394 Professor Dewar [Jan. 19, 



the results warrant the conclusion that at the zero of absolute tem- 

 perature all the pure metals would be perfect conductors of elec- 

 tricity. Under such conditions a current of electricity started in a 

 pure metallic circuit would develop no heat, and therefore undergo 

 no dissipation. Similarly, we infer there would be no Peltier effect at 

 the zero. In other words, the passage of electricity from one metal 

 to another would take place without evolution or absorption of heat. 



Further investigation, along with Professor Liveing,* on the re- 

 fractive index of liquid nitrogen and air, has led to the conclusion 

 that the refractive indices of nitrogen and air are respectively for 

 the D-ray, 1*2053 and 1*2062. In these determinations, instead of 

 using the prisms we have employed the method of Terguem and 

 Trannim, which consists in suspending in the liquid two plates of 

 glass with a thin layer of air between them, and measuring the 

 angle of incidence at which the chosen ray suffers total reflection at 

 the surface of the air. As all the vacuum vessels are either spherical 

 or cylindrical in form when filled with liquid, they act as lenses 

 which are irregular and full of striations. Further, small bubbles 

 of gas being given off in the liquid rendered any image indistinct 

 when viewed with a telescope. In order to avoid the necessity of 

 observing any image through the liquid, it was used simply as a 

 lens to concentrate the light observed on the slit of a spectroscope. 

 Under such conditions the observations were easily executed and the 

 results satisfactory. 



For some time a series of observations on the thermal opacity of 

 liquid oxygen and nitrogen have been projected. It is, however, 

 exceedingly difficult to experiment in such a way as to eliminate the 

 absorbing action of the glass vessels, and as the use of rock salt is 

 impracticable, the absorption of heat of low refrangibility remains 

 for the present undetermined. It is possible, however, to use the 

 glass vacuum vessels to determine approximately the relative thermal 

 transparency for boat of high refrangibility, such as is radiated by a 

 colza lamp. The following results represent the heat transmitted 

 through the same vacuum vessels filled with different liquids, taking 

 chloroform as the unit for comparison and correcting for differences 

 of refractive index. 



Chloroform 1*0 Liquid nitrous oxide .. 0*93 



Carbon bisulphide . . 1*6 Liquid ethylene . . . . • 60 

 Liquid oxygen .. .. 0'9 Ether 0-50 



From this result it follows that liquid oxygen is nearly as trans- 

 parent to high temperature heat radiation as chloroform, which is 

 one of the most transparent liquids next to carbon bisulphide. 

 Liquid ethylene is much more opaque. These results must, however, 

 be considered only as an approximation to the truth, and as generally 

 confirmatory of the inferences Tyndall drew as to the relation 

 between gases and liquids as absorbents of radiant heat. 



* 'On the Refractive Indices of Liquid Nitrogen and Air. By Professors 

 Liveing and Dewar. Phil. Mag. 1893. 



