TABLE 42. CHANGES EN THE LENGTH OF 1 KM OF ICE IN M 

 WITH A 1° RISE IN TEMPERATURE 



Malmgren conducted direct determinations of this coefficient in a specially constructed ap- 

 paratus simultaneously with a computation of the coefficient of expansion. Furthermore, he com- 

 pared his data with data of Patterson. Petterson had conducted his experiments very carefully with 

 artificially prepared sea ice. It developed that all the results are in good agreement. This proves 

 first of all the correctness of Malmgren's reasoning and secondly the fact that air bubbles within 

 sea ice play a secondary role in the thermal expansion of ice. The latter follows from the agree- 

 ment of the data obtained by Petterson when investigating artificially prepared sea ice devoid of 

 any air bubbles with Malmgren's observations of natural ice which contained air bubbles. 



LITERATURE: 52, 53, 62, 73, 104. 



Section 64. Thermal Conductivity 



The coefficient of thermal conductivity of pure ice, devoid of air bubbles, as an average of 

 the data of many investigators, is given by 



K = 0. 00540 g-cal/sec ■ deg x cm 



wherein, according to Lis, it decreases somewhat (approximately 0. 00001 per 1°) with a decrease 

 in temperature. 



Malmgren determined the thermal conductivity of sea ice using both direct and indirect meth- 

 ods for this purpose. 



On the basis of his indirect computations for the coefficient of thermal conductivity, Malmgren 

 gives a chart (figure 56) which represents the average changes of the thermal conductivity as 

 a function of the depth of the ice level; Malmgren shows that the rapid decrease in the coefficient of 

 thermal conductivity when approaching the upper layers of the ice is explained by the presence of a 

 multitude of small air bubbles in these layers. At a great distance from the surface, the thermal 

 conductivity of sea ice approaches the thermal conductivity of pure ice containing no air bubbles. 



As Chernigovskii indicates, according to his computations which were conducted by the same 

 method as Malmgren's computations, the thermal conductivity of fast Ice on the Kara Sea increased 

 from winter to summer and from the upper surface of the ice to the lower. Thus, at cm level, 

 it was about 0. 001, and at the 150 cm layer it was about 0. 0044. 



160 



