodictyon, Carter, Ann. fy Mag. Nat. Hist. 1882, x. p. 122. 



This form is so aberrant in its coarser anatomy that I think there 

 can be little doubt that Carter ha s done right (I.e.) in making it the type 

 of a distinct group, although we have as yet no satisfactory information 

 about the arrangement and structure of the soft parts. Although I 

 can see no sufficient reason why the name Oceanapia, Korman, should 

 give way to the above names for such species as Desmacidon jeffreysi, 

 Bowerbank, whose spiculation includes a bihamate, yet it seems 

 not undesirable to retain the older of the two for those which have 

 simply an acerate spicule. With regard to the question of syste- 



2e2 



