322 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [May 19, 



If we next turn to the fifty-two Upper Limburg species (Table X.), 

 it will be seen that nineteen are identified with English Eocene fossils, 

 one only of the number {Venus incrassata) being admitted on the 

 disputed ground of its occurring in beds so modern as the Upper 

 Marine of Hampshire. In regard to two of the species also, we re- 

 quire fuller information before we can feel sure of their agreement. 



Out of 106 Middle Limburg species, forty-eight are British Eocene, 

 but thirteen of these must be deducted by those who do not consider 

 the upper marine and the freshwater beds of Hampshire as Eocene. 

 Even after this deduction, together with two others as doubtful, nearly 

 a third of the whole are common to undoubted Eocene beds. 



Lastly, out of 106 Lower Limburg fossils, forty-four are British 

 (or well-known French) Eocene shells, and only one of these {Venus 

 incrassata) would have to be excluded as occurring only in the Upper 

 Marine of Hants. If we omit five others as doubtful, for want of ample 

 means of comparison, there remain thirty-eight Eocene species, or 

 more than a third. This is as great a resemblance as can usually be 

 affirmed of any one of the great members of the English or French 

 Eocene divisions, when compared with another, — the Barton beds, 

 for example, with the Highgate, or the Calcaire grossier with the 

 Sables inferieurs Soissonnais. 



It is natural to find a larger proportion of shells in the Lower 

 Limburg, than in the Upper, common to the older Eocene. Among 

 the nineteen shells of the latter identified with Eocene species, eleven 

 occur in the Barton clay, or high up in the English series. When 

 in the * Principles of Geology ' I placed the Mayence basin in the 

 Miocene division, my classification was in that respect inconsistent 

 with itself, for M. de Koninck pointed out to me in 1850 his reasons 

 for concluding, many years before, that a great many of the Mayence 

 fossils agreed with the Limburg and Rupelmonde strata*. M. Bos- 

 quet has since observed to me, that it is with the Middle, and not with 

 the Lower Limburg division, that this analogy with the Mayence 

 basins holds good. 



In the paper by M. Hebert before referred tof , it will be seen that 

 he enumerates twenty-four species of shells from the '* Ostrea cya- 

 thula clay" and " Sables de Fontainebleau," &c., at Paris, Etampes, 

 and other places in France, a:s decidedly identical with Limburg and 

 Rupelmonde fossils. All of these are found in the Middle or Upper 

 Limburg beds, and only nine in the Lower Limburg. This greater 

 agreement with the former arises partly from the presence of fresh- 

 water species, none of which occur in the Lower Limburg. Of the 

 twenty-four French species, four are common to the Rupelmonde Clay 

 (or 4 in 43), and nine to the Lower Limburg beds (or 9 in 106) ; 

 so that the relationship of the French fossils, taken as a whole, with 

 the highest and lowest divisions of the Belgian formations, which are 

 both of them marine, do not seem to differ essentially, 



I have met with no Nummulites in any of the Belgian Upper 



* See * Manual of Geology,' 1851, p. 177 ; where I alluded to the agreement of 

 the Belgian and Mayence beds with those of Hermsdorf near Berlin, 

 t Bulletin de la Soc. Geol. de France, t. vi. p. 459. 1849. 



