1847.] LYELL ON THE COAL-FIELD OF EASTERN VIRGINIA. 2/5 



surface concentrically furrowed and somewhat wrinkled. They re- 

 semble Posidonomya, and more than any other the P. minuta. They 

 occur in such immense numbers as to divide the shale-like plates of 

 mica into very thin laminae. Every fresh surface exhibits a layer of 

 them. In their gregarious character they resemble the genus Cy- 

 pris as it occurs in the cypriferous marls of Auvergne. 



With them is another smaller and more convex bivalve shell which 

 resembles a young Astarte, but may perhaps be the young of the pre- 

 ceding (see 6, fig. 6). The same shell was also presented to me by Dr. 

 Werth, from the Creek mines south of Blackheath, also in black shale. 



At the Deep-run pits I saw traces of the same shell, but in none of 

 these places in a perfect state : the genus Posidonomya in Europe is 

 considered to be characteristic of the triassic period, in which the 

 P. minuta occurs in England (see fig. 4. plate 28. vol. v. Geol. 

 Trans. 2nd series). I have stated, that the manner in which these 

 extremely thin shells occur by myriads, like plates of mica between 

 the partings of the shale, reminded me of the exuviae of the Cypris 

 in the freshwater marls of Auvergne. Mr. Morris has shown me a 

 recent bivalve crustacean from Bahia Blanca in South America, having 

 a delicate, crumpled or corrugated horny shell, bearing so great a re- 

 semblance to the Posidonomya, that I cannot help sharing his doubts 

 whether the latter genus may not be allied to Cypris rather than to 

 any genus of the true Mollusca, i. e. to Avicula. 



Fossil Fish. 



After much careful search, and the indefatigable exertions of my 

 friend Mr. A. F. D. Gifford, residing at the Blackheath mines, I was 

 unable to procure the remains of more than three species of fossil fish, 

 most of which, and the only perfect ones, are referable to a new genus 

 of the homocercal class. On seeing these ichthyolites (September 

 1846), M. Agassiz, ignorant of their locality, at once suggested, from 

 the analogy of European forms, that they were of the age of the lias ; 

 as being all homocercal, and the large smooth scales of one species 

 belonging to the genus Tetragonolepis. This opinion would accord 

 with the inference previously deduced from the fossil plants by Pro- 

 fessor W. B. Rogers (Trans. Assoc. American Geol. 1840-42, p. 298). 



As all the fish hitherto obtained from the sandstone (commoidy 

 called the '* new red ") of the Connecticut River, four or five species in 

 number, are heterocercal, this would indicate a distinct and proba- 

 bly a more modern date for the Virginian coal-bearing strata. Could 

 we therefore feel sure that the beds containing ichthyolites and the 

 foot-prints of birds in the valley of the Connecticut were triassic, 

 it would afford strong ground for presuming the oolitic age of the 

 Virginian coal strata. But I conceive the question is still open, 

 whether the Connecticut sandstone be of Permian or triassic date. 



I shall now offer a description of the Virginian ichthyolites, drawn 

 up chiefly from notes supplied me by Prof. Agassiz and Sir Philip 

 Egerton. 



Bictyopyge macrura, PI. VIII. This species is the most abundant, 

 occurring in strata of argillaceous sandstone about 1 50 feet above the 



