49 



SIMPLE THEORY (GAUSSIAN) 



LANDAU THEORY 



htl EXPERIMENTAL POINTS 



AVERAGE TOTAL 

 ENERGY LOSS 



MAXIMUM ENERGY 

 LOSS IN SINGLE 

 COLLISION 



fH»4*Vf« rt l| . h ri M | m( L_ 

 50 60 76 



PULSE HEIGHT- ARBITRARY UNITS 



Fig. 1. (reproduced from Igo et. al. , Phys. Rev. 89, 879 (1953) ). Frequency 

 distribution of energy losses of 31.5 Mev protons traversing 3/4" proportional 

 counter. Histogram of experimental points shows standard deviation and chan- 

 nel widths. The theoretical Landau distribution is computed from Symon. The 

 dashed curve is a Gaussian distribution based on ion pair statistics. The ordi- 

 nate is counts per pulse height interval. 



and that the work I mentioned was the first real experimental test. 



FANO: Small energy losses enter in the theory only through their contribu- 

 tion to the stopping power; there is another correction which is called the 

 Blunck-Leisegang correction which you can put in if you want to, but does not 

 amount to much. 



POLLARD: The proportional counter is a very nasty test to use in this 

 region. You are just above the ease of detection against noise if you use 500 

 ion pairs. When you get down below that it is very hard to use. That means 

 you have to have high multiplication factors. Immediately there begins to be 

 the question of where the particle goes through the counter, and all the other 

 tnings that make it hard to work with begin to come in. So I would feel there 

 would be additional experimental uncertainty that I personally would worry about. 



FANO: I thought the proportional counter with a few kilovolts had been used 

 by lots of people. 



POLLARD: It is not easy. 



FANO: No, it is not easy, but there is all that work on tritium, especially 

 by the Glasgow people. 



BURTON: Was a number of proportional counters operating simultaneously 

 in this instance? 



