482 Rev. E. Hill — On Evaporation and Eccentricity. 



peratnre would not be nearly equal to that produced under a 

 temperature which oscillated from 30 degrees above to 30 degrees 

 below zero. We may express the theorem tersely, though perhaps 

 rather obscurely, as that " The mean evaporation under varying tem- 

 peratures exceeds the evaporation at the mean temperature." And any 

 cause which increases the fluctuations will still further increase the 

 excess. Thus, " when temperatures vary about a mean, increased 

 variation produces increased evaporation." 



Every one who has dipped into the Glacial Epoch controversy will 

 see that this fact has important bearings upon it. All agree that 

 increased eccentricity of the earth's orbit must bring wider variations 

 of temperature. Such wider variations would raise the total evapo- 

 ration. Increased annual evaporation must produce increased 

 annual precipitation. At present some of the precipitation takes 

 place in rain and some in snow. Each would he augmented. At 

 the Equator there would be heavier rains, at the Poles a greater 

 snowfall and all the consequent eifects. A sun nearer than at present 

 in summer, more remote in winter, will give a colder January and 

 hotter July. Hence will follow increased annual evaporation, and 

 increased precipitation, and a Polar snowfall beyond that now 

 experienced. But no increase of total annual heat accompanies this 

 snowfall to melt it. There is reason for a greater fall, but no reason 

 for greater melting. The surplus would accumulate, and although 

 doubtless small, may with sufficient time reach any reasonable 

 amount. 



The obvious objection to this argument is one formerly adduced by 

 myself against Dr. Croll, that snow liberates in its formation as 

 much heat as it absorbs in its dissipation. And this objection is 

 valid as regards all snow produced by the meeting of warm and 

 cold air-currents. The cold current is warmed as much as the warm 

 is cooled, and no heat thereby leaves the atmosphere. Increased 

 snowfall if it resulted would be accompanied by increased melting. 

 But there is a way in which heat is lost and snow produced wliich 

 this objection does not touch. If air be heated, it rises from the 

 ground and carries its heat into higher regions of the atmosphere. 

 Here as the heat is radiated oflP, it finds a thick blanket of air below, 

 a thinner coverlet above. Much less of the radiated heat will 

 return to earth, much more will be lost in space, than for alike mass 

 of air in contact with the ground. The nprush of warm air will 

 carry with it watery vapour. The particles of vapour under this 

 more rapid radiation will soon cool down, and may reach a very low 

 temperature. When the vapour falls to earth again, it will often fall 

 as snow. Any cause which tends to increase up-currents of air, 

 increases the loss of heat, and increases the rainfall and snowfall. 

 So does any cause which increases the vapour carried aloft by those 

 up-currents. Wider extremes of temperature produce such increase 

 of vapour, and increased eccentricity produces such wider extremes. 



A less obvious, very abstruse, but serious difficulty arises from the 

 simultaneous loss of heat by radiation. I have formerly shown 

 (Geol. Mag. 1880, p. 17) that these increased fluctuations of heat- 



