138 P. W. HAKMER ON THE KESSINGLAND CLIFF-SECTION.
tion, the roots of which are still preserved on it, and that afterwards, at its extreme northern edge, a shallow depression was eroded by a stream, partly in the mottled clay, No. 2, and partly in the Chilles- ford Clay, No. 1, into which trunks and branches of trees growing at the surface of No. 2 were swept.
In the section, fig. 2, I have reduced the scale, so as to allow of its being prolonged to the northern extremity of Covehithe cliff, where the undoubted Chillesford Clay occurs capped by the Lower Glacial pebbly sand, in order that the position occupied by the mammaliferous clay and by the lenticular laminated deposit rela- tively to the Chillesford Clay both north and south of them may be more clearly seen. I see thus no reason to question that they were deposited iu a shallow depression or valley excavated by the removal of the latest of the Crag-beds, viz. the Chillesford Clay.
While, however, the posteriority of these mammaliferous and fresh- water deposits to the Crag seems thus apparent, their age relatively to the beds newer than the Crag is obscure, since there is nothing to show whether they preceded the Lower Glacial beds, which are absent from the Kessingland section, or succeeded them — a point of con- siderable interest, in reference to the great denudation which, as Mr. Wood and I maintain, followed the Lower Glacial formation when the valley-system of East Anglia was, we believe, mainly excavated. The Forest-bed which appears at intervals along the Cromer coast, on the other hand, affords clear evidence of its having preceded the entire Glacial formation, since the lowest of the Lower Glacial beds, the Till and the Pebbly sand (the latter of which is shown under 1 a as capping Covehithe Cliff in fig. 2), rest upon it — although, as has been pointed out by Mr. Wood and myself, its relation to the Crag is obscure, owing to the absence there of any deposit that we could recognize as belonging either to the Crag or Chillesford beds.
There is another matter connected with this subject to which I desire to allude, viz. the nature of the stone-bed which underlies the Crag at Thorpe, Bramerton, Horstead, and other places, which has been regarded, erroneously as Mr. Wood and I think, as an old land- surface. It is usually of inconsiderable thickness, consisting of large subangular flints imbedded in a matrix of ferruginous sand and gravel; and in it occur mammalian remains, such as the teeth of Mastodon arvernensis, a species which, according to Mr. Prestwich (Q. J. G. S. vol. xxvii. p. 118), occurs similarly in the bed of phos- phatic-nodule and other debris at the base of that much earlier for- mation the Coralline Crag. It does not present the least resemblance to such a formation as the Purbeck dirt-bed or to a freshwater deposit ; and it is, I believe, merely the marine basement-bed of the formation which overlies it, the mammalian remains found in it (which are gene- rally much worn and rolled and very imperfect) being, like those at the base of the Coralline and Ped Crags, as much derivative as are the flints with which they are associated. Pig. 3 represents the way in which, I believe, this bed may have originated, viz. by the advancing waters of an estuary which washed back its chalk shore and any Tertiary bed which may have rested on the Chalk. Had the
