96 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [JuRC 14, 



witli small circles similar to those just mentioned, but occasionally 

 surrounded by a narrow translucent interspace resembling in colour 

 the spar which filled the cavities, and the minute area within the 

 circles. A third portion, laid open nearly to the centre, differed not 

 in character from translucent slices. A slightly-abraded reverse sur- 

 face of an Atherfield specimen, presented to the Geological Society 

 by Prof. Edv/ard Forbes and Capt. Ibbetson, exhibited (fig. 23) ver- 

 tical rows of opake-white links with narrow, translucent, intermediate 

 lines ; and a little lower, where the surface had been more deeply 

 exposed, the linlcs were not arranged lineally, and were detached from 

 each other by a very narrow pale band. Part of a thin slice displayed 

 also several detached circles, the inner areas and intervening spaces 

 agreeing in colour and amount of clearness. Again, a fragment pur- 

 posely worn down unequally, gave in the upper or least abraded por- 

 tion, irregularly distributed circles, either in contact or separated by 

 very fine pale lines ; while lower down a deeper wearing laid open first 

 circles of smaller dimensions with an increased interspace ; and still 

 nearer the lower extremity, intersections of three tubes with gra- 

 dually narrov»'ing intervals in which circles could just be detected. 

 From these observed characters, it is inferred that the meshes of the 

 exterior network were not mere pores in the outer layer of the coral, 

 but tubuli possessed of as distinct walls as the visceral tubes. 



Additional cavities, so far as concerned the upward growth of the 

 coral, were developed chiefly in the middle of the branch, converging 

 to the very centre as they descended ; but occasionally a minute open- 

 ing surrounded by mature tubes w^as detectable near the outer or re- 

 ticulated zone. 



As regards the position of Siphodictyur,i in a general classification, 

 little doubt can be entertained of the coral having been formed by an 

 ascidian polype. The simple abdominal cavities, uncrossed by trans- 

 verse laminae, were clearly adapted for the reception of digestive 

 organs, which did not change their position ; and the length of the 

 tubes is not greater than that requisite for complicated viscera with 

 rigid tentacula. Among the great groups comprising that class, the 

 nearest agreement is with the family Tubuliporidse of M. Milne Ed- 

 wards, and to which he has removed Hornera (pp. cit.). Some of the 

 best-know^n genera however, as Tubalipora and Crista, do not thicken 

 conspicuously, if at all, the exterior of their tubes ; and the mode in 

 which the process is effected in Hornera differs, as already stated, 

 materially from that in Siphodictyum. In this respect the fossil 

 resembles the Escharidae, the additions being purely local, and pro- 

 ducing external irregularities of surface ; but the absence of all signs of 

 an operculum essentially distinguishes the extinct coral from Eschara 

 and allied genera, and associates it with those forming the Tubulipo- 

 ridee*. It is therefore proposed to place Siphodictyum provisionally 

 in that family of ascidian zoophytes. 



The fifth fossil had not apparently been named ; but when first 

 examined, it was conjectured to be that which is alluded to in the 



* Consult M. Milne-Edwards, op. cit., Mem. sur les Tubulipores, p. 5. 



