IONIZATION IN LIQUID WATER 169 



liquid and vapor. It is evident that there is very httle at the present 

 time in the way of a quantitative basis for theoretical study of the effects 

 of the liquid state. One may conclude that there is moderately strong 

 evidence for a significant decrease in /' in the liquid state, and that at 

 present the most likely estimate of the extent of this decrease gives about 

 20-30 per cent. In view of the important fact that the value of /' in the 

 vapor is so much greater than the relevant ionization potentials, such a 

 decrease in the liquid state must be regarded as fairly plausible. 



VIII. Ionization in Liquid Water 



Even if the anomalously great stopping power of liquid water should 

 ultimately prove spurious or greatly exaggerated, this would not evi- 

 dence that liquid water must behave like a gas in respect to other pene- 

 tration phenomena — for example, and especially, in respect to the 

 total ionization. And, if a small stopping-power anomaly should re- 

 main, this may imply, and provide a key to the understanding of, great 

 anomalies in other properties. 



The total ionization is most conveniently measured by the mean en- 

 ergy required to form an ion pair, commonly denoted by W. The de- 

 pendence of the ionization of a gas on molecular structure is familiar 

 from a great number of studies of gaseous systems. Indeed, Bragg, in 

 1913, in his statement of the additivity rule for molecular stopping 

 powers, pointed out that additivity does not hold for the total ioniza- 

 tion. Nevertheless, the patently naive statement that liquid water does 

 exhibit behavior similar to a gas in respect to total ionization is very com- 

 monly encountered in publications concerned with interpretations of 

 radiation effects on chemical and biological systems. Such an assump- 

 tion may lead to grave errors. 



As mentioned above, the ionization of liquid water by high-energy 

 radiations has never been directly studied; the natural conductivity is 

 too great to permit ion collection by any conventional means. Modern 

 techniques of rapid ion collection may offer some hope of collecting the 

 electrons produced, however, although the slowest electrons might still 

 interact too quickly with the medium. (The interaction of very slow 

 electrons with liquid water is by no means well understood, but recent 

 advances in the interpretation of the interaction of electrons with crys- 

 tals may provide a valuable basis for analyzing the corresponding prob- 

 lem with water.) The positive ions must interact extremely rapidly, 

 dissociating to OH + H"*", the latter "hydrating" in a period of the 

 order of magnitude of the dielectric relaxation time for water, that is. 



