388 



Scientific Proceedings, noyal Dtihlin Society. 



The temperature of the wire was taken as lOO^C, when the jackets 

 were heated, witli no blast on. The resistance as measured at room- 

 temperature and at 100° allowed the other temperatures to be calculated. 

 Since the temperatures varied with the blast, the method adopted of 

 representing the results by " apparent " velocities saves a large amount 

 of unnecessary calculation in plotting the curves. At a velocity close to 

 the critical velocity the temperature was found to be 90°0., and this is 

 consequently taken as the temperature of- the air at that point. The 

 curves No. 12 (Plate XXXIII.) represent the results of this experiment. 

 We can best express them in tabular form as below (Table II). Sutherland's 

 formula is again used to calculate the viscosity of the air. 



Table II. 



The values of the ratios in brackets in the last column, as well as the 

 values of the constant K calculated on the last line, indicate the excellent 

 agreement between theory and experiment. 



It appeared of interest to push tlie inquiry further and to examine 

 how, the law applied to other gases. In this investigation it is the ratio ij/p 

 that is important, and a gas is required for which this ratio differs from its 

 value for air. Witli the exception of coal-gas, tlie gases most easily obtained 

 in quantity are given in the accompanying table (Table III). 



- Taki.i; ITL 



