XV. ELECTRONS, NEUTRONS, AND ALPHA PARTICLES 549 



d W 



density of gas used 32.5 



where Je is the observed ionization in electrostatic units per cubic 

 centimeter. When the ionization chamber is filled by a gas having 

 the same atomic composition as the walls, the only limitation on its 

 size is that the neutron intensity should be sensibly constant through- 

 out the volume. Neutron ionization measurements have been made 

 in chambers containing gases approximating the walls in atomic 

 composition bj' Gray and Read (93) and Aebersold and Anslow (91). 



In a private communication, Prof. Failla informed the author that II. H. 

 Rossi, working in his laboratory, has found that tissue having the composition 

 H = 9.75%, O = 70.98%, C = 15.71%, and N = 3.55% may conveniently be 

 represented by walls composed of: gelatin, 20.15%; glycerin, 5.18%; water, 

 66.23%; sugar, 8.43%; and a gas mixture having the following composition 

 (partial pressures in cm. of Hg): methane, 29.22 cm.; hydrogen, 50.12 cm.; 

 ox5^gen, 48.96 cm.; aii-, 3.53 cm. The gas mixture, however, is explosive. 



It is always more convenient, however, if air can be used in the 

 chamber instead of a special gas filling. If the results are to be 

 readily and quantitatively interpreted, two conditions must be ful- 

 filled. The dimensions of the chamber and the air pressure must be 

 such that the particles generated in the walls lose a small fraction of 

 their energy in crossing the chamber, and it must be possible to esti- 

 mate the value of p,„, the mass stopping power of the walls relative 

 to air. 



A knowledge of the quality of the neutron radiation being meas- 

 lued and reference to tables of ranges (Table IV and Fig. 13) would 

 generally'' indicate whether the former condition is likely to 'be ful- 

 filled, but it can also be tested experimentally, for, if it is fulfilled, 

 the observed ionization will be strictly proportional to pressure for 

 all pressures up to the working pressure. 



Many interesting tests of this kind were made by Aebersold and 

 Anslow with a variety of gas pressures and wall materials. Provided 

 the chamber wall is made of a material about as rich in hydrogen as 

 tissue, the condition is well fulfilled for gas volumes of the order of 1 

 cc. and for netitrons having upward of 2 m.e.v. energy. 



The mean stopping power of the wall material, p„, varies with the 

 energy of the recoil protons and therefore with the neutron energy. 

 Data given by Aebersold and Anslow and by Gray {S3), reproduced 

 in Table IV, enable p,„ to be estimated with sufficient accuracy for 

 neutrons of energy between 0.1 and 15 m.e.v. 



