Mr. H. E. Strickland on the genus Cardinia. 103 



shells of the coal-measure fossils are much thinner and weaker 

 than in those from the lias. Fourthly, the shells from the coal- 

 measures rarely exhibit any trace of a lunule, and when present 

 it is more diffused and indistinct than in the liassic species. 

 Lastly, the Cardinia from the lias were wholly marine in their 

 habits, while there are strong grounds for believing that the spe- 

 cies from the coal-beds inhabited fresh, or at most brackish water. 

 This is shown by the fact that these Unio-likc shells are almost 

 invariably found in the beds of shale accompanying the coal, and 

 not in the really marine formations of the same age. Now whe- 

 ther we suppose the coal to have grown in situ like peat, or to 

 have been washed by currents into certain localities (both which 

 theories are no doubt true in certain cases), we cannot deny the 

 coal to be a terrestrial production ; and therefore when we find a 

 particular family of mollusks constantly, and almost always ex- 

 clusively, accompanying the beds of coal, we have a very strong 

 presumption that these animals had a lacustrine or estuarine 

 habitat. 



It is true that in some cases, as in Coalbrook Dale, at Halifax, 

 at Glasgow, and in Belgium, the coal-measures contain an admix- 

 ture of these bivalves with various marine genera ; but this does not 

 necessarily prove them to be marine species, for they may either 

 (as suggested by Mr. Prestwich in his memoir on Coalbrook Dale, 

 f Geol. Proceedings/ vol. ii. p. 405) have been washed down into an 

 estuary and there become mixed with marine shells, or by a depres- 

 sion of the land the sea may have washed the marine shells into 

 the marshes tenanted by these supposed freshwater species. And 

 it is important to remark, that in the carboniferous limestone, a 

 strictly marine formation immediately preceding, and in some 

 cases alternating with the coal-measures, these peculiar bivalves 

 rarely if ever occur. 



Por these reasons I think we ought to abstain from classing 

 the shells of the coal-measures with the well-marked and clearly- 

 defined genus Cardinia of the lias. I do not indeed mean to as- 

 sert that the carboniferous group of shells really belong to the 

 Unionidce, where they were formerly classed, for they want the 

 supplementary anterior muscular impression which distinguishes 

 that family* ; but I think they may be for the present regarded 

 as a distinct family, probably lacustrine, and possibly allied to 

 Unionida, but the precise characters of which, and especially the 

 structure of the hinge, are as yet unascertained. Perhaps Dr. Car- 

 penter, whose researches on the microscopic structure of shells 

 have opened to us a new element for the determination of fossil 



* Mr. G. B. Sowerby, in his ' Genera of Recent and Fossil Shells,' stairs 

 that he could find no difference between the casts from the coal -measures 

 and those which he made from the inside of recent Unios, but he had perhaps 

 overlooked the supplementary muscle of the latter. 



