382 Andr. Lundager 



First of all, it was of importance to find a drift which had 

 been lying somewhat undisturbed since the autumn, and which had 

 merely been increased, without, at other times, with changes of the 

 wdnd, having yielded an}^ of the snow at one time deposited; as in 

 such a case, a lower temperature would have availed itself of the 

 opportunity to penetrate to a greater depth than, later in the winter, 

 and in accordance with the thickness of the layer of snow, it might 

 be expected to be found. Therefore the winter of 1906 — 1907 was 

 more suitable for these measurements than the following winter 

 when less snow fell, and no great snowdrifts lay unaffected by 

 such erratic snowstorms, which might at times sweep away layers 

 one metre in thickness, which layers would be replaced only when 

 a storm from the direction of the prevalent wind (NW.) brought 

 new layers upon the ordinary deposits to leeward. For the measure- 

 ments, this lack of a constant layer of snow causes the results to 

 be arbitrary, and makes it impossible to establish absolute values 

 for the range of the temperature in the snow. 



Even at a depth of 2 metres, I did not reach the rock beneath 

 the snow; but this was as hard as ice, and of a consistency that 

 only allowed of a very lengthened process of melting. As a rule, 

 a vegetation which is covered by such an enormous layer will see 

 daylight only so late in the year that it will never, in any case, be 

 able to reach the flowering stage, and will not even be able, every 

 year, to awake to life. 



The figures show that the temperature is considerably higher 

 downwards in the snow. Still, the mean temperature for the month 

 stands rather high in the layer of snow, at a depth of about 30 cm., 

 I think. 



Twenty-four hours later, in the same hole, the snow-walls 

 of which had been exposed to the temperature of the air, the 

 following measurement was taken at a depth of 2 metres: — 22.5. 

 The surface, like the air, being — 32.0. 



Again, on April 7, I took measurements in the same place. But 

 the layer of snow had increased by about 45 cm. in thickness. At 

 this depth a hard crust was then lying, which indicated the surface 

 of the snow on February 23 ; as this was now well hidden beneath 

 the new cover of snow, it also showed a rise in temperature (from 

 — 28.5 to — 21.5) owing to the fact that now the radiation of heat 

 from below was prevented. 



The same circumstance was noticed at a depth of 1.60 metre, 

 which almost corresponds with the depth of 1 metre on February 23 

 ( — 18.4 against — 20.0). Here also a hard crust was found, which, 

 earlier in the winter, had been the surface of the snow. 



