of Deter mining the Elementary Electrical Charge. 227 



that the density of the particles constituting an emulsion o£ 

 gum gamboge is the same as the density of gum gamboge in 

 mass. Fourth, from the experimental stand point, the difficulty 

 in measuring accurately the small differences in level which 

 he is obliged to use (less than *1 mm.), and in counting 

 correctly the number of particles in a given level, are such 

 as to render easily intelligible such lack of consistency as is 

 shown in Perrin's different determinations, which give values 

 of N (the number of molecules in 1 gram-molecule) varying 

 from 5-4xl0 23 to7'lxl0 23 . 



Broglie's result* is 4*5 x 10~ 10 . It is obtained by measuring 

 the velocities in an electrical field of the charged particles of 

 tobacco smoke, the mean radius of these particles being 

 obtained from kinetoscopic records of the mean displacement 

 which they undergo in a given time because of th» j ir Brown- 

 ian movements. The method involves the assumption of 

 Perrin's value of N, and tor both theoretical and experimental 

 reasons will scarcely claim to represent an accurate deter- 

 mination of e. 



Moreau's t measurement of the charge carried by the ions 

 in flames depends upon Perrin's determination of e. Hence, 

 Moreau's final result, viz. 4'3x 10~ 10 must involve any errors 

 contained in Perrin's work. 



If then, equal weights he assigned to all the recent deter- 

 minations of e by methods which seem least open to question, 

 the most probable value of e at present obtainable is that 

 given below : — 



Planck 4"69xlO- M 



Rutherford & Geiger 4*65 „ 



R gener 4*79 „ 



Begeman 4*67 „ 



Millikan 4*b\5 .. 



Mean 4*69 



a 





§ 9. Conclusion. 



1. The temperature of the fog chamber a very few seconds 

 after expansion in the form of apparatus shown in the figure 

 (p. 212) is essentially the temperature of the room. 



2. The balanced drop method herewith presented for the 

 determination of e involves an experimental error of not more 

 than 2 per cent., and is entirely free from all theoretical un- 

 certainties except such as .re incident to the application of 



* Broglie, Le Radium, vol. vi. p. 208, Juillet, 1909. 

 t Moreau, C. It. vol. cxlviii. p. 1255, 10 Mai, 1909. 

 Q. 2 



