CAUSES OF THE GLACIAL PERIOt) 567 



moisture. This conclusion is based on actual observation. Different 

 authorities may not agree as to the amount of effect produced by changes 

 in CO2, but all appear to agree that there is some effect. When it comes 

 to changes in ocean currents, however, the case is not so clear. Here the 

 hypothesis is necessarily not based on direct observation, but on a highly 

 complicated chain of reasoning. A periodic reversal of oceanic circula- 

 tion appears to be the only way in which Glacial epochs can be explained 

 according to the carbonic acid hypothesis, but the actual occurrence of 

 any such reversal because of the action of CO2 has never been demon- 

 strated, even on a small scale. In the very nature of the case, any such 

 demonstration is impossible, for the phenomenon must occur so slowly 

 that many centuries would be required to detect it. In brief, the chief 

 importance of COg would seem to be in the production of climatic eras 

 far longer than Glacial epochs. It may cause long periods of mild cli- 

 mate when equatorial conditions prevail far toward the poles, or equally 

 long periods when the atmosphere is relatively free from CO2 and the 

 earth^s temperature falls somewhat. This would mean that the influence 

 of the composition of the atmosphere would vg|,ry with changes in the 

 extent and elevation of the lands. In connection with the form of the 

 solar hypothesis here presented we fully accept the idea that both defor- 

 mation of the earth^s crust and changes in the amount of carbonic acid 

 in the atmosphere have been and will continue to be among the chief 

 causes of climatic changes whose length is measured in hundreds of 

 thousands or millions of years. They do not seem, however, to have been 

 anything like so effective in producing changes measured in hundreds, or 

 thousands, or even tens of thousands of years. 



THE CYCLONIC SOLAR HYPOTHESIS 



General discussion. — If we have reasoned correctly in our exclusion of 

 other hypotheses, the only one which seems to be competent to explain 

 Glacial epochs and the minor cycles shown by the California trees is the 

 solar hypothesis. In its "caloric" form it does not seem to stand the 

 test, for present changes of climate do not agree with changes in solar 

 temperature. Moreover, from the point of view of the physicist, it seems 

 beyond the bounds of probability that the sun's mean temperature should 

 change sufficiently often and with sufficient rapidity to cause the observed 

 terrestrial phenomena. The cyclonic form of the hypothesis seems to be 

 free from such objections. We have already seen that there is a striking 

 agreement between the changes of solar spots and variations in storms 

 and winds. We have also seen that there is no inherent reason why the 

 activity of the sun's surface, especially in its magnetic or electrical con- 



