J. tS. Lee — Recent Ice-action in N. Cltina. 19 



connected witli the so-called pebble-bed of northern Honan, then the 

 age-problem can be at once solved ; for we know definitely that the 

 pebble-bed is everywhere else immediately overlain by the loess. 

 3Ieanwhile, let us turn to the fossil evidence. Cases are known that 

 the so-called Pterophijlluni carhonarium ScLenk associates with 

 forms closely related to Nilssonia. This would mean that the upper 

 jjart of the coal-bearing series either belongs to the Permo-Mesozoic 

 or even entirely to post-Palaeozoic. In any case the possibility of 

 the boulder-bed belonging to the Pemio-Carboniferous can be 

 absolutely excluded. It might be, however, urged that the ice- 

 marked rock-fragments could have been originally produced in some 

 period earlier than the late Tertiary or early Pleistocene, and became 

 re-arranged before the dawn of loess time. But if this be the case, 

 how could the gigantic blocks of rocks, entirely unknown in the 

 neighbourhood, be brought to their present resting place ? How can 

 the unstratified state of the deposit be explained if the material was 

 wholly transported by water ? Why are some of the fine striag and 

 most of the sharp edges of the boulders or fragments remarkably 

 well preserved ? It would be extremely difficult, if not impossible, 

 to find a satisfactory explanation save on the basis of recent 

 glaciation. 



(2) During a reconnaisance in the Tatung Basin, Xorthern Shan-si, 

 T was struck above the village of Kou-tsuen (about latitude 40^ 5' N., 

 longitude 113^ 15' E.) by a U-shaped valley which extends a little 

 less than 10 miles in a west-east direction, and shows a wonderfully 

 uniform width throughout its entire length. In the valley there 

 occur strange boulders consisting of gneiss, schists, basalt, and many 

 otlver kinds of igneous rocks. They are not derived from the hUls 

 on both sides of the valley, nor from the watershed that separates 

 this valley from the broad expanse of lowland which forms the 

 western part of the basin ; for all these hills are composed of piled-up 

 horizontal or nearly horizontal strata of Jurassic sandstones, which 

 on the top of the watershed are covered by the loess and near the 

 entrance of the valley are replaced by the Palfpozoic coal-bearing 

 series aiid the Sinian Limestone. Thus it is beyond doubt that the 

 boulders are not of local origin. As there appears no sign that the 

 present valley had once extended across the watershed, the boulders 

 could not have been transported by a pre-existing river whose 

 source niight have extended far back into the iiiountains on the west 

 of the Tatung Basin. 



The side of the valley is sometimes so steep, particularly in its 

 lower part, that it almost merges into a clifT, assuming the shape of 

 a snubbed hill-spur. Near the base of one of such clifis a block 

 of dislodged sandstone was found, which exhibits coarse parallel 

 furrows on one of its surfaces. The bottom of the valley is filled up 

 by stratified but hardly consolidated rock-debris, attaining a thick- 

 ness of more than 20 feet in many places where the river erosion 

 has gone so far as to pennit measurement. Above this deposit 



