176 BELL SYSTEM TECHNICAL JOURNAL 



tained in a certain manner.* On slud\ing this curve, it is found that 

 the critical speed at which 180° deflections arc most frequent is too low. 

 This indicates that an incident electron approaching an atom is 

 accelerated toward it, by virtue of the total charge of the electrons 

 on the spherical shell not quite compensating the nuclear charge; 

 the speed U which figures in the equations is therefore greater than 

 the measured speed with which the electrons are fired at the target. 

 (This interpretation also serves to explain the lobe observed on the 

 lowest-speed curves for magnesium, and suggests the reason for the 

 lobes of the curves for platinum.) 



The curves are satisfactorily explained, if we build the magnesium 

 atom in this manner: a nucleus of charge \2e, two electrons so near 

 it that the central charge is efTectually lOe, and a spherical shell of six 

 electrons with a radius of 1.28*10~' cm.; the other four electrons 

 much further out, perhaps dispersed and wandering through the 

 metal. The only arbitrary assumption made is that about the two 

 deep-seated electrons; the radius R of the shell and the number of 

 electrons upon it are prescribed by the curves, once that assumption 

 is made. If we assume three deep-seated electrons, R becomes 

 1.15"10~' cm. and the number of electrons in the shell drops to five. 

 The shell must be the L-level, and the deep-seated electrons con- 

 stitute the X-level.' 



The energ\- required to remove the loosest or outermost electrons 

 of the atom is generally determined, as is well enough known, by 

 smiting the atom with an electron instead of with one of the radiation 

 quanta used in extracting the inner electrons.'" Usually the quantity 

 measured is simply the energy which the striking electron must have, 

 in order to convert the atom or molecule into a positively-charged 

 ion; the negative charge removed from the atom is assumed with- 

 out proof to be a single electron. On the other hand, J. J. Thomson 



* Imagine an electron incident at angle on the target surface, and deflected 

 through angle <t> (in the plane of incidence; by an atom which it meets after pene- 

 trating a distance d in a straight line. If it continues in a straight line from the 

 (loint of deflection until it emerges, it travels a distance x = d (l-|-cos 8 'sec (\f/-6)), 

 where ^ = 7r — 6. This distance x will be the same for any two values ^i and ^2 of ^, 

 such that ^t 1 4-^2 = 29. Insofar as the number of deflected electrons emerging with 

 speed suflicient to reach the collector depends on x, it will be the same for both 

 values of if/. The curve representing the ratio of the number of electrons reaching 

 the collector, for two such angles, plotted versus U, is exempt from this correc- 

 tion, and can be directly compared with a theoretical curve. 



' Or we could assume that there were no deep-seated electrons, and give seven 

 electrons and a radius 1.54' 10~' cm. to the shell; but then we should have nothing 

 to serve as a A'-level. 



'"Generally the frecjucncy required to extract the outermost electron with a 

 (|uantuni lies in the Juost iiu-oiuenieiit region of the spectrum for praclicil work. 



