1881.] Hordivell Cliffs, Hampshire. 145 



" These Headon Hill sands are usually called the Upper Bagshot 

 " beds ; but it appears to me that it cannot be but a source of con- 

 " fusion to base our classification of the Upper Eocene strata on the 

 "poorly fossiliferous deposits of the London basin rather 1 than on 

 " the richly fossiliferous deposits of the Hampshire basin. It is 

 " true that at Alum Bay the Headon Hill sands have not yielded 

 " any fossils, but the equivalent beds at Hordwell contain a by no 

 " means scanty fauna, in which we find the same admixture of 

 " marine and freshwater forms which characterises the overlying 

 " Headon strata. As moreover we detect in these beds the emi- 

 " nently characteristic Cerithium concavum, it seems clear that we 

 " must regard them as constituting the lowest member of the 

 " Headon group." Accordingly in the table comparing the English 

 beds with foreign equivalents [I. c. p. 167] the whole of these Upper 

 Bagshot sands, Lower and M. Headon beds are paralleled with the 

 Mortefontaine sands, which are placed above the St Ouen freshwater 

 limestone. [This error is avoided by the omission of the St Ouen 

 limestone in a similar table, Pop. Sc. Rev. iv. New Ser. 1880, p. 133.] 



Since Oliva Branderi, Sow., occurs abundantly in these sands, 

 a shell characteristic of the Barton beds and not passing up into 

 the M. Headon marine beds, it has seemed to us that the view 

 just cited was not to be accepted without further evidence. 



Accordingly we made a journey to France to examine the 

 Mortefontaine bed, collect its fossils for comparison with those 

 from the Long Mead End sands, and trace its connection with the 

 St Ouen limestone, and beds above, as well as with the " Sables de 

 Beauchanrp" below, — which latter is equivalent to our Barton 

 deposits. 



If Searles Wood had been able to determine the species in 

 the numerous genera which he collected from this bed, doubtless 

 the true affinities of the bed would have been recognised, and 

 Forbes' opinion would scarcely have been challenged. The result 

 of our collecting from the Hordwell sands is given in a table, in 

 which we have also noted in separate columns such of the species 

 as pass down into Barton beds, or upwards into Middle Headon 

 marine deposits, as well as those recognised in the Mortefontaine 

 sands of the Paris basin — here deemed equivalent. 



We have compared our specimens with those in the Edwards 

 collection in the British Museum, which were determined by Ed- 

 wards himself; in cases where we have not accepted his names 

 we have in an appendix given reasons for so doing. 



The result of our work is that out of a total of 28 species 

 which we collected from the Long Mead End sands, 15 occur in 

 the Barton beds below, while 10 pass up into the Middle Headon 1 . 



1 This is including the Brockenhurst bed at the base of the M. Headon, where 

 many Barton species cease without extending up into the Venus-bed. 



