676 BELL SYSTEM TECHNICAL JOURNAL 



In his latest experiments Lenard went down to electron-speeds so low 

 that he was working near, or even below, onset of ionization; but in- 

 stead of dwelling on these, I will take up the experiments which are 

 designed primarily for the study of slow electrons — experiments, the 

 first of which were made in Lenard's own laboratory by his associates, 

 H. F. Mayer and C. Ramsauer. They tried in somewhat different 

 ways to realize in practice the ideal scheme of apparatus which in the 

 last few pages I have been taking for granted. Before depicting their 

 devices, I revert for a moment to equation (4) above. 



Suppose that the collector is located at the distance x from the aper- 

 ture where the beam of Qq electrons per second enters the gas: the 

 number Q of electrons and the charge Qe reaching it in a second should 

 conform to the epuations: 



Q = (2oexp(- Nax), (8) 



\og{Qe) = \og{Q,e) - Nax. (9) 



Thus on plotting the logarithm of the collector-current against x one 

 should get a straight-line graph, and the slope of the line should give the 

 value of Na, therefore the value of a when the density of the gas is 

 known. It should suffice to slide the collector along the direction of the 

 beam, and plot the logarithm of the current against the distance 

 through which it has slid, measured from any arbitrary zero. Or al- 

 ternatively, one might keep the collector stationary at some fixed 

 distance Xo from the point of entry, and vary the density of the gas. 

 Plotting logarithm of current against N, one should get a straight-line 

 graph, and the slope of the line should give the value of Xijcr. It is 

 then unnecessary to worry about the possible presence of residual 

 vapors of unvarying density not recorded by the pressure-gauge, for 

 they would contribute only an additive term to N, not affecting the 

 slope of the line. 



It must be granted that the interpretation of the data is seldom 

 quite so simple. The quantity Qo may vary with the density of the 

 gas; some observers determine the total emission of electrons from the 

 source (or something which they take to be proportional thereto) for 

 each value of N separately, and plot the logarithm of the ratio of Qe to 

 this latter as function of N. The beam may diverge even in a vacuum, 

 or what the experimenter takes to be a vacuum; some measure Qe 

 twice for each value of x, once with a vacuum and once with gas, 

 and plot the logarithm of the ratio of the two as function of x. Mayer 

 varied both N and x and combined the results in the hope of thus 

 eliminating the effects both of the divergence of the beam in a vacuum, 

 and of residual gases. ^ 



' It is not always easy to make out from a paper just what procedure the observer 

 has followed, nor how far he has traced the curves of Q vs. x or Q vs. N. 



