and Heat-conduction in rarefied Gases. 61 



be regarded as an actual vacuum with respect to heat-conduc- 

 tion, we had a thermometer manufactured which, by means of 

 slips of glass, could be put into two different envelopes. These 

 were selected so that with medium pressures the times of cool- 

 ing of the thermometer in them were in the ratio of 1: 2. If 

 now the best vacuum possible was produced in the envelopes, 

 the result was that the times of cooling approached to within a 

 few per cent, of equality. 



Atmospheric Air. 



p. App. I. App. H- 



millim. seconds. * seconds. 



760 .. . 171 114 



148 .. . 234 114 



95... 270 116 



0-5 . . . 280 154 



Best vacuum . 576 576 



If the cooling-spaces were filled with hydrogen or carbonic 

 acid and then exhausted as completely as possible, the results 

 were : — App. I. App. II. 



seconds. seconds. 



Hydrogen vacuum . . 588 578 



Carbonic-acid vacuum . 586 578 



We thence concluded that, in these vacua, the quantities 

 of heat abstracted from the thermometer proceeded almost ex- 

 clusively from radiation. It is true it follows from the gas- 

 theory that, from certain degrees of rarefaction onwards, the 

 quantity of heat conducted is dependent only on the num- 

 ber of the molecules in the unit of space, and independent of the 

 dimensions of the cooling- space ; but with such degrees of rare- 

 faction that quantity of heat is an extremely small fraction of 

 that carried over with normal pressures. 



We purpose to utilize such vacua for the determination of 

 the absolute emissive power of black bodies and the depend- 

 ence of its value on the temperature. 



Using the water-value of our thermometer (to be sure only 

 approximately ascertained), from our experiments up to this 

 time a value of the emissive power of glass is obtained which 

 nearly agrees with that ascertained by M. Lehnebach*; for 

 many reasons, however, we lay no stress on this accordance. 



On the dependence of emissive power on temperature we 

 possess the law of Dulong and Petit, according to which its 

 value is proportional to a', where a, referred to Centigrade, 

 = 1-0077. But in the experiments on which Dulong and Petit 

 founded that law the rarefied air of the cooling-space must, 

 * Pogg. Ann. vol. cli. p. 96. 



