58 TEMPERATURE. 



packed, the low temperature and entire absence of melting preventing the snow crystals 

 combining into anything approaching solid ice. Unfortunately no actual measurements of the 

 density of the Barrier snow are available, but it is almo.st certain that it was only -2 or -3 of 

 that of ice. In McMurdo Sound on the other hand the surface is sea ice, which at the end 

 of the winter has a very thin coating of snow. On account of the high wind which blows 

 through the Sound there are then large patches of ice with no snow at all and the snow 

 which does exist is large grained and very compact. 



The thermal conductivity of ice is very much greater than that of snow, and the con- 

 ductivity of snow is greater the greater its density. 



Let us now consider the temperature of the surface of the Barrier during a bright sunny 

 day and the following night. During the day time the sun shines on the snow and al- 

 though a large proportion is reflected, some is absorbed. As the conductivity of snow is very 

 small this heat is not conducted downwards to any considerable depth, but is all retained 

 near the surface ; also on account of its small density the thermal capacity of snow is small. 

 Thus a very little heat, being retained near the suiface and acting on a substance of small 

 heat capacity, raises the temperature of the snow near the surface appreciably. During 

 the night the snow surface radiates heat to the clear sky and its temperature falls, but 

 those conditions which caused its temperature to rise while the sun was shining now act 

 in the opposite direction. On account of the small heat capacity of the snow a small loss 

 of heat by radiation lowers the temperature con.siderably, and as the low conductivity pie- 

 vents heat being conducted from below, the temperature of the surface may f^ll very greatly 

 during a clear night. 



Let us now consider that the Barrier is changed into ice with just sufficient snow on 

 the surface to give it the same absorbing power. The amount of heat absorbed \vi)l be the 

 same as before, but owing to the much greater conductivity of ice — at least a hundred times 

 as great as that of snow — changes of temperature at the surface are accompanied by a 

 greater transference of heat into the mass of the ice. This keeps the surface temperature 

 lower during the day, but during the night when the air temperature falls this same amount 

 of heat can agam reach the surface, and if the radiation is particularly great still more heat 

 which was stored up in the ice will be conducted outwards and so prevent the temperature 

 of the surface falling so low as it would have done had the mass been snow through which 

 heat can only pass slowly to the surface. If now instead of the Barrier we consider a sheet 

 of ice floating on the sea, the temperature of which cannot fall much below freezing point, 

 we have a still greater protection against an excessive fall of temperature during the night ; 

 for the ice with its high conductivity is always conducting beat from the warm sea below 

 and is therefore always a little warmer than the mean temperature of the air above it. 

 Thus as soon as the surface temperature falls owing to radiation, the heat from the sea 

 stored up in the ice becomes available to counteract any large fall of temperature of the 

 surface due to radiation. 



Thus we see that the chief difference between a snow and an ice surface is that the 

 latter prevents low radiation temperatures, while excessive radiation always produces low 

 temperatures over the snow. 



We have now the explanation of the different daily variations of temperature over the 

 Barrier and over McMurdo Sound. Very low night temperatures are possible over the former 

 owing to the bad conductivity of the snow, while over the latter the high conductivity 

 of the ice and the store of heat retained in it from the warm sea below prevent low tem- 

 peratures during the night. It will be remembered that the observations showed that the 

 day temperatures of the Barrier were nearly the same as those at Cape Evans, while the 

 night temperatures on the former were many degrees lower than those at the latter. 



