96 



but that it decreases regularly by 4° until it becomes tlie 

 well known difference of 19°, as the following table will 

 show — 



Boiling-points. 



It appeared to me of interest to compare the boiling 

 points of other normal compounds, selecting of course those 

 only of which the boiling points have been carefully deter- 

 mined and corrected for pressure and expansion of the 

 mercurial column of the thermometer above the vapour. 

 The result of this investigation is that in most of the other 

 series the difference between the boiling points also steadily 

 decreases by about 2° ; but I am not in a position to state 

 whether this decrease ceases when the difference becomes 

 19°, as we do not yet know a sufficient number of compounds. 



