464 MELLON! ON THE NOCTURNAL COOLING OF BODIES. 



deduced from the researches of Leslie, Petit and Dulong, will he 

 effaced from scientific works and replaced by measures of greater 

 accuracy. 



The parallel movement of the thermometers with pohshed ar- 

 matures, in the closed and open vessels, and above all, the ex- 

 tremely slight difference in their indications, show that in both 

 cases these instruments give the true temperature of the stra- 

 tum of air in which they are plunged, and consequently the 

 closing of the vessel is useless, Avhen the nature of the experi- 

 ments does not require an extreme degree of precision in the 

 measures. It is moreover quite evident that if the presence of 

 the vessel (the sides of which tend to increase the cooling of the 

 thermometer, by preserving it from the radiation of the ground, 

 by reflecting towards the sky the heat radiated by the instru- 

 ment towards the earth's surface, and by maintaining a certain 

 calm in the air surrounding it) does not sensibly alter the indi- 

 cation of the thermometer relatively to the temperature of the 

 stratum of air in which the instrument is plunged, this tempe- 

 rature will be given with so much the greater accuracy by the 

 thermometer provided with a simple metallic case, and freely 

 suspended by long metallic threads, or placed on a support 

 formed of tubes of tin-plate, as we have above described. 



The means of obtaining the true temperature of the air being 

 known, nothing is easier than to determine the different degrees 

 of cold, that is to say, the depressions below the temperature of 

 the air, produced by the nocturnal radiations of different sub- 

 stances. In fact, it is sufficient to apply these substances on the 

 armatures of a certain number of thermometers, introduced into 

 their respective conical recipients and exposed to the free air 

 during calm and serene nights, together with the thermometer 

 armed throughout with polished metal, which gives at each in- 

 stant the temperature of the air, and which, for greater perspi- 

 cuity, we shall call the atmospheric thermometer. All the ther- 

 mometric reservoirs ought to be at the same height. Each ther- 

 mometer being compared, at several periods, with the atmo- 

 spheric thermometer, the frigorific action of the substance which 

 envelopes it will be equal to the difference between the two in- 

 struments, when this difference is preserved invariable during 

 two or three consecutive observations. 



