100 G. F. Becker — Certain Astronomical 



peratnre at different epochs * The glacial epoch is relatively 

 speaking so recent that the composition of dry air was probably 

 much the same then as now ; but even moisture affects absorp- 

 tion of heat rays, greatly increasing it, so that relatively moist 

 air tends to become hotter while dry air tends to sink below 

 the average temperature. 



It is notorious that climate is enormously affected by oceanic 

 currents, themselves due to winds of prevalent direction, and it 

 would therefore be in vain to seek any very close relation 

 between temperature and solar radiation in littoral areas. On 

 the other hand, in the interior of the continents, where air 

 currents alone distribute the heat received from the sun, it 

 seems possible that a relation may show itself. Considering 

 the facts presented above it seemed to me that a reasonable 

 trial hypothesis would be this: — The average variation of tem- 

 perature from a certain mean in purely continental areas during 

 a great season may be nearly proportional to the mean rate 

 at which solar energy is received during that season. 



In an ideally continental climate the summers will be hotter 

 and the winters will be colder than in any real climate, since 

 the actual tendency is always toward an equalization of tem- 

 perature. Hence to test the trial hypothesis it is required to 

 know the lowest latitudes to which mean winter isotherms de- 

 scend in continental areas, and the highest latitudes which 

 mean summer isotherms reach. In this enquiry the lag of 

 heat effects would have to be taken into consideration, so that 

 the seasons for temperature would not be divided by the equi- 

 noxes. I do not know of any isothermal charts suitable for the 

 enquiry in this form. The next best material would be charts 

 for the two extreme months, January and July, for though 

 such charts would give temperatures exceeding the means in 

 intensity, these temperatures would probably be proportional 

 to the means. In other words it is probable that the tempera- 

 ture curve for a representative cold winter is derivable from 

 that of a representative warm winter by projection. 



Proceeding on this assumption I collected data from Mr. A. 

 Buchan's January and July charts based on the mean observa- 

 tions for eleven years.f The lowest latitudes to which each 

 isotherm descended in January in Asia and in North America 

 were noted and the highest latitudes to which they ascended in 

 July. They are recorded in the following table. 



* It is substantially certain that the sun's own heat has varied. The record of 

 its variation must exist in the rocks, but as a palimpsest. The greater equability 

 of the earlier climates seems to me explicable, at present, only by greater atmo- 

 spheric absorption. 



f Encyc. Brit., 9th ed. Art. : Meteorology. 



