10 CARBONICOLA, ANTHRACOMYA, AND NAIADITES. 
and their very varying forms, I have come to the conclusion that either they were 
capable of inhabiting both lakes and estuaries on the one hand, and the open sea 
on the other, or else the marine and fresh water forms are so similar in appearance 
that they can only be identified by reference to those which may occur along with 
them in the bed. Thus when, (as in Coalbrookdale and Lancashire), {I doubt the 
latter locality’] we find them associated with undoubted marine forms, we can 
only conclude that they themselves were inhabitants of the sea; but if they 
happen to occur unaccompanied by such well-recognised forms, then we may 
assume that they represent lacustrine or estuarine conditions, the probability 
being that had the strata been formed under the sea marine shells would have 
been preserved along with those of this genus.” I have shown above that the 
mixture of marine forms with Carbonicola (Anthracosia) is by no means so 
common as might be supposed from this remark, and that it may be accounted 
for in other ways than those advanced by the Professor; and I am sure that the 
true reason of this group of shells being a “‘ bane to paleontologists”’ is because 
they have never been systematically collected and studied. It is perfectly 
astonishing to see the paucity of specimens of these shells in museums, and more 
so in the museums of large towns situated in coal-measure districts. It is, 
indeed, a matter of very considerable difficulty to obtain these fossils in situ, and 
unless the greatest care is observed, and the collector possesses an intimate 
acquaintance with the naked-eye characters of the various shells and rocks on the 
waste-heaps, it is very easy to infer that fossils which lie close together on the 
mound, belonging to widely separated beds, are from the same horizon. Another 
obstacle arises from the carelessness of many collectors in not labelling each 
specimen immediately it has been brought home, and often putting them by with 
others, so that shells from various beds become mixed and the horizons forgotten. 
I can testify myself to the frequent difficulty and almost impossibility of arriving 
at any but the approximate horizon of many specimens ; often I have found some 
of my best specimens lying weathered-out on the rubbish-heap of a pit-bank years 
after the pit has ceased working, and which, from the hardness of the matrix and 
1 On the authority of Prof. Phillips (article on ‘‘ Geology,” ‘Encyclopedia Metropolitana, 
1834, p. 590), who, in speaking of the marine shells in the Lower Coal-measures of Lancashire and 
Yorkshire, says, “In the midst of this series of Gannister Coal two layers of these shells [Unio] 
occur, one of them about the middle of the series, considerably above the Pecten Coal, and the other 
near the bottom, considerably below that coal.” This extract is quoted by Binney in two papers on 
“The Marine Shells of the Lower Coal-measures” (‘ Manchester Geol. Soc. Trans.,’ vol. i, p. 82, 
1841; vol. ii, p. 75, 1860). Mr. Bolton, assistant keeper of Owens College Museum, writes me :— 
“In the Rossendale area of the lower coal-measures a very persistent bed of ironstone shale with 
ironstone, the whole charged with Carbonicola in the state of casts, occurs immediately over the 
Bassey or Salts Mine. The same bed has been found at Helpit Edge, Saddleworth. Iam sure that 
careful search would prove its existence in the Haslingden district.” 
