Vol. 50.] ME. CHAS. DAVISON ON SNOWDRIFT DEPOSITS. 483 



covering of mud was formed on the surface. After 18 days the 

 snow had entirely disappeared, and a deposit nearly an inch thick 

 was left behind. In this there was not the least sign to be detected 

 of any surface of separation between the mud from the second and 

 that from the fourth layers. The only trace of their original division 

 was the presence of the mica-flakes in the upper part of the deposit. 



This experiment is, I think, of some value, though not perhaps 

 decisive — on account of the small scale on which it was made. To 

 prove the general absence of stratification in snowdrift deposits, I 

 would rather trust to the absence of any reason why stratification 

 should exist. 



Arrangement of Mica-flakes in Snowdrift Deposits. — In the Chinese 

 loess, Eichthofen has observed that mica-flakes are arranged in- 

 discriminately and not horizontally, and it was on this account that 

 mica-flakes were mixed with the dust in the fourth layer of the 

 preceding experiment. The flakes were generally small, most of 

 them about y^ inch or slightly more in diameter, but some were 

 larger, and a few were about | inch long. While the snow was 

 melting, these flakes appeared in the mud extruded on the surface, 

 but there was no sign whatever of a horizontal disposition. When 

 the snow had completely disappeared, from the largest flake to the 

 smallest they were inclined at all angles to the horizon, and in 

 every direction. 



Occasional Stratification in Snowdrift Deposits. — Pools of thaw- 

 water are sometimes formed on the surface of decaying snow ; streams 

 of melted snow may carry into them some of the surface-mud, and 

 true aqueous deposits may thus be intercalated in the snowdrift 

 formation. These pools are no doubt caused in some cases by the 

 varying thickness of the incipient snowdrift deposit. " Round a 

 ship which has wintered in the ice," says Payer, " there is gradually 

 accumulated a mass of rubbish of all kinds, of which cinders form a 

 considerable constituent. These, when thrown out in small quan- 

 tities, sink at once into the snow, while larger quantities act as a 

 non-conducting layer. Hence we were surrounded by a maze of 

 holes, big and little, alternating with plateaux, under which winter 

 still continued to linger. When thaw-water made its appearance, 

 all this was transformed into a succession of lakes and islands, 

 which we bridged over by planks." l 



During rapid tbaws, the water rushes down valleys and ravines 

 in torrents, and mud and gravel are spread out in sheets over the 

 snow and ice at their mouths. 2 



But in both these cases, unless the deposits are laid down directly 

 on snowdrift formations having no thick layers of snow and ice 

 beneath them, the original stratification must be largely interfered 

 with by differential movements due to unequally rapid melting of 

 the snow below. 



1 Payer, vol. i. p. 251. 



2 Kares, vol. ii. pp. 12, 55-56, 65-66 ; Scoresby, vol. i. p. 476. 



