Vol. 51.] THE LOESS AND OTHER DEPOSITS OP SHANTUNG. 251 



The stones, where we have examined them, are rolled and water- 

 worn, very different from the torrent-debris of the present streams ; 

 and they are, as might be expected, chiefly of local origin : schists, 

 gneiss, and quartzites in the Archaean region, sandstones and lime- 

 stone in the Carboniferous area. 



In Europe, especially in England and France, what are usually 

 classed as loess-beds often lie upon true river-gravels, as is the case 

 with the Thames and the Somme. We are inclined to think that 

 much of this material is not true undisturbed loess (as is shown by 

 its sweeping down the valley-sides and not lying like a flat cake), 

 but that it is either of different origin, or re-arranged as is so much 

 of the loess of China. Nevertheless, if the loess has been deposited 

 tranquilly in a quiet sea not very unlike the present Yellow Sea and 

 Gulf of Pe-chi-li, during a gradual submergence, the river-gravels of 

 the sinking land would be covered up, as doubtless some of the old 

 valley-gravels of Shantung are by the Marine Sand. 



In Europe traces of man's handiwork are found associated with 

 the loess ; in China up to the present this is not so, and our searches 

 were fruitless, but as yet hardly a single observer has dealt with 

 this question in China. A few prehistoric stone-implements are 

 known, but all are of far more recent date than the loess. 1 



We look upon the Basement-gravels as being to the loess sea very 

 much what the Upper Greensand was to the Chalk sea — a series of 

 beaches extending their limits as the sea-waters crept onward during 

 slow subsidence. 



III. The Tertiary History of North China. 



Comparatively little is known of the Tertiary formations of North 

 China, but in the provinces of Chekiang and Kiangsu the following 

 may be noted. 



Beginning at Ning-kong-jow (lat. N. 29° 45', long. E. 121° 30'), 

 some 12 miles from Ningpo, is a very marked series of flagstones. 

 Commencing in rough shingles, almost boulder-beds, some of the 

 fragments having a cubical content of 6 to 8 feet, they pass 

 upwards into fine siliceous flagstones. These, not improbably repre- 

 sentatives of the Siwalik Sandstones of India, though as yet they 

 have apparently yielded no fossils, are probably slightly over 1000 

 feet in thickness. They are associated with or underlie an equally 

 thick mass of volcanic tuffs, locally called the Daying Beds. 

 Succeeding these, apparently, are a series of red sandstones, 

 characteristic about Hang-chow (lat. N. 30° 15', long. E. 120° 

 30'), which may conveniently be called ' the Hang-chow Sandstones.' 

 These are largely represented in the valley of the Yangtse, appear- 

 ing about Nanking and still higher up in Kiangsi and Hupeh, and 

 at Hankow (lat. N. 30° 35', long. E. 114° 25') are largely used 

 for building purposes. 



Following these again is a thick series of siliceous gravels, 



1 Cf. Evans, 'Stone Implements,' pp. 103, 116; Southall, 'Rec. Orig. Man,' 

 p. 391 ; Skertchly, ' Across Shantung,' Hongkong, 1892, pp. 5, 6. 



