Hakdcastle. — The Tiniaru Loess as a Climate Register. 327 



growing dust-heap played the part of an observant bystander', 

 taking notes of certain climatic phenomena as they succes- 

 sively arose. The record of the lowest separable layer, marked 

 off by a baud in which both drought-veins and rust-granules 

 occur, may, I would suggest, be read as follows : — 



1. A phase of cold, producing great icefields and glaciers in 

 the highlands, which scud down floods of sludgy waters, inun- 

 dating the lowlands, and creating fields of dust, from which 

 the winds picked up and deposited here a bed of loess up to 

 10ft. thick where the contemporary denudation was slight. 

 I This is the thickest of the layers.) 



2. A phase of improving climate, during which the glaciers 

 diminished and the supply of dust ceased, probably in part 

 through the trapping of the glacier silt in lakes or pools, occu- 

 pying basins scooped out by the previously extended glaciers. 

 The climate here continued wet, however, for even where the 

 slope of the surface afforded good drainage the rust-granules 

 characteristic of wet soils were formed. 



3. The climate further improved, becoming dry enough in 

 summer to crack the ground to the depth of a few feet, and 

 drought- veins were formed. 



4. The moist climate returned, the formation of drought- 

 veins ceased, and that of rust-granules was resumed. 



5. With increasing cold the glaciers again advanced, and 

 the supply of dust was resumed, this recommencing the series. 



If right in the main, this reading may be wrong by contain- 

 ing a redundant clause. There may have been but one phase, 

 not two, of wet climate, giving rise to the production of rust- 

 granules. If but one, there is some reason to suppose that it 

 was related to the return of the cold phase rather than to the 

 retreat of that phasQ. A ground for this supposition wnll be 

 staffed later on, in the suddenness of considerable improve- 

 ments in climate. 



The series of variations of climate registered in the first 

 layer of loess appears to have been fully repeated but once. 

 There are only two of the buried subsoils, so far as my ob- 

 servations show, that contain the drought-veins produced by 

 dry climate cracking the ground, the second of these being 

 near the top of the deposit. I have not been able to determine 

 the exact number of marks of pauses in the deposition, but 

 there are in one good section five or six distinct and a few 

 indistinct ones. As the faces of the cuttings arc coated with 

 rain-wash it is not easy to count these soil-bands with cer- 

 tainty. As each of them indicates a long period when the 

 supply of dust ceased, and as we trace the dust to an origiil 

 dependent upon a certain condition of climate, each of these 

 bands registers an absence of that condition — in other words, 

 an important alteration in the climate. 



