72 



SCIENCE. 



[N. S. Vol. XII. No. 289. 



from Pekin into Mongolia by way of Kalgan, I 

 observed the following facts : 



(1) The loess is a wind deposit without doubt. 

 Along the Tsing-ho, a river joining the Yang- 

 ho near Kalgan, I found that all the north and 

 south tributary valleys had slight deposits of 

 loess in sheltered spots along both sides, and on 

 the south or southeast slopes of the mountains. 

 In the east and west valleys the north side of 

 the valleys, that is the south slope of the 

 mountains exhibited loess hundreds of feet 

 steep, and clinging in sheltered spots to the 

 very summit of the mountains more than 5000 

 feet above tide. 



On the other side of these east and west 

 valleys the loess deposits are practically want- 

 ing, except in gullies where there would be a 

 lull in the wind. 



The Chinese, who have overrun the Mon- 

 golian border, make use of this firm perpen- 

 dicular cleaving loess for excavating houses 

 which stand well. So the towns are usually 

 found on the south or southeast slope of the 

 mountains, where they have the loess to build 

 in, or to build with, and also the sunny south 

 exposure. 



As a rule, depending on the local physical 

 structure of the country, these deposits are 

 rather more on the southeast than south side. 

 In other words, the prevailing winds, then as 

 now, blew from the northwest, down over the 

 plains of Mongolia, the escarpment of which 

 runs from northeast to southwest. 



(2) In the valleys it often shows modification 

 by water action. In the valleys and even half 

 way up the mountains bands of rock fragments 

 usually very angular are of common occurrence. 

 These are of local origin and in all cases could 

 be easily accounted for. They were either 

 talus accumulations from the hill back of them, 

 or else were deposited by some temporary 

 stream which was formed by one of the sudden 

 and terrific rains to which this section is sub- 

 ject during the summer months. 



In oue of the pits northwest of Kalgan there 

 is a U-shaped deposit four feet across, of well- 

 rounded gravel, some of the pebbles being 

 three inches in diameter. It looks as if a 

 stream of considerable size and superloaded 

 with gravel from the hills near by had run 



over the loess at this point for a short time 

 during the latter's period of deposition. 



Lower down in the valley of the Yang-ho, 

 100 or 200 feet above the present river, espe- 

 cially where side streams have built up deltas 

 at the point of emergence from the mountain 

 passes into the valley, beds of sand, gravel and 

 loess are interstratified. Probably this loess is 

 material brought down either by the main river 

 when it was at a higher level or by the side 

 stream and deposited in slack water. 



(3) There was some special period of rapid 

 deposition, and that in quite recent time. Now 

 this loess is everywhere deeply channeled by 

 the little streams that are cutting it away. A 

 very characteristic channel is one 20 to 30 feet, 

 deep, 3 feet wide at the base, and from two to 

 three times as wide at the top. Such miniature 

 canyons will often be cut back a few hundred 

 yards from the valley. Evidently this loess was 

 deposited very rapidly at one time and then for 

 some reason, probably lack of material, ceased 

 to accumulate. 



At present there is enough wind to do the 

 work if it had the material at hand. Having 

 been for seven hours in a dust and sand storm, 

 between Hsiian-Hua-Fu and Kalgan, I feel cer- 

 tain that the present wind forces are suflBcient 

 to deposit loess much more rapidly than it 

 would erode away, provided it had the mate- 

 rial. As it is the wind deposits now forming are 

 entirely different from the loess. The drifts are 

 in the same sort of places, but instead of being 

 an impalpable dust are sand. At Hsiian-Hua- 

 Fu the city wall is banked to the very top with 

 drifts of sand, but no loess. 



At some recent time the winds must have 

 had an excessive amount of this peculiar fine 

 dust at its command, and the dust must have 

 come from the plains of Mongolia. Whether 

 this material was supplied by glacial grist, 

 furnished by glaciers coming down on to the 

 Mongolian plains from the elevated mountain 

 region to the northeast, or not, remains to be 

 seen. One thing is certain. The glaciers never 

 extended down to the edge of the Mongolian 

 plateau in this region (Lat. 40° North). 



(4) This deposit is very recent, for many of 

 the smaller streams have not yet cut their way 

 through it to the rock. This is in marked 



