SECT. 7] 



THE PHYSICS OF SEA-ICE 



833 



These phenomena were investigated by Malmgren (1927) and very Httle 

 additional work has been done. His results are summarized in The Oceans, 

 pp. 73-75, where tables are given showing the relationship of specific heat to 

 temperature and salinity, the variation of the coefficient of thermal expansion 

 with the same variables, and the total heat required to melt one gram of sea-ice 

 of given salinity and temperature. This last table gives an effective latent heat 

 for melting. For the freezing sea-water, an equation of Malmgren frequently 

 used is 



^^1^4: 



(2) 



2 4 6 8 10 15*. 



Fig. 3. Specific heat of sea-ice as a function of temperature and salinity. 



where L, Lp are the latent heats of fusion of sea-ice and pure ice, and Si, 8w 

 are the salinities of the sea-ice and sea-water. Malmgren's results will not be 

 repeated here except that some of the data on specific heat are plotted in Fig. 3. 

 The thermal conductivity of sea-ice is, like most other properties of this odd 

 material, dependent on its physical state. The air bubble content has a marked 

 influence. Malmgren found values from 1.5-5.0 x 10"^ cal cm~2 sec"i, with the 

 higher values applicable to the lower portions of an ice cover. The effect of gas 

 content and pressure on the thermal conductivity of ice was investigated by 

 Vlasov and Uspenskii (1931) and by Shuleikin et at. (1931). They showed that 

 for pure ice an increase of pressure decreased the thermal conductivity, but 

 that for porous ice containing carbon dioxide a pressure rise increased the 

 conductivity. At constant pressure an increase in gas content reduced the con- 

 ductivity. A number of observers have reported values of thermal conductivity, 

 all within the range found by Malmgren. These values have usually been 

 computed from measurements on growth rates of ice sheets and are thus not too 



