of Deter milling the Elementary Electrical Charge. 215 



When the water in these jackets was two or three degrees 

 •above the temperature of the room, that is, when the lower 

 electrode was but a degree or two warmer than the air above 

 it, the cloud formed by a sudden expansion would always 

 •evaporate rapidly from the lower plate up, so that the 

 appearance was as though it were rising instead of falling. 

 It required not more than two seconds for the entire cloud 

 between the plates to evaporate from the lower plate to the 

 upper. If then the cloud could not last with a difference in 

 temperature of not more than 2° between itself and the lower 

 electrode, it seems certain that it would evaporate much 

 more quickly if the difference in temperature were 12° 

 or 15°. 



§3. The Viscosity of Air Saturated with the I r apours of Water 



and Alcohol , at 2(\° C. 



A large amount of work has been done within the last 

 decade upon the exact determination of the coefficients of 

 viscosity of dry air. According to three independent deter- 

 minations, all made with the utmost of care, the value of this 

 constant at 2G° 0., the temperature of the room in these 

 experiments, is 0*0001863. This number is the mean of 

 the values -000186:} obtained by Breiterbaeh *, -0001865 

 obtained by Schultze f, and "0001861 obtained by Fischer J. 

 Mr. Fred Allison made a careful determination for me in 

 the Ryerson Laboratory of tin 4 ratio between the viscosity of 

 dry air at 26° and that of air saturated with the vapours, 

 first of water, and then of alcohol. He obtained for the 

 times of outflow of given volumes through the same capillary 

 the numbers contained in the following table : — 



IstObs. 

 2nd Obs. 



3rd Obs. 



Dry Air. 



250-6 sec. 



Saturated Air 



(Water). 

 256-0 sec. 



Saturated Air 



(Alcohol). 

 252-6 sec. 



250-6 „ 



256-0 ., 



252-6 „ 



250-4 „ 



256-1 .. 



2^2'C „ 



Means 250*5 ., 256*1 ., 252-6 



Applying these corrections there results for the viscosity 

 of air saturated with water vapour at 2l!>° C, ?; = "0001904, 



* P. Breiterbaeh, Aim. der P/n/sik, v. p. 168 (1901). 



t H. S> c hultze, Ann. der Physih, v. p. J 57 (1901). 



X \V. J. Fischer, Phys. Rev. xxviii. p. 10-4 (1909). Fischer makes 

 seven determinations at a mean temperature of 24 *2 O. and obtains 

 *l = -0001852. This reduces to -0001861 at 26° C. 



