40 WRITINGS OF JOSEPH HENRY. [1855- 



earth's orbit which approaches nearest the sun is constantly 

 changing its place, and in time the order will be reversed; 

 the greatest amount of heat from this cause will be on some 

 day in July, and the least in January. But this change is 

 so slow, that no appreciable effect has been produced during 

 the historic period, A slight variation also takes place in 

 the distance of the earth and sun when nearest to each other; 

 but this also is confined to such narrow limits, that it is 

 entirely insufficient to account for the changes undergone 

 in the earth's temperature, as indicated by fossil plants and 

 animals, and cannot, on account of its slowness, have had 

 any appreciable effect upon the temperature of any part of 

 the earth since the first records of civilized man. If there- 

 fore it be true, as some suppose, that the seasons have 

 changed in different parts of the earth within the memory 

 of man, the effect must be due to other than to astronomi- 

 cal causes. 



The earth is approximately a sphere, and consequently, 

 the sun's rays strike it obliquely at all places, except those 

 over which it is precisely vertical. The amount of variation 

 on this account can readily be calculated ; the sun's beam 

 may be considered as a force, and resolved into two parts, 

 one of which is parallel to the surface of the earth, and the 

 other perpendicular to it, the latter alone producing the 

 result. The intensity of the sun's beam will be the greatest 

 at the equator, and will gradually diminish to the poles. It 

 is true the sun does not continually remain vertical at the 

 equator, but the average result in the course of the year, is 

 nearly the same as if this were the case ; since the greater 

 amount of heat received while he is at the north just com- 

 pensates for the less while at the south. The average tem- 

 perature of any given place, in consideration of the obliquity 

 of the rays which the earth would receive if uninfluenced by 

 other conditions, can be obtained by multiplying its equa- 

 torial temperature into the radius of its parallel of latitude; 

 or (in more technical language) into the cosine of the latitude. 



From this formula, which we owe to Sir David Brewster, 

 we have calculated the following table, which exhibits the 

 astronomical and observed temperatures of the valle}^ of the 



