A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 37 



Experiment shows that the currents are of about the same order when 

 charged water nuclei from an intense high-pressure jet and when charged 

 phosphorus nuclei are passing longitudinally through the condenser. But as 

 the number of water nuclei as tested by coronas are, even in the coadensation 

 chamber, not above lo' per cubic centimeter, while the number of phosphorus 

 nuclei may reach 10', the charge q in electrons per nucleus is large for water 

 nuclei and small for phosphoinis. Similarly one may expect the water nucleus 

 derived by a mechanical process to be larger than the initial phosphorus nu- 

 cleus derived chemically, so that k is larger in the latter case. 



Hence it is assumed that in case of water nuclei, k is negligible in com- 

 parison with qEU'/{r,_ — r,) and equation (2) becomes, if qU'=U 



-C dE/dt=i6.'j Vn,{eq) (i- e- -377 li>: + r,)UE/V (,r,-r,)y 



This equation, which fits the phenomena very well, predicts saturation as the 

 exponent is essentialh- dependent on E. 



On the other hand, in case of phosphorus nuclei, k is large in comparison 

 with. qEU'/(r,— r,), for here a single electron travels with many nuclei. The 

 exponential term in (2) vanishes or 



— C dE/dt = i6.'j Vn.fiq'EU'/k {r,-r,) 



which is virtually Ohm's law. 



An endeavor has thus been made to explain the two types of conduction in 

 question, the charged water nucleus type and the phosphorus nucleus type, by a 

 simple self-contained hypothesis. I have not, however, been able to complete 

 the numerical details to my satisfaction, and will therefore leave the subject 

 here without further comment. 



19. Charge and conduction. — The data have shown that positive as well as 

 negative charges are dissipated by water nuclei, immediately after they have 

 been produced, and that the ionization, if it may be so called, is quite of the 

 order of that of phosphorus, while the nucleation is much smaller. After being 

 stored but a few minutes, the nucleation loses all but a few per cent, of this 

 propertv of conduction, behaving in this respect again like phosphorus nuclei. 

 The number of nuclei does not appreciably vary in the same time. The char- 

 acter of the ionization (whether positive or negative nuclei are in excess) remains 

 intact so long as it can be observed. Hence the large initial and the eventual 

 very small conduction (a few per cent, of the original value) may be regarded 

 as two successive phases of a single continuous phenomenon, either of charge or 

 ionization or conduction. It seems to me therefore that it is not necessary to 

 distinguish the initial charges from the initial ionization. The experiment as a 

 whole shows an attenuation of the Lenard effect, continuously through infinite 

 time. One is at liberty to refer the conduction either to charged nuclei or to 

 ionized nuclei unless some distinctive definition is adopted. Both occurrences 

 are similarly reduced. The present case of river water is one in which there is 

 an excess of negative over positive nuclei. In other cases (pure water) the 

 reverse may be the case, or, again, there may be an absence of an excess of 



