112 H. KEEPING AND E. B. TAWXEY OX THE BEDS AT 



fifth, which was the proportion given in the paper referred to (op. 

 cit. p. 150). If we compare with this the ratio of Barton forms in 

 the list of commoner Headon-Hill given above (supra, p. 105), we find 

 that 30 per cent, pass up from Barton beds (while in the complete list 

 of the Headon-Hill fauna the porportion is also about 29 per cent.). 

 Moreover, if we first eliminate the forms that occur also at Brocken- 

 hurst, so as to obtain what we may consider as specially Headon 

 species, the proportion becomes even lower. On all hands the 

 palaaontological evidence seems perfectly in accord with the strati- 

 graphical. 



Relation of Colwell Marine to Brockenhurst Fauna. — In order to 

 see whether the Colwell bed is more nearly related to the Brocken- 

 hurst than is the Headon-Hill bed, we will first take the proportion 

 of Barton forms in it for comparison with similar treatment of the 

 Headon-Hill catalogue. Examination of the list gives us a pro- 

 portion of 29 per cent, of Barton forms in the Colwell-Bay bed ; we 

 saw above that, in the Brockenhurst bed, the ratio was about 50 per 

 cent, and in the Headon marine bed 29 per cent. An inspection of 

 the list of more characteristic Colwell and Headon marine fossils 

 (supra, p. 105) showed that these faunas are practically identical 

 Xow we see that their proportion of Barton forms is nearly equal, 

 and far lower than in the Brockenhurst bed. 



To complete the proof from fossils, if any such is needed, we may 

 inquire whether there are more Brockenhurst forms peculiar to 

 Colwell Bay than to Headon Hill. Examination of the lists 

 shows that only the following Brockenhurst species occur at 

 Colwell Bay and not at Headon Hill, viz. Scalaria tessellata 

 and Tellina affinis, the latter passing up from Barton beds ; 

 while those occurring at Headon Hill and not at Colwell Bay are 

 Marginella cestucvrina and Cardita paucicostata — two only in each 

 case, which amounts to perfect equality. If, on the otber hand, we 

 count those common to the Colwell and Headon marine beds, and not 

 occurring at Brockenhurst, we find twenty-six species in this 

 category. ^Ve are at a loss to understand how any one could 

 imagine that the Brockenhurst fauna is identical with that of the 

 Colwell-Bay bed and newer than that of Headon Hill. 



That the Colwell-Bay bed is stratigraphically identical with the 

 Headon Middle Marine we hope that we have sufficiently proved ; 

 and the fact is confirmed by fossil evidence. The same twofold 

 proof has been brought forward to demonstrate that the Brocken- 

 hurst bed, where present, lies at the base of the marine Headon 

 beds and immediately above the Lower Headon. This bed is 

 absent at Colwell Bay and Headon Hill, but is seen at Whitecliff 

 Bay, Brockenhurst, and Lyndhurst, 



Since the Middle Headon includes every thing between the fresh- 

 water Upper and Lower Headon, it must be allowed to include the 

 Brockenhurst beds, though that special fauna was not known when 

 names were given to these groups of strata. It would cause the 

 greatest inconvenience to abandon the term Middle Headon, as it 



