1865.] DAWSON COAL-FOEMATION. 143 



represents only a fraction of the vegetable matter which was actually- 

 produced. It has been estimated that it would require eight feet of 

 compact vegetable matter to produce one foot of coal*; but if we 

 reckon the whole vegetable matter actually produced in the process, 

 I should suppose that five times that amount would be far below 

 the truth, even in the most favourable cases ; while there is evidence 

 that in the Carboniferous period many forests may have flourished for 

 centuries without producing an inch of coaly matter. 



15. Bivalve Shells. — All the Lamellibraiichiate shells, which are 

 so numerous in some of the shales and bituminous limestones of the 

 Joggins that some of the beds may be regarded as composed of 

 them, belong to one generic or family group. They are the so- 

 called Modiolas, Unios, or Anodons of authors. I proposed for them 

 some years ago the generic name of Naiadltes, and described six 

 species from the Coal-measures of Nova Scotia, stating my belief 

 that they are allied to UnioJiicla', and that their nearest analogue may 

 be the genus Bysso-anodonta of D'Orbigny, found in the Kiver Pa- 

 panaf. Mr. Salter, however, to whom I sent specimens, regards 

 these shells as belonging to his new genera Antliracomya and Antlira- 

 coptera, the former being supposed to be allied to MyadceX- More 

 recently Giimbel and Geinitz have described similar shells from 

 Thuringia as belonging to the genera Unio and Anodon, and regard 

 my Naiadltes carhonariiis {Antliracoptera carhonaria of Salter) as a 

 Bre{ssena§. As these shells swarmed in the waters of the Coal-for- 

 mation estuaries or lagoons, facts tending to the elucidation of their 

 habits and affinities are important with reference to the coal ; I would 

 therefore make the following remarks in relation to them : — • 



(1) Under the microscope, the shells of the thicker species, asiV«m- 

 dites carhonarlus, present an internal lamellar and subnacreous layer, 

 a thin layer of vertical prismatic shell, and an epidermis — these 

 structures being entirely similar to those of UnionidcK. In the 

 thinner species, as in N. Icevis, only the prismatic coat appears, and 

 in this the prisms are in some instances placed obliquely. These 

 thin sheRs, however, show evidence of an epidermis. (2) The 

 ligament was external, there seem to have been no teeth, the shell 

 was closed posteriorly ; but there are indications of a byssal sinus. 

 Mr. Salter describes the epidermis as wrinkled posteriorly ; but this, 

 with the exception of the rings of growth, appears to me to result 

 from pressure. The shells are equivalve, and have the external 

 aspect of Unionidce or Ilytilidce. (3) I know of no instance in 

 Fova Scotia of the occurrence of these shells in the strictly marine 

 limestones, nor have any properly marine forms of MoUusca been 

 found with JSfaiadites in the Coal-measures. (4) The mode of their 

 occurrence precludes the idea that they were burrowers, but favours 

 the belief that they were attached by a byssus to sunken or floating 

 timber. On the whole I think that the balance of probability is in 



^ Dana's Mnniial, p. 367. 



t Supplement to Acadian Geology, 1860. 



I Quart. Journ. Geol. Soc. vol. xix. p. 79. 



§ Neues Ja,hrbat;h, 1864. Geological Magazine, May 1865. 



