INTIMATE STRUCTURE OF THE SHELLS OF BRACHIOPODA. 39 



verrucosa, these passages appear to be prolonged inwards, by infundibular extensions of 

 the shell, which project for some distance into its cavity, as shown in the annexed 

 woodcut. 



Family.— DISCINID^ and LINGULID^ 



The structural character of the shell, in these two families, is altogether peculiar. 

 For in the first place it is composed of a great number of very thin laminae, arranged 

 parallel to the surface ; to which plan of structure we have seen that an approach is 

 presented in Crania. Next, the substance of these shells (in the recent species at least) is 

 horny, instead of being calcareous, thus resembling the periostracum of other shells ; 

 and further, it is penetrated by minute tubuli of extreme minuteness, resembling those of 

 dentine, instead of by large perforations. These tubuli run parallel to each other 

 (Plate V, fig. 16, a a), usually in an obhque direction, though near the margin of the 

 shell they may often be seen to be parallel to the surface. In Discina they are commonly 

 arranged in fasciculi, so that their transverse section presents a series of clustered dots, 

 as seen at b b, sometimes, however, presenting the arrangement seen at c ; both of these 

 plans suggest the idea of a cellular structure as the original basis, but of this I have not 

 been able to obtain any satisfactory evidence. 



Attention has been drawn by Mr. Daniel Sharpe to an apparent departure of 

 a very remarkable kind from the usual plan of structure, in the genus Trematis} 

 the shell of which he describes as composed of two layers, the external one being 

 regularly punctuated, like that of the Terebratulidce, whilst the internal is destitute of all 

 trace of punctuations. This he endeavours to show to be only an extension of a plan 

 of structure common among other Brachiopoda, namely, that the internal layer is 

 frequently imperforate towards the centre of the valve. On this general statement I have 

 already had occasion to remark (p. 24, note) ; and have now to show that it is altogether 

 inapplicable to the case before us. For I have ascertained, by careful examination of the 

 shell of Trematis, that it is formed on the general plan of the shells of the BiscinidcB ; and 

 that it has not in reality the slightest relation to the Terebratulidan type. The external 

 markings are of the most superficial character possible, and no more indicate perforations, 

 than do those of Porambonites. They are confined to the outermost of the numerous thin 

 laminae of which the shell is composed ; and although I cannot detect in these the minute 



^ Quarterly Journal of the Geological Society, vol. iv, p. 67. 



