p.2

1867

It has been proposed *declined the mean [to] temperature of the year by a regular mathematical curve, the curve of sines, having one maximum and one minimum in the year, which are precisely half a year and [undue?], and exactly midway between these are two points where the curve intersects the line of mean temperature corresponding to those two days, one in the spring and the other in the autumn when temperature is the same as the mean of the year. The entire year is supposed to be divided into 360 degrees, and the amplitude of the wave (or radius) or [the] greatest departure from a straight line is one half the annual range of temperature.

In the same number of Silliman's Journal that contains the article by Prof. Everett, Prof. E. Loomis shows quite conclusively, that actual curve as derived from observation does not correspond with this regular curve of sines, being sometimes above and sometimes below it; and therefore concludes that this curve does not represent climates, with that degree of accuracy which science requires.

Upon endeavoring to find the causes that make the march of temperature thus vary from that regular curve which from astronomical phenomena we might expect, it occured to me that we ought here have a measure of the effects of local causes upon the temperature of any station.

The curve of sines may be taken as the normal or standard curve, representing what the temperature would be each day were the surface of the earth one uniform plane, without mountains, without oceans

*J.D. Everett. Am. Jour. Science (2d series) 85 p 18