32 NUCLEATION OF THE UNCONTAMINATED ATMOSPHERE. 



seen throughout the condensations in connection with the rain and 

 blurred coronas. Apart from this, the numbers obtained throughout 

 are quite out of keeping with any similarly observed ionizations. If, 

 however, free electrons appear only at the destruction or at the origin 

 of nuclei, the association of few ions with many nuclei at any time 

 subsequent to their origin is well accounted for, as already suggested 

 in the earlier paper. It is only while the nuclei are being produced 

 that the ionization and the nucleation must be of the same order ; for 

 the latter persists while the former vanishes at once. Finally the fol- 

 lowing results are implied, at least for the physical structure of air 

 saturated with water vapor : 



Air (dust-free) is inseparably intermixed with large and small nuclei, 

 whose number (to be reckoned in millions per cubic centimeter) rap- 

 idly increases as the order of molecular size is approached. There 

 seems to be no objection to looking upon these nuclei as a kind of 

 colloidal (air) molecule, particularly as such molecules are frequently 

 producible by the means (Bredig) which produce nuclei. If a large 

 number of free atoms is suddenty introduced into any region (and this 

 is probably what the radiation of the above kind virtually does) the 

 result is not merely a production of typical molecules but of a large 

 concomitant of graded nuclei. 



Practically any given nuclear status of air is a counterpart of the 

 intensity of the ionization of the medium in which the nucleation 

 originated, to the effect that the superior limit of size of the nuclei 

 and their number increase with the ionization. But there is no case 

 of ionization free from nucleation, be the exciting cause a mere radi- 

 ation as above, or ignition, combustion (including the low-tempera- 

 ture cases like phosphorus), or high potential discharge, or violent 

 comminution as in the case of water nuclei. 



EXPERIMENTS WITH DUST-FREE AIR ENERGIZED BY RADIUM. 



31. Effect of radium in hermetically sealed glass tubes. A modi- 

 fication was now introduced by inserting a small vial of thin glass 

 (walls 0.04 cm.) containing about o.oi gram of impure radium 

 (strength io,oooX) in the B end of the rectangular apparatus, Chap- 

 ter I, figure i. For clearer vision, the eye at the goniometer was 

 placed at about 35 cm. from the nearer glass plate of the fog chamber, 

 and the apertures so obtained, larger than the customary values by 

 about 10 per cent, appropriately corrected. The data of most of the 

 tables are reproduced in the chart containing figures 26-32. 



The filtered air was first examined without interference, as in table 

 1 8, in the complete absence of radium, and a similar test precedes all 

 the subsequent observations. The results for air are without perio- 



