M. K, A. Zittel on Fossil Hexactinellida. 267 



examples, like the roiiglily silicified specimens, only present 

 us with a picture of the external form ; they are quite unfitted 

 for the investigation of the structural characters. 



Attempts at Classification hy Saville Kent, Carter, and 

 MarsJiall. 



All previous attempts to express the relationships of the 

 Hexactinellida in a systematic classification were necessarily 

 confined to the living representatives of the group, on account 

 the complete want of knowledge of the minute structure of the 

 fossil forms. As Bowerbank includes under the " Fibro- 

 siliceous Sponges" both the Hexactinellida and the Lithistida, 

 and his monograph, which appeared in small divisions, really 

 consists only of descriptions of species, it cannot receive any 

 particular consideration here. 



The first attempt at a system is due to Saville Kent *. 

 According to this author the Hexactinellida are divisible into 

 two suborders : — 



CORALLIOSPONGIA, Gray. 



Sponge-body Avith a skeleton consisting of interwoven or 

 isolated spicules, never reticulate and coherent. Gemmules 

 without spicules, 



Calicispongia, S. Kent. 



Sponge-body with an anastomosing or coherent reticulate 

 skeleton. Reproductive gemmules membranous, without 

 spicules. 



W. Marshall t has already given expression to the doubts 

 that exist as to the systematic estimation of the gemmules, 

 about which so little is known. With fossil forms this cha- 

 racter is, of course, useless. But the grouping of the genera 

 also a])pears not to be natural. Thus in the first suborder we 

 have Euplectella and Hahrodictyon, together with Farrea and 

 A2}hrocallistes ; whereas the first two evidently show much 

 nearer affinities to Hyalonema, Askonema, IloUenia, &c. That 

 the Lithistid genus ' MacAndrewia, Gray, still finds a place 

 among the Coralliospongia, is probably due to an oversight. 



A complete revision of the living Hexactinellida known up 

 to the year 1873, with a systematic arrangement, was pub- 

 lished by Carter in two excellent memoirs on Hexactinellida 



* Monthly Microsc. JoiU'n, vol, iv. p. 242. 

 t Loc. cit. XXV. p. 146, 



