least three successive beds, showing that, contrary to the Baron's idea, the mass is really stratified, though the uniform character of its constituents and their extremely mobile nature have for the most part effaced any decided marks of deposition.
To account for the origin of the formation, Baron von Richthofen has started the extremely ingenious explanation that the beds of the Chinese Loess have been formed on dry land, his principal reasons for this assumption being that the beds contain remains of land-shells and land-animals to the exclusion of marine or even, so far as known, freshwater species, and that no depression of the eastern portion of the continent is sufficiently recent to allow of their deposition under the surface of the sea.
I shall deal with these objections in reverse order, and afterwards state some reasons against the subaerial theory.
Evidence of late depression in North China. — First, the Baron states that there is evidence to prove that the north of China has not been submerged to the depth of 6000 feet within a recent geological epoch. Without arguing as to the difficulty of proving a negative of this sort, I shall only state that to my mind there is abundant evidence, irrespective of the Loess itself, to prove that China, as far north as the Yellow River (beyond which my personal experience does not extend), has since the commencement of the Tertiary period been the scene of very considerable depression. Proofs of this, I believe, are to be found in the upper Nanking sandstones and conglomerates and their succeeding rocks. These sandstones, in almost perfectly horizontal strata, stretch from the south of Nanking through northern Anhwei as far at least as Ting-yuen-hien in the Fung-yang prefecture, being especially characteristic at Luchow-fu, in the centre of this district. The upper portion of these rocks I believe to represent the Tatung gravels of Baron von Richthofen. These Tatung gravels extend through the south-western portion of Anhwei, forming in many localities the bed of the present valley of the Yangtsze, are seen in still greater development in Hupeh, as at Hwang- chow and Wuchang-hien, and probably reach as far west as Ichang, at the foot of the gorges of the upper Yangtsze. I have met with them myself at San-kia-tientsze, some thirty miles from Bung-yang- fu, in the north-east of Anhwei province, where they form a bold escarpment looking over what was at one time the Yellow Sea, but now constitutes the alluvial plain of Kiang-peh. I do not know how much further in either direction these beds extend ; and, besides, I wish to confine myself to facts within my own observation. If local conditions at Kiukiang on the Yangtsze are to be trusted, these gravels pass upwards into the Kiukiang laterite, a deposit occupying likewise a considerable area in Anhwei, Kiangsi, and Hupeh.
It seems improbable that these rocks, which extend over so large a space, are otherwise than marine, though at present no fossils have been collected in them. The bold escarpments of the hills on either side of the Yangtsze occur in localities where it seems impossible to ascribe them to fluvial action. They are much more suggestive of ancient coast-lines. They are, besides, not confined to the valley of
