We will not do more than mention the latter, which is of course derived 

 from the former : it resembles in its characteristics, such as vertical cleavage 

 and compactness, the true Loess. 



The true Loess, which forms the bulk of the Huang-Vu formation 

 encountered by us on our journey, is again divisible into two classes, con- 

 taining a greater and lesser proportion of separated clay respectively. To the 

 former I give the name of Shao-fu, which has already been explained in the 

 earlier part of this Chapter ; for the latter we may still use the name Loess. 



The extent of the Huang-fu formation is very great : it covers, roughly 

 speaking, an area of 200,000 square miles. Occurring as a sort of mantle 

 overlying the greater part of Shansi, Shensi, and Kansu, it extends into Honan 

 and forms also the great plains of Chihli. But nowhere are such vast deposits 

 found as in the Province of Shensi, north of the Wei Ho. 



The physiography of North Shensi has already been likened to a great 

 basin, and it is in this basin that we can see loess at its best. It covers the 

 whole of the sedimentary rocks to an average depth of 1000 feet. In some 

 places — notably along the sides of the Yellow River — it gets rather thin, but 

 in the country immediately south of the Ordos Desert the depth increases to 

 about 2,000 feet. The greater part of these Loess deposits are cut up by the 

 action of water into rounded hills of uniform height. 



To the east, south and west of Fu Chou, in North-Central Shensi, the 

 Huang't'u formation assumes the form of immense plateaux, divided from one 

 another by deep valleys, and cut up along their edges by deep and narrow ravines 

 (Plate 56). The surfaces of these plateaux are perfectly flat ; and all are equal 

 in height to one another, and to the loess hills extending over the rest of the 

 basin. The hills immediately adjoining the plateau-area are the more flat- 

 topped, showing a gradual transition from the plateau to the hill form. From 

 these facts it may be argued that the whole of the North Shensi basin was at 

 one time in the form of a great loess plain, formed during a long period of 

 drought, when little or no erosion was taking place. I cannot conceive of such 

 a plain having been formed under the weather-conditions prevailing at the 

 present time. It is certain that deposition is going on nowadays, but I do not 

 think its rate can be in any way equal to that of erosion. Many facts may be 

 cited to show the enormous rate at which the present torrential rains, 

 occurring in summer, erode the soft Loess. Some twenty miles north-east of 

 Yti-lin Fu, where deep deposits of loess occur along the Ordos border, huge 

 ravines or canons, many miles in length and hundreds of feet deep, cut through 

 the Great Wall into the Ordos. Everything points to the fact that these have 



128 



