CONTEMPORARY ADVANCES IN PHYSICS 



129 



of Fig. 9 was obtained. Let the air-equivalent x of the total thickness 

 of the screens be varied, and let the average number of ions produced 

 per particle in the chamber be measured and plotted as function of x. 

 Recalling Fig. 7 and what was said in respect to it, the reader will see 

 that the resulting curve should have a peak wherever the integral 

 distribution-in-range curve has a step. This has been verified several 

 times, and there are cases in which these peaks have been taken as 



uJ 50 

 ffl 



2 



1.5 2.0 2.5 3.0 3.5 



AIR EQUIVALENT IN CENTIMETERS 



Fig. 11 — Integral distribution-in-range curve of the fragments resulting from the 

 bombardment of boron by protons. (Oliphant & Rutherford) 



clearer evidence for the existence of groups than the shape of the 

 integral curve itself (Fig. 10). Peaks may also appear in a curve of 

 which the ordinate is the total ionization produced in the very thin 

 chamber by all the fragments which enter it. 



Anyone at all acquainted with physical experiments will readily 

 suspect that the steps of actual integral cuives "ought to be" steeper 

 than they are. I mean : that he will form the hypothesis that perpen- 



