246 MESSES. SKERTCHLY AND KINGSMILL ON [May 1 89 5, 



or in close proximity to the superficial deposits, and the nature of 

 the authors' expedition gave them ample opportunities for examin- 

 ing the beds ; indeed, every exposure was noted, and a continuous 

 section constructed, as well as a detailed map of the Chefu district, 

 and a more general one of this and the neighbouring provinces. 



One of us (T. W. K.) had unhesitatingly declared for the marine 

 character of the loess ; but, being anxious not to invalidate the results 

 of the journey by any preconceived theories, it was agreed that the 

 two should work out the deposits conjointly, and that no opinion, 

 until accepted by both on independent testimony, should be placed 

 on record. This condition became the easier because it was by no 

 means clear that the superficial deposits in the neighbourhood of 

 Chefu, on which our investigations began, in any way represented the 

 loess. And, as a matter of fact, it was only by slow degrees, as we 

 advanced westward and noted the gradual changes in colour and 

 structure, that we recognized the representative character of the 

 local formations. 



The older rocks of this part of Shantung consist entirely of 

 quartzites, schists, gneiss, and crystalline limestones, all belonging 

 to the Archaean series, with intrusive granites and diorites. They 

 form picturesque mountainous hills, rising to 1 500 feet, which wind 

 about and so form almost land-locked valleys, whose beds are from 

 200 to 300 feet above sea-level. The loess sweeps from the sea- 

 level up into these valleys, and, looking down upon them and over 

 the plain from the hills, the terrace-structure is so marked that 

 one's memory harks back instinctively to the terrace-deposits of 

 Utah ; those, however, not having been so much interfered with by 

 cultivation, are far more distinct. The loess here runs up to about 

 350 or 400 feet, and all cols or gaps between the hills above that 

 height are free from it. Small streams flow down these valleys and 

 meander across the plain, cutting deep into the superficial beds and 

 exposing numerous fine sections. 



The material hereabouts is a dark, bluish-brown, loamy clay, 

 weathering yellowish-brown, often sufficiently sandy to form a 

 good brick-earth, much used in the district. The burnt bricks are 

 blue, red bricks being made by hydrating the ferric oxide, a process 

 which simply consists in sprinkling the red-hot bricks with water. 



The clay is stratified, the planes of deposition being marked by 

 different shades of colour in some instances, more often by lines of 

 pebbles where the deposit is close to the hills, and by the horizon- 

 tally of the calcareous nodules which are locally known by the 

 descriptive name of ' stone-ginger.' They are exactly comparable 

 with the similar nodules of ' race ' in the European loess. 



The character of the local rocks has clearly impressed itself upon 

 the deposit around Chefu, and save its tendency to split in vertical 

 planes it bears very little resemblance to typical loess. At first we 

 did not recognize it as such, but attributed it to local streams. It 

 was not until we had mapped it, and followed it for long distances, 

 that two facts dawned upon us : firstly, the impossibility of any 

 stream, flowing through small land-locked valleys with very small 



