A New Precision Method for the Determination ofemfor Electrons

Abstract

The excellent focusing properties of crossed electric and magnetic fields have been utilized in the development of a new, precision method for the determination of $\frac{e}{m}$ for electrons. This method differs from previous methods in that the final equation for $\frac{e}{m}$ does not involve the velocity explicitly. Moreover, focusing criteria have been worked out which effectively eliminate any possible influence of electron energy upon the value of $\frac{e}{m}$. That this is a great source of error and uncertainty in other methods is shown by the great difference between the energy of the electrons before and after emergence from a slit. This effect is too great to arise from a contact potential difference but it can be attributed to direct electron bombardment of the slit and the subsequent formation of a surface charge on it. The magnitude of this charge is not constant but varies between 9 volts and 24 volts, depending upon the applied accelerating potential. The value of $\frac{e}{m}$ obtained with the present apparatus is, $\frac{e}{{m}_{0}}=(1.7571\ifmmode\pm\else\textpm\fi{}0.0013)\ifmmode\times\else\texttimes\fi{}{10}^{7}$ e.m.u., where 0.0013 is the probable error derived from a least squares solution of a set of observations for various electric and magnetic field intensities. Other sets differed from this by less than 1: 5000. The mechanical accuracy of the present cylindrical condenser sets the limit on the precision attainable with the present apparatus. However, the error due to this cause is less than the probable error stated above. The limitations of the present condenser can be reduced considerably through the use of a new condenser designed in accordance with kinematic principles. The method presented here for the production of magnetic fields of great uniformity and a new, precise cylindrical condenser would permit a determination of $\frac{e}{m}$ to be made with the method of crossed fields to within an accuracy of 1: 3000.