358 Prof. J. S. TownsenJ on Ionization produced 



of 0"003()6254 is 273"03 ; thus Ave see that we might reason- 

 ably suppose the value of the freezing-point obtained from 

 hydrogen to be in error by as much as 0°*08 C. 



It might perhaps be urged as an objection to the above 

 reasoning that the difference in the increase of pressure 

 observed by M. Chappuis has not been proved to be due to 

 the influence of the Avails of the vessel : and that such a 

 difference might very well be caused by those unavoidable 

 errors of experiment which accompany even the most skilful 

 observations. Such an objection should not be disregarded ; 

 but it remains of importance to notice that a method, carried 

 out by a competent observer, which might have indicated 

 that the fifth significant figure remains unaffected by the 

 nature of the containing vessel, entirely fails to furnish such 

 a result. 



If there is really some physical cause at work which intro- 

 duces an error comparable with 0°'08 C. into the uncorrected 

 estimate of the freezing-point, then such a cause might affect 

 the second place of decimals very perceptibly in measurements 

 of differences of temperature between 0° C. and 100^ C. This 

 would happen if the disturbing agency were introduced some- 

 what irregularly between 0° C. and 100*^ C, as I have already 

 shown in the first part of this paper {he. cit. pp. 139-111). 

 It would thus appear that the errors introduced by the walls 

 of the thermometer may be more serious than the thermo- 

 dynamic correction to either gas-scale ; and as we do not 

 know for which of the two gases this error is more likely to 

 arise, we cannot really tell which of the two constant volume 

 thermometers most correctly measures differences of tem- 

 perature on the thermodynamic scale. 



XLY. On Ionization produced hy the Motion of Positive and 

 l^egative Jons. 



To the Editors of the Philosophical Magazine. 



GeXTLE3IEX, 



IX the July number of the ^Philosophical Magazine^ 

 Dr. Stark has published some remarks on the theory of 

 ionization by collision which refer to my work on that sub- 

 ject. It might be gathered from his remarks that the theories 

 which I have proposed and worked out represent the A'iews 

 previously held by other physicists, and that in the reference 

 1 made to his work I have greatly misrepresented the theories 

 which he has put forward. 



