1898.] 



on Liquid Air as an Analytic Agent. 



823 



in each were absolutely identical, so that the results are simply com- 

 parative. The general ratio of heat propagation found for two 

 substances when different sets of double-walled tubes of about the 

 same form and proportion were compared, remained substantially 



J:''iG. 3. — Three tubed blowu on to oae ^tem, so that the exhaustion in each 

 would be identical. 



A, filled with lampblack between the inner and outer tubes. B, annular 

 space left empty. C, filled with silica between the tubes. A', B', C, the same 

 tubes in section. 



constant when a high vacuum was reached. A confirmation of the 

 results was generally made by noting the time required to evaporate 

 the whole of the air from each tube. The annular vacuum space 

 had generally a thickness of 4 to 5 mm., and was in each case com- 

 pletely filled up with the solid. In reality, however, the absolute 



3 H 2 



