sphere. The layer formed by one generation of plants would, in effect, have been absorbed by the next without any addition being possible.
Dust-storms, however, have been suggested as a source of supply. Unless these passed over deposits of the Loess itself, I know of no other source for the necessary ingredients. Clay would not be acted on by the wind ; sand, of itself, would not suffice to form the peculiar mixture of ingredients ; limestone rocks have never, to my knowledge, been so disintegrated by the action of the atmosphere as to become reduced to dust capable of being transported in the manner suggested. The means are therefore utterly inadequate to the end.
Easy removal of the Loess by rain. — There are, however, other grounds of objection. Had the Loess been formed as dry land, there is no reason why it should not only have utterly ceased to increase, but should be actually undergoing a rapid destruction. There is evidence sufficient to prove that its waste now is greater than at former periods, owing to the ignorant destruction of the trees with which, tradition states, it was once covered. This destruction has increased the frequency and force of the annual floods ; but as long as rain fell or rivers ran, denudation of one sort or other must have been going on. Denudation above the sea-level is, in fact, as necessary an accompaniment of running water as is deposition below it. When it is remembered how sensitive is the Loess to the slightest contact with water, and that ever since the emergence of the Loess district from the waves every stream within its limits must have been continually engaged in the work of denudation, the difficulty of accepting the subaerial theory becomes a practical impossibility.
Marine origin of the Loess. — Rejecting, then, as untenable the theories which would assign a glacial, freshwater, or subaerial origin to this peculiar formation, little remains except to class it as marine. As yet, except in its peculiar structure, such as its mechanical resemblances to chalk, an undoubted deep-sea formation, no internal evidence has been discovered to guide us to this conclusion. Specimens of the clay were sent by Mr. Pumpelli to the United States, and examined microscopically by Mr. Edwards ; except, however, some small green crystals, pronounced not to be organic, nothing peculiar presented itself. There are here no means of making microscopic examinations, and but few for scientific operations of any kind ; inductions have therefore to be founded on what, under more favourable circumstances, would be deemed insufficient proof, trusting to subsequent rigid investigation to prove or disprove their truth. The real origin of the Loess will probably be proved by the close examination of its microscopic structure ; but even this must be to a certain extent taken in connexion with its external conditions, as Microzoa may readily be of derivative origin.
Probable Geological relationships of the Loess. — The assumption of a marine origin for the Loess of eastern Asia leads up to most important geological deductions. It extends, as has been stated above, from the south of the Yangtsze, in the prefecture of Chinkiang, far into
