38 METEOROLOGY 



ice crystals or ice grains which were left having increased in size at the expense of 

 the grains which had been partly vaporised, partly absorbed, while thawing, to 

 swell the volume of the larger granular crystals. Underneath this eight or nine 



Feet. Inches. 

 ■ I Finegrained white hard snov^. 



I 



I 



« • 





A. Enlarqed. 



5/70JV showing development of neve crystals 

 becoming successively larger downwards. 

 There is a considerable amount of 

 interstitial airspace due in part to 

 sublimation, in part to closer packing of 

 tiie neve crystals below which thus leave 

 open air spaces above them. 



Fig. 7. 



Ice crystals . developed out of the snow, 

 found neKt' to the sea ice Just soutli of 

 Nordenskjold fee Barrier Tongue. 



Natural size. 



SKETCH SHOWING EFFECT OF SUBLFMATJON ON SNOW OVERLYING 

 SEA ICE NEAR NORDENSKJOLD ICE BARRIER 



inches of crust snow and ice granules was an older surface of crust snow. The 

 structure of the whole is illustrated in Fig. 8. 



It seems probable that the formation of this crust snow is somewhat in this 

 way : — The sun on bright clear days warms the air above the snow and even in the 

 interstices between the snow grains for some little depth down.* Consequently 

 evaporation during the wai'mer part of the day becomes very rapid, chiefly at 

 the surface of the snow, but also penetrating to a depth of several inches. In 

 cases where the snow seldom actually thaws, as on the vast nev^ fields of the 

 high inland plateau, little but waste of the snow surface takes place during the 

 warmer parts of the day ; but as the surface temperature falls at night the ice 

 vapour still rising from the interstices between the snow crystals beneath the surface 

 is deposited in a solid form on the snow crystals at or near the surface, and so tends 



* That tlie sun's heat passes rapidly through a considerable thickness of ice, is abundantly shown by 

 the temperatures taken by us in shafts sunk through the lake ice of Cape Royds, as explained elsewhere in 

 this memoir. Its penetration through snow would no doubt be much slower. Unfortunately we did not 

 onduct any experiments on this subject. 



