432 



NA TURE 



[August 31, 1882 



would cause great inconvenienee. The term Middle Headon, 

 based as it is on the cla-sical work of Edward Forbes, is dem- 

 and definite. Again, it would entail the abandonment of the 

 names Upper and Lower Headon also ; and the non-occurrence 

 of the Brockenhurst series, or its representative, in Colwell Bay 

 admits of no recognition on the west side of the island, and 

 therefore the cla-sification would be based upon a defective 

 appreciation of the beds. 



Von Kbnen, in 1S64, justly correlates the fauna, and since 

 then, in 1866, the coral fauna has been de cribed by Dr. 

 Duncan. 



Messrs. Tawney and Keeping, in their paper on the beds at 

 Headon Hill and Colwell Bay in the Isle of Wight, uphold the 

 work done by the Geologinal Survey, maintaining the correctness 

 and integrity of the two Survey Memoirs, and the horizontal 

 and vertical sections of the Tertiary beds of the Isle of Wight. 

 Prof. Judd differs from the identifications and stated succession 

 of the beds in Totland and Colwell Bays. He introduces two 

 new series at Headon Hill, a marine and a freshwater (?) in addi- 

 tion to those which have been universally accepted for the last 

 twenly-five years {Quarterly journal of the Geological Society, 

 vol. xxxvi.). 



The sections prepared by Prof. Judd also differ very consider- 

 ably from that of the Geological Survey, or those lately pre- 

 pared by Messrs. Keeping and Tawney, during their late exami- 

 nation of the beds under notice. These are the marine series 

 known as the Middle Headon or Middle Marine. Prof. Judd 

 places them at the sea level near Widdick Chine. Conse- 

 quently, between the top of the marine be.l and that of the Bern- 

 bridge limestone, there would be, on Prof. Judd's theory, 250 

 feet of beds, such being the altitude of the cottage on the 

 Warren which marks the summit of the Bembridge limestone. 

 This thickness mu-t, however, be reduced by ico or 105 feet, 

 which is the altitude of the top of the Middle Headon at this 

 point. This 105 feet of beds, or another frehwater and another 

 marine have no existence; they can only be accounted for by 

 counting the Lower and Middle Headon twice over. Now the 

 only marine beds are those of the Middle Headon, inclosed 

 between the altitudes of 70 feet above the sea level ; the others 

 are all fresh water. 



The point wherein Prof. Judd's section differs from the 

 Survey, and that of the authors, arises from the belief that a 

 second marine series, termed the "Brockenhurst series," with 

 another freshwater below, in all 105 feet, is intercalated above 

 the Upper Headon — these two believed new formations having 

 that portion of the section allotted to them which is occupied by 

 the freshwater Osborne marls and part of the Upper Headon. 

 It must lie remembered that there is no positive evidence of the 

 existence of this second marine (Brockenhurst) ser.es at the spot 

 where the Geological Survey place the Osborne tnarls. Careful 

 examination fails to reveal these said to be additional beds. It 

 is clear, therefore, that no bed having the | eculiar fauna of the 

 Brockenhurst bed occurs at the west end of the i-la d ; its place 

 100, if found, would be at the base of the Middle Headon, and 

 not above the (-'/•per, where it has been wrongly assigned. 

 Messrs. Keeping and Tawney, in their paper, object to the 

 correlation ol the Brockenhurst with the Colwell Bay bed — 

 which is identical with the marine (Middle Headon) bed of 

 Headon Hill. Thus the 105 feel of itrata have no existence. 



The Middle Headon, which is denuded away from the top of 

 the cliffs in the centre of Totland Bay between Western and 

 Clime-, has been discovered in the Totland Bay brick- 

 yard, which lies a little inland id this portion of the cliff, thus 

 ly showing that this bed was continuous above the lop 

 of the cliff, consequently linking the Warden Cliff expo ure to 

 that of Headon Hill. They are visibly and ab-olutely con- 

 tinui Us with tnose of Colwell Pay. 



Palaontological Evidence. — The equivalency of the Colwell 

 Bay and Brockenhurst beds is a point to be definitely settled. 

 Mi. st careful li^ts of fi ssils have been prepared from collec ions 

 mule both from the Middle Headon at Colwell Hay, aid 

 Headon. We find that out of fifty-seven species at Colwell 

 Hay, fifty-three occur in the Middle Marine of Headon Hill, or 

 93 per cent. This clearly proves the identity of the horizon in 

 the two localities. 



The well-known shells Cerithu/u coneavum and C. ventri- 

 cositm occur both in Colwell Bay and Headon Hill, and on the 

 same hcrizons. C. coneavum appears to have a less restricted 

 range at Headon Hill than C. ventricosum, occurring abundantly 

 there through the greater part of the Middle Headon series. It 



has also been found in the "Venus bed" of Colwell Bay. 

