582 BELL SYSTEM TECHNICAL JOURNAL 



Another great magnet of a peculiar and original construction, de- 

 veloped at the Cavendish, was then applied both to spectra displaying 

 line structure and to the spectrum of RaC, with notable success; 

 while the technique of determining distribution-in-range curves has 

 been improved to such an extent that it now almost rivals the magnets 

 in its capacity of distinguishing separate groups in an alpha-ray beam. 

 The theorem of the unique speed is therefore like so many another 

 theorem of physics; it was valid so long as the delicacy of the experi- 

 mental methods was not refined beyond a certain point, its validity 

 ceased as soon as that point was passed. 



To enter now into detail : 



The long-range particles were discovered by observing scintillations, 

 a method of singular delicacy and value, but having great dis- 

 advantages: all the observations being ocular, it is wearisome and 

 taxing, not every eye is capable of it, and there is no record left behind 

 except in the observer's memory or notes. Tracks of some of these 

 particles were later photographed in the Wilson chamber, but it 

 is a long research to procure even a few hundred of such photographs, 

 and yet even a few hundred are not sufficiently many for plotting a 

 really good distribution-in-range curve (the disagreements between 

 the early work with scintillations and the subsequent work with 

 Wilson chambers are rather serious). The best available curve is that 

 which Rutherford, Ward and Lewis obtained with the method of the 

 "differential ion-chamber," of which the principle is as follows: 



When an alpha-particle (or, for that matter, a proton) traverses a 

 sheet of matter, its ionizing power or ionization per-unit-length-of- 

 path — we may take one mm. as a convenient unit of air-equivalent — 

 varies in a characteristic way with the length of path which the particle 

 has yet to traverse before being stopped completely. The ionization- 

 curve at first is nearly horizontal, then rises to a pretty sharp maximum, 

 then falls rapidly to zero.^^ Suppose now that the particle traverses a 

 pair of shallow ionization-chambers, each containing a gas of which 

 the thickness amounts to not more than a few mm. of air-equivalent 

 (the same for both) and the two separated by a metal wall of negligible 

 air-equivalent. Suppose further that the metal wall is both the 

 negative electrode of the one chamber and the positive electrode of 

 the other, and that it is connected to the electrometer or other de- 

 tecting device. The charge which is perceived is then the difference 

 between the ionizations in the two chambers. If these are traversed 

 by a particle which is yet far from the end of its range, the difference 



"The curve for protons is exhibited in Fig. 7 of "The Nucleus, Second Part," 

 p. 124. 



