71 



appreciably: if the dielectric constant is unity, one kilovolt per centimeter gives 

 about a ten per cent increase at room temperature. The data on recombination 

 agree fairly well with the theory. From the variation of the current with pres- 

 sure we can get some idea about the attachment distance. For air at 100 atmos- 

 pheres the order of magnitude would be about 500 A. 



Certain details in the data for fields up to one kilovolt per centimeter 

 seemed a bit puzzling at the time. I suspected that the collection field might 

 affect the motion of the electron before attachment. This would make a difficult 

 theoretical problem. Alternatively, I suspected multiple ionization. Now that 

 we know more about this, it might be worth while to take another look. 



Incidentally, I might have one or two suggestions, although they are hardly 

 of the kind to decide between 10 and 100 A, more likely to decide between 100 

 and 1000. One trick would be put in some scavengers deliberately. We can do 

 that if we add some cationic soap to the water, enough to form micelles. These 

 are probably spherical, typically of about 15 A radius, at potentials of the order 

 of 0.2 volts. You can load them with a great variety of dirt, such as any hydro- 

 carbons, dyes, small particles of graphite, whatever you think the electron 

 will stick to, and you can put in styrene. Many variations are possible. From 

 light scattering you can determine how many micelles you have and how big they 

 are. This is one suggestion. 



About one question which came up earlier: just what effect has the dielectric 

 relaxation? The relaxation time in water is of the order of 10" 1 * seconds. In 

 other water-like solvents it can be varied by many orders of magnitude; in gly- 

 cerol it is many microseconds. Chemically, of course, that differs a bit from 

 water, but you might compare it with methanol. For an intermediate range, 

 take glycol -- and you can mix the solvents. Certainly you can vary the relaxa- 

 tion times by many powers of ten without much change in the proportions of car- 

 bon, hydrogen, and oxygen, and you can see how that affects the results of ir- 

 radiation. 



LINSCHITZ: How about a comparison between the radiation chemistry of 

 liquid and gaseous water? 



MAGEE: It is an entirely different process in the gas phase. The radiation 

 chemistry of gaseous water has been done, hasn't it? 



ALLEN: A long time ago by Duane and Scheurer (22) and recently by Guenth- 

 er and Holzapfel (23). 



BURTON: Two years ago Dr. Onsager suggested an experiment which we 

 are still thinking about and Dewhurst, of our laboratory, plans to do. Dr. On- 

 sager suggested that in order to find out more about liquid water we should find 

 out more about liquid ammonia. I don't know how the relaxation times compare, 

 but we certainly have a different situation so far as electron trapping is con- 

 cerned, in liquid ammonia as compared with liquid water. The experimental 

 sequence would be to compare liquid ammonia and liquid water and then to com- 

 pare these other things. For example, take glycerol, methanol, and glycol. 

 Put them all, say, in water, so that you have approximately the same concentra- 

 tion of hydrogen, oxygen, and carbon, and then maybe you can find something 

 here which is systematic. 



GARRISON: Rollefson, of California, is working on ammonia. He has some 

 preliminary data. 



