536 



J. F. BLAKE OX THE UPPER 



and Rhynchonella corallina were noted. On the same level these 

 change into Diceras-beds of characteristic material, with many ex- 

 amples of that genus ; and finally, at Merry-sur-Yonne, an immense 

 cliff, 200 feet in height, and carved into fantastic shapes, frowns over 

 the river, and is geologically an imstratified heterogeneous mass of 

 Coral Rag, Diceras-beds, and rubble all together. Huge masses of 

 Thamnastrcea, delicate branches of Car yophy Ilia, and fan-like growths 

 of Thecosmilia here and there ornament the surface ; and in the inter- 

 spaces innumerable specimens of Dicer as, Gardium corallinum, stout 

 spines of Gidaris Jlo rig emma, andTricJiites a foot in length are crowded 

 together. It is the finest example of a Coral Rag visible either in 

 the basin of Paris or anywhere in England, and it is, undoubtedly, 

 impossible to confound it with the much higher beds described as 

 occurring at Bailly. Rocks of this description form the boundary 

 of the picturesque valley through which the railway runs past 

 Chatel Censoir to Coulanges. At the latter place there is an almost 

 equally fine cliff, composed of a white limestone as beautiful as that 

 of Angy, though apparently on a lower level, being really only 

 a modification of the Coral Rag of Merry. Here, too, the fossils are 

 in profusion and perfect preservation ; the finest examples of various 

 species of Diceras occur, associated with Nerincea Cotlaldina ?, N. 

 subnodosa, Gardium corallinum, Hinnites, Terebratula insignis, 

 Gidaris florigemma, corals growing in reefs, and sponges. This 

 great corallian development proves conclusively that Diceras, 

 Nerincea, and Gidaris florigemma are not always characteristic in 

 any sense of distinct horizons, and that the lowest portion of any 

 possible rocks which may be referred to the Corallian may be 

 characterized by the latter species. If the series has been rightly 

 traced in our journey westwards, these great coralliferous masses 

 must be the grander development of the coral-growths of the base 

 of the series in the Aube, only represented by thin bands at the 

 base of the lithographic limestones in the valley of the Armancon. 

 Much discussion has arisen on the correct position of these lower 

 coral-beds. At first they were confounded with the upper Diceras- 

 beds and supposed to be repeated by some fault (3, 6). Certain 

 marly limestones developed at Vermanton, with a proportion of 

 Oxfordian fossils, were, at that time, considered to belong to Ox- 

 fordian strata (9, 14). These Yermanton limestones were after- 

 wards proved by Raulin to overlie the Coral Rag of Chatel Censoir 

 and Merry (15) ; and as they were still considered by him as Upper 

 Oxfordian, it followed that the Coral Rag below should be called 

 Middle Oxfordian, and as such it appears in the monograph by 

 Leymerie and Raulin quoted above (25). Cotteau, however, snowed 

 by an exhaustive examination of the fossils (16), first, that the Coral 

 Rag was essentially "Corallian" in the usual sense, and, secondly, that 

 the Yermanton marly limestones, though from their similar lithology 

 they appear more closely allied to the Oxfordian than to the coral- 

 beds below, yet, on the whole, are really Corallian. It is noteworthy 

 that though these marly limestones are quoted in the earlier descrip- 

 tions as marls " a Amm. marantianus" (15), the only Ammonite given 



