484 MR. CHAS. DAVISON ON SNOWDRIFT DEPOSITS. [Aug. 1 894, 



VI, The Origin of the Loess. 



I propose now to refer as briefly as possible to the geological 

 problems mentioned at the beginning of this paper. 



The snowdrift theory of the loess is in great part described in the 

 preceding pages. During the Glacial period a large part of Europe 

 was unoccupied by ice. In the districts adjoining the ice-sheet, a 

 similar though milder climate must have prevailed. The winters 

 must have been long and the summers comparatively short ; the 

 former being characterized by heavy storms, violent gales, and much 

 snowdrifting. Abundant dust-material would be provided by the pro- 

 ducts of glacial erosion. During the summer most of the snow, except 

 in sheltered places, would melt, and the snowdrift deposits would 

 be formed chiefly in the valleys, more thinly on the higher ground and 

 plateaux. Thus, every year a new layer would be added, each layer 

 merging imperceptibly into that of the previous season, the whole 

 covering high and low ground alike, except that it would attain a 

 greater thickness in the more sheltered places, and would be 

 altogether absent in exposed spots on hillsides and crests. "With 

 the advent of a warmer climate and the retreat of the ice-sheet, 

 there would be a gradual expansion of the snowdrift formation, 

 overlapping the glacial deposits along their border, but diminishing 

 in thickness the more rapidly the ice withdrew. 



In many respects the snowdrift theory bears a close resemblance 

 to the seolian theory of Baron F. von Eichthofen. But they obviously 

 differ in some important points. I will now endeavour to show, 

 (i) that the snowdrift theory accounts for the more prominent 

 features of the loess, and (ii) that it escapes several of the objections 

 which have been urged so forcibly against the aeolian theory. 



The snowdrift theory seems to me to give a satisfactory explana- 

 tion of the following peculiarities of the loess, which distinguish it 

 so strongly from the other formations with which we are acquainted : 

 (1) its independence of altitude above the sea-level ; (2) its occur- 

 rence in uniform sheets over plains and table-lands, and the concavity 

 of its surface when developed between two ridges ; (3) its homo- 

 geneous composition and structure ; (4) the complete absence 

 of stratification in pure loess, and the indiscriminate arrangement 

 of included mica-flakes ; (5) the occurrence of angular grains of 

 quartz ; (6) the inclusion of layers of angular fragments in loess 

 near hillsides ; (7) the great quantity of bones of mammals (this 

 will be referred to in the next section) ; (8) the presence of land- 

 shells and the preservation of delicate shells ; and (9) the peculiar 

 character of the fauna, resembling that of sub-Arctic steppes or 

 tundras. 



I have not referred here to the calcareous concretions, the root- 

 like tubular structure, or the tendency to vertical cleavage, all 

 characteristic features of the loess. The concretions are in all 

 probability a subsequent formation, and do not here require a 

 special notice. The explanation of the others given by Eichthofen, 

 if it holds for the aeolian theory, will also probably hold for the 

 snowdrift theory. 



