276 Spontaneous Ionization of Air and other Gases. 



Month. 



1st Min. 



1 



1st Max. 



2nd Min. 



2nd Max. 



Feb 



J 4h. 



J 4h. 



6h. 

 3h. 



10 h. 12 h. 

 8 b. 6 b. 



14 b. 14 h. 

 14 h. 14 b. 



19 b. 21 b. 

 21b. -lb. 



March 



April 



4h. 



5h. 



9 b. 8 b. 



14 b. 17 b. 



21 b. 24 b. 



June 



J 4h. 



5h. 



9 b. Sb. 



16 b. 10|h. 



22 b. 22| h. 



August .... 



. 4h. 



4h. 



8 b. 6 b. 



15 h. 14 b. 



21 h. 23 b. 



Means.... 



. 4h. 



4ih. 



9 b. 8 b. 



14 b. 14 h. 



21b. 23 h. 



The close agreement of all these figures, and especially of 

 the means, seems to be sufficient to establish a connexion 

 between the two quantities involved. It may be noted that 

 the atmospheric potential curves represent observations in 

 fine weather only, while our observations were taken quite 

 irrespective of weather conditions. So. far as we could tell 

 the weather conditions were entirely without effect on the 

 ionization. 



An attempt was made to analyse these curves into Fourier 

 series without success. The curves were smoothed out and 

 account was taken of the first four components, but no con- 

 stant results were obtained. The apparatus has recently 

 been improved, and in future months it is hoped that obser- 

 vations may be taken more continuously. In this way 

 smoother curves may be obtained, and these may possibly 

 analyse. 



Briefly, then, the paper seeks to establish the following 

 points : — 



(1) The ionization in a closed vessel undergoes a per- 



manent increase for a considerable time after the 

 gas contained in it has been enclosed, but this in- 

 crease depends on the nature of the vessel, being 

 comparatively large for lead and tin vessels and 

 negligible for zinc vessels. 



(2) Superimposed on this variation of the ionization is a 



periodic variation having two maxima and two 

 minima each twenty-four hours. 



(3) In their main features the curves representing this 



periodic change are the same as the curves repre- 

 senting the variations of atmospheric potential, and 

 it is hoped in a later paper to establish a connexion 

 between the two variations. For a possible con- 

 nexion the reader may be referred to Richardson's 

 letter to ' Nature'*. 

 The authors welcome another opportunity for expressing 

 their indebtedness to Professor Thomson. 



* ' Nature,' April 26, 1906, p. 607. 