 Thus both stratigraphies! and palaxmtological evidence are Ui 

 harmony. All evidence tends clearly and conclusively to show 

 that there is only one marine seiies in this section, viz. the 

 Middle Headon of Edward Forbes,, which is interstratifieel 

 between the freshwater Lower and freshwater Upper Headon ; 

 while there is no evidence of the Brockenhurst bed occurring 

 anywhere in the west of the island. 



rVhitecliff Bay and the New Forest. — The Brockenhurst bed 

 was recognised at Whitecliff Bay by the Rev. O. Fisher in 1864, 

 where it occurs in the lowest 2 feet of the Middle Headon series. 

 No less than 70 species have been collected here out of 104 

 known at Brockenhurst. Many species are peculiar to it, but all 

 are identical with those of the w ell-known section in the railway 

 cutting near Brockenhurst. Many species are confined to this 

 horizon and do not pass up into the "Venus bed." Thus the 

 Brockenhurst fauna at Whitecliff Bay number 70 species, at 

 Brockenhurst 104, and of these only 18 occur in the Middle 

 Marine beds of Colwell Bay, or are common to Whitecliff Bay and 

 type locality. Eighty-three Barton or Brack lesham species pass 

 up, 25 to the Middle Marine of Colwell Bay, and 36 to the 

 Brockenhurst bed of Whitecliff Bay, or these two localities 

 yield the above number of Bartonian forms. To still further 

 illustrate the value of the Middle Headon series of the Isle of 

 Wight and elsewhere, I may mention certain characteristic fossils 

 that occur in several zones. The "I onus bed" of the Geo- 

 logical Survey is about 30 feet thick at Colwell Bay, Headon 

 Hill, and Whitecliff Bay, and contains the following well- 

 marked shells, Murex sexdentatus, Milania fasciata, Cerithium 

 duplex, C. ventricosum, C. coneavum, and Nerita aperta. 

 Shells characteristic of the Brockenhurst bed and confined to it 

 are Valuta sufura/is, Lciostoma ovatiim, Peeten bellicostatus, 

 Modio/a nysti, Cyprina nysti, and Cytherca solandri, var. 

 atttnuata. In the Koydon zone occurs Valuta geminata, and 

 now here else in England. PUurotoma transversa/ ia, /'. subdenli- 

 culata, Cardita deltoidea, and Protocardmm hanUniense are in 

 both the Roydon and brockenhurst zones, but not known in the 

 Venus bed. Certain species range through the Middle Headon 

 series and occur nowhere else. These are Pisania labiata, 

 PUurotoma headonensis, Cancel/aria muricata, C. elongata, Leda 

 prop nqua, Cytherea suborbicularis, Psammcbia •estuarina, and 

 Corbicula obovata. The Brockenhurst zone is restricted to the 

 lower 2 feet of the Middle Headon, and it lies immediately on 

 the eroded surface of the Lower Headon. An error certainly 

 has been committed in the New Forest Section, in assigning the 

 place of the Brockenhurst series above the Middle Marine or 

 Mieldle Headon. This is at variance with facts at Brockenhurst 

 and Whitecliff Bay, and this misapprehension as to the strati- 

 graphical position of the Brockenhurst bed refutes the theory as 

 to the occurrence of this bed high up in the Headon Hill. It is 

 not in existence there. 



Wiih reference to the affinities of the Brockenhurst fauna it 

 has been stated that "nearly one-third of the Hordwell and 

 Headwell Hill marine shells are Barton fo-ms, and not more 

 than one-fifth of those occurring at Brockenhurst, Colwell Bay, 

 and Whitecliff Bay, are found at Barton." We should not 

 expect the Venus Bed or Middle Marine would have more 

 Barton species than the Brockenhurst Bed, seeing that the 

 former occupies a higher zone in the Middle Headon series. 

 The percentage of Barton forms, according to Mr. Tawney, in 

 the Whitley Kidge bed, i- 42 per cent. ; a lower proportion than 

 at Whitecliff Bay, arising from the number of corals being special 

 to the Whitley locality. At Whitecliff Bay the Barton group 

 has 52 per cent., and the proportion of Barton forms from all the 

 Brockenhurst localities, including the Roydon zone, is 48 per 

 cent. , and the percentage of the Barton forms in the Middle 

 Headon of Headon Hill is found to be 29 percent. ; the conclu- 

 sion, therefore, from fossil evidence is that the Headon Hill 

 marine bed is later in age, and higher stratigraphically than the 

 Brockenhur.-t bed, the proportion of Barton forms in the latter 

 being nearly 50 per cent., and not one-fifth, as stated. The 

 result is in strict accordance with their stratigrapbical positions. 

 It is equally important to test by fossil evidence whether the 

 Colwell Bay Venus bed (Middle Headon) is more nearly related 

 to the Brockenhurst than is the Headon Hill bed. In Colwell 

 Bay the observed Barton forms are 29 per cent, in common, and 

 the same percentage in the Headon Hill bed, while in the 

 Brockenhurst bed they were 48 per cent. To test still more the 

 proof from palteontological evidence, it is stated, on the same 

 authorities, that there are only two sfecies in each case common 



