56 E. P. Culverwell — Theory of the Ice Age. 



I. Examination of Sir Egbert Ball's " Cause of an Ice Age." 



The first pages whicli deal with the astronomical explanation of 

 the Ice Age are pp. 43-4:7. Here Ball practically agrees with Croll ; 

 the only difference being that while Croll takes — 239° F. as the 

 " natural zero," or temperature, to which the earth would fall if 

 the sun-heat were removed, Ball takes — 300 F. Thus, all the 

 objections to Croll's argument C (see the paper in the January 

 Number) apply equally here. In addition to what I have there 

 pointed out, I give the following discussion to show how the method 

 on which Ball and Croll rely, if properly followed, confirms my 

 contention. 



Let us try to apply this method to the temperature of Great 

 Britain, amending it where it is in direct conflict with known 

 physical laws. We must proceed somewhat as follows : Since the 

 temperature reaches its maximum and minimum in July and January 

 respectively, we know that in each of those months the total heat 

 received is equal to the total heat lost. Now, the loss arises from 

 two causes — first, heat lost by radiation from the earth; and second, 

 heat lost by transference to other parts of the globe. Similarly the 

 gain arises from radiation from the sun, and from heat transferred 

 from other parts of the earth. We can strike a balance between 

 the heat lost and gained by transference, and call it the heat gained 

 by transference ; but we must, of course, remember that this will be 

 a negative quantity if more heat be transferred northward than is 

 received from southward. 



Now we assume with Croll that the radiation from the surface of 

 the earth is to be treated as a direct radiation into space ; that is to 

 say, we neglect what is probably one of the most important factors 

 in terrestrial temperatures, the blanketing effect of the atmosphere. 

 Then, assuming the truth of Stefan's law, we suppose the loss by 

 radiation to be proportional to (4:60 + 0*' Hence, if H be the heat 

 transferi-ed, and S the sun-heat and stellar radiation received, tliea 

 since in January and July, when the temperature is stationary, the 

 heat received must equal the heat lost, we get the equation, 



S+ 8 = A{4:Q0 + tY 

 where A is the coeflScient of radiation. As we do not know the 

 value of H, we must make various assumptions and see what their 

 effect will be. First, suppose that in July JI=z 0. Then, taking 

 lat. 55° as representing Great Britain generally, we have for July 

 S = 1700, and t = 60° F., and for January S = 265 and t = 40° F., 

 the 1700 and 265 being, of course, on an entirely arbitrary scale. 

 Hence for July, 



1700 = ^(460 + 60)* 



which gives A = 



(520)* 

 Hence for January, 



= 1455 



