148 WRITINGS OF JOSEPH HENRY. [1855- 



the whole night, as well as different nights, while the other, 

 namely the radiation from the air, would vary from hour to 

 hour, since it depends on the varying temperature of the at- 

 mosphere. 



By obtaining a series of observations in different states of 

 the atmosphere an assumption could be made as to the fixed 

 temperature of space which, when subtracted from the temp- 

 erature observed, would give the radiation of the column of 

 the atmosphere. 



Since it was impossible to cut off all the heat from the in- 

 strument except that which it received from the sky and air 

 above, and since it was exposed to but two-thirds of the 

 celestial hemisphere, some correction was necessary to reduce 

 the observed temperature to the true one. This was found 

 by making an artificial sky, formed of a zinc vessel about 

 forty inches in diameter, the bottom coated with lampblack, 

 and the whole filled with a refrigerating mixture. Beneath 

 this the " actinometer " was vertically placed at such distances 

 as to expose it successively to one-quarter, one-third, and two- 

 thirds of the hemisphere; and by repeating these experi- 

 ments with different temperatures of the artificial sky, it was 

 found that if from the temperature of the surrounding air f 

 of the lowering temperature of the actinometer were taken 

 away, the temperature of the artificial sky would be obtained, 

 since the same ratio would obtain in the case of the real sky. 

 In order to find therefore in all future experiments the temp- 

 erature which the actinometer ought to assume under the 

 radiation from space and the air above, it was only necessary 

 to subtract the degree given by the instrument from the 

 temperature of the surrounding air and multiply this by 

 f. From a series of observations thus corrected, he found 

 for the fixed part of the temperature given by the instru- 

 ment, or in other words the temperature of space, a value of 

 — 142° C. or — 222° F. This temperature is much lower than 

 that obtained before from considerations of a more theoret- 

 ical character. M. Pouillet however thinks that it cannot 

 be far from the true temperature of celestial space, since a 

 thermometer placed upon the coldest part of the earth and 



