638 Dr. A. M. Tyndall on the Critical 



of the approximations employed, the agreement was as good 

 as could be expected. 



This expression may also be tested by carrying out direct 

 measurements of the critical field at a point. 



Direct experimental measurements of the critical field at 

 a hemispherically ended point were first made by Cbattock 

 in 1891*, the method being that of measuring the mechanical 

 force on one end of the point, the other end being shielded 

 from induction. The accuracy of this work was, however, 

 vitiated by the fact that the needles which were used as 

 points had tapered sides ; these contributed to the pull which 

 was measured. Later measurements by Ohattock and the 

 author in 1910 1 on platinum wires at atmospheric pressure 

 led to the following empirical relation for a positive point 

 between the critical field (in E.S. units) and the radius of 

 the hemisphere a, 



Xa°' 45 = 85. 



The results for a positive point are alone given, because in 

 this case the value of the field may be derived from the pull 

 with greater certainty. In 1914 Zeleny % used this principle 

 to obtain the critical field at a liquid point, and obtained the 

 empirical relation 



Xa°*=56-9. 



For purpose of comparison these relationships between 

 X and a at atmospheric pressure are shown in curves 1, in 

 which the values of X in E.S. units are plotted as ordinates 

 with the values of a as abscissse. The dotted line (i) is the 

 theoretical curve, the full line (ii) is the experimental curve 

 of Ohattock and the author, and the broken line (iii) the 

 experimental curve of Zeleny for a liquid point. 



Edmunds points out a remarkable agreement between his 

 results and those of Zeleny at small values of a, values which, 

 however, are far smaller than any with which experiments 

 were made. It should be noticed, however, that the results 

 of Ohattock and the author lie considerably nearer to the 

 theoretical curve in the region of experiment, and the 

 agreement particularly with the smaller points is striking 

 considering the approximations necessary in the theoretical 

 development. At this pressure one would hardly expect the 



* Ohattock, Phil. Mag. vol. xxxii. p. 295. 



t Ohattock and Tyndall, Phil. Mag. vol. xx. p. 285 (1910). 



+ Zeleny, Phys. Rev. vol. iii. p. 69 (1914). 



