50 



deduction of still more important hypotheses. It is evident that 

 for the purpose of examining into the structure, recent specimens 

 are far superior to fossil ones. 



Having repeatedly and carefully examined the structure of the 

 shells of the Arachnoidisci in various ways, I purpose detailing 

 a statement of the result of my investigations, particularly as in 

 some few points I differ with the conclusions arrived at by other 

 observers. 



I trust that if the matter of this paper should be deemed uninte- 

 resting, I may be pardoned, from its being the means of placing in 

 the Society's possession the accompanying very beautiful drawing, 

 most accurately executed by Mr. Legg, who kindly undertook to 

 assist me by performing so difficult, but necessary, a part as the deli- 

 neation of the objects to be described. 



The following observations apply to the shells alone, and not to 

 the animals inhabiting them (if animals they be), for although the 

 specimens I submitted to examination were recent, they were not in 

 my possession until they had been dried for some months. 



These shells are not, strictly speaking, bivalves, although capable 

 of being separated into two corresponding portions ; but are more 

 properly multivalves, each shell consisting of two discoid portions 

 and two annular valves exactly similar respectively to one another. 



Each of the discoid valves is also capable of a further separation 

 into two circular, but dissimilar portions, which separation, however, 

 is by no means easy of accomplishment, owing to the very fragile 

 nature of the substance of the shell : in fact, I have only been able 

 to effect this separation by accident, although this I have done suffi- 

 ciently often to satisfy me that all the discoid valves are composed 

 of the two portions adverted to, and this fact is capable of very dis- 

 tinct verification by another means to be presently described. 



The two parts of the disks consist of, firstly, a very thin membrane, 

 somewhat flexible and elastic, and capable of resisting the action of 

 boiling nitric acid, and on this membrane are the characteristic 

 spider's-web-like markings, which have given rise to the name of the 

 genus. It is situated externally in both disks of each individual. 

 Plate XI., fig. 1. 



Secondly, each disk is composed of a siliceous frame-work, admi- 

 rably adapted to support and strengthen the outer membrane, and 

 bears a very close resemblance to a circular Gothic window. Plate 

 XI., fig. 2. 



Each of these portions I have been able to obtain separately, by 



