112 C. Bar us — Ionization of Water Nuclei. ■ 



dimensions (?\ = '60, r, = '32, I = 50 cm ) the result was -6 X 10 5 . 

 The data for complete saturation were (after correction)* about 

 3 X 10 r ', agreeing fairly well with the present cases where the 

 ionization is necessarily undersaturated, because of the influx 

 tube. 



S. Evanesence of the Charges of Water Nuclei. — The same 

 remarkable contrast between the initial charges and the subse- 

 quent charges on the nuclei will now be observed, if only a 

 little time is allowed to intervene, that was already pointed 

 out for phosphorus. Table 2 refers to nuclei produced in the 

 receiver A, figure 1, by allowing the accumulating water to 

 run off by the cock, Tc. They were then conveyed to the con- 

 denser C, about five or ten minutes later, by aid of the Mariotte 

 flask, M. The conduction is enormously reduced, though for 

 positive charges in the condenser it is greater than for negative 

 charges, showing that an excess of negative nuclei has persisted. 



Table II. — Ionization of water nuclei after 5-10 minutes. dV/dt = 2*5 

 lit/min. Capacity 72 cm . 

 Insulation Insulation Corrected 



Charge. before. a. after. a. 



+ at 

 30 volts -019 a= -009 -010 



— at 



30 volts -005 a = — -002 '007 



In Table 3 the data refer to nuclei which were left in the 

 vessel A for about one hour after they were produced. The 

 original ionization has all but vanished ; nevertheless there is 

 still an excess of negative nuclei, as shown by the greater 

 leakage of positive charges in the condenser. If negative 

 nuclei had been more rapidly precipitated in the intervening 

 hour in A, the reverse should have been the case, there should 

 have been an excess of positive nuclei, and negative charges in 

 the condenser should vanish more rapidly. 



Table III. — Ionization of water nuclei, 1 hour after production. dV/dt = 



2-5 lit/min. 

 Insulation Insulation Corrected 



Charge. before. a. after. a. 



+ at 

 20 volts -013 a = *009 .004 



— at 



20 volts '003 a ='001 *002 



Tested for coronas, after about one hour, 1,000-2,000 were 

 left, or about 1/3 of the original 5,000, a result quite out of 

 proportion with the loss of ionization. The electrical and con- 

 densational phenomena are thus again distinctly separated. 



