POLAR ICE-CAPS AND SEA LEVELS. 357 



some little time like that, with an ice-cap 2,000 feet 

 thick and it is not at all improbable that that is 

 the present condition and suppose that under 

 some of Croll's supposed .astronomical changes 

 snow began to fall twice as fast as now, and that 

 it went on snowing for 2,000 years, the result 

 would be that there would be a gain at the rate 

 of, in round numbers, 2 feet per annum for 100 

 years twice as fast as at present for 1,000 years. 

 Set off against that the sliding down, and you 

 get a very complicated problem. The terrestrial 

 temperature would go for nothing. The thermal 

 conductivity may be expressed thus : How much 

 heat would be conducted per annum through the 

 27 metres of rock ? You can make a calculation, 

 remembering that there are 31.* million seconds 

 in a year. The temperature increases by one 

 degree centigrade per 2700 centimetres, and the 

 thermal conductivity of average rock (in gramme- 

 water thermal units per square centimetre, per 

 i per centimetre of rate of variation of tempera- 

 ture), is '005. Thus we have the calculation : 



^ XIO Voo 5 . 



2700 



