58 Emission and Transmission of Heat 



supposing these experiments to be perfectly exact, the results are 

 only relative and therefore cannot be used in practical applica- 

 tions. 



842. In the second edition of this work I have given the re- 

 sults of some experiments on the conductivities of the materials 

 under consideration, but the methods then employed were not 

 exact enough to give in every case sufficiently accurate results. 

 I have taken up these researches again using varied and more 

 precise methods; the details of which will be found in an appen- 

 dix to this edition, I shall limit myself here to a review of the 

 methods used, and a statement of the results obtained. 



843. In the first method which I have employed, the 

 material under experiment was enclosed between two spheres of 

 thin copper; the inner sphere was filled with warm water con- 

 stantly stirred, and the outer sphere was submerged in a bath, of 

 considerable size, of water at the ordinary temperature; this latter 

 was constantly stirred in the immediate vicinity of the sphere. It 

 can be shown by calculation that the cooling of the warm water 

 follows Newton's law, provided that certain relations between 

 the temperatures of different points of the spherical envelope hold 

 true at the beginning of the experiment, and these relations are 

 satisfied by bodies which are not too poor conductors of heat. 

 The coefficient of conductive power may then easily be deduced 

 from the rate of cooling. For quartz sand, mahogany sawdust 

 and starch paste, the rate of cooling followed exactly the law in- 

 dicated above, and I was able to obtain the conductivity of these 

 materials. But for cotton and powdered charcoal the cooling fol- 

 lowed a more rapid law and no definite conclusions could be 

 drawn from the experiments except that in the case of cotton the 

 rate of cooling under the same circumstances was sensibly the 

 same for densities varying from .0077 to .076, and I arrived at the 

 conclusion that the conductivity of this material was independent 

 of its density and in consequence that the conductivity of the fibre 

 composing it must be equal to that of motionless air. 



844 . Figure 1 68 represents a vertical section of the apparatus, 

 A B CD is a copper frame placed in a leadplated iron vessel filled 

 with water at ordinary temperature. The lower member B C of the 

 frame is provided with three toothed wheels E fand G (fig 169). 



