The term _loess_, now generally accepted, has been used to designate a tertiary deposit appearing in the Rhine valley and several isolated sections of Europe; its formation has heretofore been ascribed to glaciers, but its enormous extent and thickness in China demand some other origin. The substance is a brownish colored earth, extremely porous, and when dry easily powdered between the fingers, when it becomes an impalpable dust that may be rubbed into the pores of the skin. Its particles are somewhat angular in shape, the lumps varying from the size of a peanut to a foot in length, whose appearance warrants the peculiarly appropriate Chinese name meaning ‘ginger stones.’ After washing, the stuff is readily disintegrated, and spread far and wide by rivers during their freshets; Kingsmill[168] states that a number of specimens which crumbled in the moist air of a Shanghai summer, rearranged themselves afterward in the bottom of a drawer in which they had been placed. Every atom of loess is perforated by small tubes, usually very minute, circulating after the manner of root-fibres, and lined with a thin coating of carbonate of lime. The direction of these little canals being always from above downward, cleavage in the loess mass, irrespective of its size, is invariably vertical, while from the same cause surface water never collects in the form of rain puddles or lakes, but sinks at once to the local water level. One of the most striking, as well as important phenomena of this formation is the perpendicular splitting of its mass into sudden and multitudinous clefts that cut up the country in every direction, and render observation, as well as travel, often exceedingly difficult. The cliffs, caused by erosion, vary from cracks measured by inches to cañons half a mile wide and hundreds of feet deep; they branch out in every direction, ramifying through the country after the manner of tree-roots in the soil--from each root a rootlet, and from these other small fibres--until the system of passages develops into a labyrinth of far-reaching and intermingling lanes. Were the loess throughout of the uniform structure seen in single clefts, such a region would indeed be absolutely impassable, the vertical banks becoming precipices of often more than a thousand feet. The fact, however, that loess exhibits all over a terrace formation, renders its surface not only habitable, but highly convenient for agricultural purposes; it has given rise, moreover, to the theory advanced by Kingsmill and some others, of its stratification, and from this a proof of its origin as a marine deposit. Richthofen argues that these apparent layers of loess are due to external conditions, as of rocks and débris sliding from surrounding hillsides upon the loess as it sifted into the basin or valley, thus interrupting the homogeneity of the gradually rising deposit. In the sides of gorges near the mountains are seen layers of