ON THE HIGHER EOCENE BEDS OF THE ISLE OF WIGHT. 417 



passed away. Opposite the very apex of the hill a huge mud stream 

 descends from a sort of "irque, like a glacier, to the sea ; a little west a 

 second stream also finds its way to the Solent. The rest of the escarp- 

 ment, save where local slips have occurred, is overgrown with a dense 

 tangle of vegetation. The mud streams are fed by slips from the terraces 

 overhanging them, and in the terraces the upper Hamstead beds are 

 exposed and can be worked. The lower Hamstead beds can be got at 

 in the low cliffs met with here and there along the shore, while the Bem- 

 bridge marls are best seen in the rather extensive flats exposed between 

 tides. In proceeding eastward from the mud streams the dip brings lower 

 beds to the surface in succession. 



The beds have been described in the Survey Memoirs by Forbes and 

 Bristow, but the death of the former whilst the work was in progress 

 undoubtedly led to its being published in the imperfect state in which it 

 still remains. The marine beds at the summit are, according to these 

 authors, twenty feet thick. The next thirty feet beneath are more or 

 less brackish, and easily distinguished by the presence of Cerithiuin. The 

 thickness of the remainder of the beds, down to a shelly band called the 

 ' white band,' forming the ' middle estuarine and freshwater,' is stated 

 to be fifty feet, but it <inust evidently exceed 100 feet, for the ' white 

 band ' is within about forty feet of the sea level, and the hill, as we have 

 seen, is 210 feet high. About thirty feet down in them, and forming the 

 ledge of the second terrace, is a bed of compact and distinctly laminated 

 clay which may be distinguished as the ' Leaf bed ' of the Hamstead 

 series. It is full of a peculiar creeping root, from one half to nearly an inch 

 in diameter, and jointed at intei'vals of three feet or more by rounded 

 nodes an inch across, from which radiate closely set straight filamentous 

 rootlets, two or three inches in length, and clothed with fibrils. The 

 scars left by the i-ootlets are small and mammillated, so that when by 

 chance a node has been severed and deprived of rootlets, it has all the 

 appearance of an echinated fruit. The roots are jet black, without netted 

 markings or veins, and contrast strongly in colour with the whitish 

 macerated fragments of sword-shaped leaves which accompany them. 

 This contrast in colour and preservation appears due to the fact that the 

 former were buried in mud even whilst living and never exposed on the 

 surface, whUst the latter had reached the last stage of decay before being 

 silted over. No leaves whatever are found in this bed except Nelumhium 

 and even these are of such extreme rarity that, during about a week's 

 search, only three undeveloped leaves and a part of a developed one were 

 met with. The figure exhibited (plate IV.) is from a nearly perfect speci- 

 men in the British Museum. The leaves, so far as outline and venation 

 go, appear to belong unquestionably to Nelumhium, and the root-stocks 

 are, according to Heer, of the same plant. They seem, however, from 

 their present appearance, to have been quita hollow, like cane roots 

 whilst those of Nelumhium, are so fleshy and succulent that they are cut 

 up into lengths and largely eaten as a vegetable. Nelumhium is at pre- 

 sent an inhabitant of Asia, the Philippines, and Australia, and formerly 

 grew in the Nile. It has been found fossil in some of the Tertiariea 

 of Central and Southern Europe, and I believe the same species occurs in 

 Antrim. 



We are also fortunate in discovering the rare fruit shown in fig. 11 

 (plate III.). It was apparently a rounded or subangular capsule with thick 

 walls and two or more chambers. The seeds are numerous and compressed 



1887. E J. 



