56 



temperature indicated by a thermometer in the shade and exposed 

 to the sun's rays, and, determining by theory the proportion of air 

 traversed at these altitudes, he inferred the loss at a vertical inci- 

 dence to be ygths of the whole. On these slender foundations some 

 philosophers have ventured to found most important conclusions on 

 various subjects of transcendental speculation ; — such as the temper- 

 ature of the sun's surface, and the temperature of planetary space. 



Facts, however, of much greater value may be looked for, since the 

 invention of Sir John Herschel's actinometer. Proceeding like Lam- 

 bert on the principle, that the heating and illuminating powers of 

 the sun's rays are proportional, this instrument measures the light 

 by measuring the heat they produce. But, avoiding the great defect 

 of Lambert's method, viz. the assumption, that the stationary condi- 

 tion of the thermometer is proportioned to the heating cause, a law 

 which would only hold, if the cooling influence on the thermometer 

 were invariable ; Herschel proposes to ascertain, by an extremely 

 delicate thermometer, ^firsl, the velocity of heating in the sun's rays, 

 or the number of degrees passed through in a given short space of 

 time, and then, secondly, the rate of cooling in the shade. The alge- 

 braic difference of these expresses the excess of one effect over the 

 other. These data will obviously supply the means of measuring 

 the true intensity of the sun's rays ; since the velocity of changing 

 temperature is known by Newton's law to be the measure of the pro- 

 ducing cause. 



Having made these preliminary explanations, the author proceeded 

 to observe, that, at the request of Sir John Herschel, he had made 

 some observations with two instruments entrusted to him by that 

 gentleman during his visit to the Continent in 1832. Twenty series 

 of observations were made on six different days in Swizerland at va- 

 rious altitudes, the comparative observations being made by Profes- 

 sor Kamtz of Halle in the overland of Berne. The columns of air 

 varied from 5000 to 7000 feet. On one day, observations were made 

 every hour from sun-rise to sun-set. Every observation indicated 

 increased radiation at the higher station, and the general diminution 

 appeared to be not less than ^th for the thickness just mentioned, 

 which is afar greater proportion than was assigned even by Lambert. 



The author found, by an extensive series of experiments made at 

 the Observatory of Paris, that, in favourable circumstances, the nu- 

 merical estimate of radiation by Herschel's instrument may be relied 

 on to Yffo*^ ^^ '*® amount. 



