96 PRINCIPAL FORBES ON AN EXPERIMENTAL INQUIRY INTO 



safe limits of observation. In fact, if we compare the average ratio froml 0° to 

 100° Cent., and again from 100° to 200° Cent., we shall find them to be almost 

 identical. They give for the value of p, the number 1 6023. This represents the 

 proportion in which the papered bar dissipates its heat more rapidly than the 

 naked bar. 



93. For the direct radiating or emissive power of the two surfaces, I had re- 

 course to the kind aid of Mr Balfour Stewart, not having had recently con- 

 veniences for making the experiment myself. He used the thermo-electric pile, and 

 he found the experiment to be attended with considerably greater difficulty than is 

 commonly attributed to it. I believe that Mr Stewart is not yet satisfied as to 

 the reliability of his methods of observation ; but the four best series of experi- 

 ments made in February and March 1864, gave the emissive power of paper 

 compared to iron as 5*8 to 1.* The value of q is therefore 5-8. 



94. Hence by the previous investigation — 



The value of x, the heat dissipated by radiation from naked iron (the dissipation 

 by convection being always =1) is ^ = 5T g- ro60 = 0-116. In the case of the paper 



surface, x is 58 times greater, or = 673. In other words, of the heat dissipated 

 from the bar in Case I., nearly T 9 ths are lost by convection, and ^th by radia- 

 tion. In the paper-covered bar (Case II.;, only T %ths are lost by convection, and 

 3*0 ths by radiation. f 



95. From this it appears that the principal agent in the dissipation of heat in 

 these experiments is Convection and not Radiation ; nay, that the effect of the 

 latter is comparatively almost insensible, when naked metallic bars are used. 

 This of itself tends to explain the systematic deviation of the statical curve of 

 temperature (Art. 64) from the logarithmic law. The experiments of Dulong 

 and Petit show that the dissipation of heat due to Convection increases not as 

 the excess of temperature simply, but as its fth power nearly (more exact! v 

 1-233). This accords so far with what has been said of the variation in the rate of 

 cooling in Art. 84 ; but it gives no adequate explanation of the inflections of the 

 curves of Plate V. at higher temperatures. Were it not for the unquestionable 

 precision of Dulong's admirable experiments, in which the law of cooling due 

 to the contact of air was verified as high as 260° Cent., one might have not un- 

 reasonably supposed that the energy of convection was relatively less at higher 

 temperatures. 



* This corresponds nearly to the relative emissive power of glass and polished silver used bv 

 Dulong. 



\ For in Case I. the whole heat lost from a point having a given temperature being represented 

 by the number I'll 6, that due to Convection is 1, that due to Radiation is -116. In Case II. the 

 total loss is 1673, whereof 1 is due to Convection, and 673 to Radiation. 



