94 ROYAL SOCIETY OF CANADA 



then filled with freshly filtered air and allowed to stand, the ionization 

 of the enclosed air gradually increases for some days and ultimately 

 reaches a steady value. Owing to this eii'ect it was found necessary 

 in comparisons of the ionizing power of radiations existing in any two 

 localities or under any two conditions to make the observations, with 

 the ionization chamber under precisely the same conditions in both 

 circumstances. 



The different comparisons were made, therefore, either with a 

 receiver freshly cleaned and freshly filled with filtered air immediately 

 before taking the observations, or else with a receiver containing air 

 which had been allowed to remain in it till the steady state had been 

 reached. 



Further, as it was impossible always to obtain observations at 

 different times with the atmospheric conditions the same as to pressure 

 and temperature, it was assumed that the ionization obtained in all 

 localities would vary directly with the density of the air. in the receiver, 

 and in making any reductions which were necessary in order to reach 

 values which were comparable, assumption, which is amply warranted 

 by the measurements of McLennan and Burton ^ on the ionization of 

 air at different pressures, hns been adopted. 



In this connection it may be well to emphasize the extreme impor- 

 tance of taking every precaution in making observations such as are 

 described in the present paper, to secure absolute uniformity in the 

 conditions of the measuring receivers. With the different receivers 

 used in the present investigation it was found that when the cylinders 

 were thoroughly scoured and washed in the manner described above, 

 the conductivity of freshly filtered air admitted into the chamber was 

 always the same at any particular observing station, and thus, by always 

 working under these definite conditions, it was possible to obtain very 

 definite results. 



Numerous investigators in this field of research have experienced 

 considerable difficulty in arr'ving at concordant results, but if the pre- 

 cautions mentioned are taken, it is possible to obtain a thoroughly 

 reliable value for the conductivity of air enclosed in any metal receiver. 



(h) — Daily variation in the conduciiviiy of atmosplieric air. 



Among other investigators, Wood and Campbell,^ McKeon ^ and 

 Strong* have observed daily variations in the conductivity of air con- 

 fined in meftallic vessels, and, inasmuch as it was not practicable to 

 take observations in different localities at the same time of day in the 



' McLen-nan and Burton, Phys. Rev. C^), 1903. 



'Wood and Campbell, Phil. Mag.. Feb., 1907. 

 » McKeon, Phys. Rev., Nov., 1907. 

 * Strong, Phys. Zeit., Feb. 15, 1908. 



