﻿and the Theory of Atomic Structure. 569 



the satellites was observed *. It is very difficult to draw 

 any definite conclusion from a comparison of Plate I. on the 

 one side and Plates II., III. on the other side, as these 

 plates have been taken with different tensions. But at any 

 rate it is very remarkable that no appreciable difference 

 between these plates could be observed. A great difference, 

 however, was observed between Plate IV. and the other 

 plates. 



On this plate only one line could be observed on the short 

 wave-length side of a. x . This line was extremely faint, and 

 seemed to have about the same wave-length as a 3 . As the 

 diagram-lines on this plate seemed not to be quite as strong 

 as on the other plates, a fifth plate was taken under the same 

 conditions as Plate I., but with a time of exposure of 

 \\ minute. On this plate the lines a 2 and a 2 were certainly 

 not stronger than on Plate IV., whereas the satellites could 

 very well be seen. 



The great difference between Plate IV. and the other plates 

 is easily explained if we assume that by an impact with one 

 single /3-particle more than one electron at the same time 

 may be removed from the atom. The energy connected 

 with the removing of an electron from the Li-level corre- 

 sponds for Ag to 3350 volts. Thus, if two electrons 

 should be removed at the same time from the L-shell, the 

 tension on the tube must be at least more than 6700 volts. 

 It is hardly probable that the maximum tension on the tube 

 should have this value in the case of Plate IV. ; therefore 

 the lines corresponding to a double ionization of the L-shell 

 could not appear on this plate. The one satellite which still 

 was present on Plate IV. may be due to atoms which have 

 lost at the same time one electron from the L-shell and one 

 electron from the M-shell. 



Further, these experiments indicate that it should be 

 possible not only to obtain information about the state of the 

 atom in which a certain satellite is emitted, but also to gain 

 more insight into the laws governing the disturbance of an 

 atom by an impact with a /3-particle. This was shown by 

 some photographs taken of the /3 2 satellites (/3 n and /3 J2 ) of 

 silver with different tensions. These lines are very broad 

 and diffuse. Under conditions differing about the same as 

 those of Plates I. and III. in the above table, no change in 



* From a comparison of Plates II. and III. it appears that the intensity 

 of the spectral lines is not proportional to the current through the tube. 

 There is no real difficulty in this, as we have to expect that the intensity 

 of the spectral lines is proportional to the current density in the focus- 

 spot on the anticathode rather than to the total current. 



