Points in Climatology. 147 



present reasoning so obviously inconclusive as that in which 

 he endeavours to show that my objection to the reliableness of 

 his dates for glacial epochs, on account of the insufficiency of 

 the fundamental data for the secular variations of the planetary 

 orbits, falls to the ground. My objection and his statement 

 in reply I can leave to the judgment of the reader who chooses 

 to refer to them. 



I conceive that some general remarks on the nature of the 

 problem will be of more value than a further analysis of Mr. 

 CrolPs reasoning. It is an observed fact that we now have a 

 glacial epoch at a comparatively moderate height in the atmo- 

 sphere, and on the tops of most high ranges of mountains far 

 removed from the equator. It is evident that if, at any former 

 epoch, the state of things at the surface of the ground was the 

 same that it now is at the height of two or three miles in the 

 atmosphere, there must have been a glacial epoch. To what 

 causes are we to attribute the cold of the upper regions of the 

 air ? There are two known causes, but we cannot assign an 

 exact quantitative effect to each. 



I. The passage of air from the lower to the upper regions is 

 accompanied by expansion, and the reverse motion by com- 

 pression, which would naturally result in the upper regions 

 being colder than the lower : the exact amount of cooling, 

 supposing no disturbing cause to come into play, is readily 

 computed, and has, I think, been assigned by Professor Sir 

 William Thomson and others ; but I need not now refer to 

 the results. 



II. Researches on radiant heat seem to show that the atmo- 

 sphere absorbs the extreme rays of the spectrum, especially 

 those of greatest wave-length, more powerfully than the rays 

 of mean wave -length. The rays radiated by the earth are of 

 longer wave-length than the great mass of those received by 

 the sun. The natural result of this selective absorption would 

 be to make the temperature of the earth higher than if there 

 were no atmosphere, or if the atmosphere exercised no selective 

 absorption on heat-rays. It seems probable that this selective 

 absorption is clue very largely, if not entirely, to aqueous 

 vapour in the air. If this be so, an epoch of dry air would be 

 a glacial one. 



A crude test of the efficacy of the first cause might be de- 

 vised. In order that it may act, it is essential that there shall 

 be a continuous interchange of air between low and high 

 altitudes. Now if there are any high table-lands so extended 

 that, in their central portions, the air has not during several 

 days an opportunity to be replenished from lower regions, such 

 air should be warmer than that at an equal height on isolated 



L2 



