the Skeletons of Mollmca, Crustacea and Echinodermata. 385 



parallel to its surface ; from this network branches seem to pass 

 both upwards and downwards, as if to communicate with corre- 

 sponding reticulations in other layers. The most characteristic 

 example of this kind which I have met with, is in the genus 

 Lima, of which the whole shell is minutely traversed by these re- 

 ticulated tubes (fig. 5). In most families whose shells are entirely 

 membranous (that is, destitute of the prismatic cellular structure), I 

 find the tubular structure present in some genera and absent in 

 others ; so that I am inclined to think that it may serve as an 

 important aid, in separating genera which are otherwise closely 

 allied. 



The tubular structure is often well preserved in fossil shells, 

 and is distinctly seen in the Avicula longicostata already referred 

 to. As this example is one of peculiar interest, in regard to the 

 value of the microscope in determining the nature of fossil shells, 

 I subjoin a summary of the characters by which I judge it to be 

 not an Avicula, but nearly allied to Lima or Plagiostoma. 



Avicula y and other > • , ^ , . Lima, Plagiostoma. 



MargaritacecB. ^^^^"^« <^ longicostata. and allied PectinidcE. 



External layer. External layer. External layer. External layer. 



Prismatic cellular No cellular struc- Coarsely plicated Coarsely plicated 



structure. ture. membranous membranous 



structure. structure. 



Internal layer. Internal layer. Internal layer. Internal layer. 



Nacreous. Not nacreous. Tubular struc- Tubular structure. 



ture. 



It is worth notice, that in the genus Pecten there is not the 

 slightest trace of the tubular structure, whilst the shell of Lima 

 is more completely tubular than any I have examined. Hence 

 we might consider the difierence between them to be more than 

 generic; a conclusion which harmonizes exactly with that of M. 

 Deshayes, founded upon the structm-e and habits of the animals 

 by which these shells are respectively formed. 



The foregoing may be regarded as specimens of the important 

 results which I think may be anticipated, from the application 

 of the microscope to the structure of shells. When once marked 

 differences in the arrangement of the organic elements of shell 

 are shown to exist, the question naturally arises, how far these 

 may be adopted as characters in classification ; and I think it will 

 be evident, from the facts I have stated, that this may be done 

 to a very considerable extent. There are several famiUes among 

 which I have not yet discovered well-marked points of difference, 

 especially in Univalve shells, and in Lamellibranchiata with the 

 lobes of the mantle united; but I have little doubt that such 

 characters may be ascertained, by a more extended inquiry than 

 I have yet had the opportunity of making. Except in the case 



