178 Prof. Sydney Young on the 



The percentage difference between the highest and lowest 

 ratios in the various comparisons — including those with fluor- 

 benzene — are given in the following Table : — 



Table XX. 





CH.OH 

 C 6 H 5 F 



C 2 H,OH 



0,H 7 OH 



C 2 H 5 OH 



C 3 H 7 OH 



3 H 7 OH 



C 6 H 5 F 



C 6 H 5 F 



CH 3 OH 



CH 5 OH 



2 H 5 OH 



Absolute temperatures at cor- \ 



10-5 



391-0 



3-5 



3-2 



12-9 

 157-0 



13-8 



602-0 



4-1 



30 



13-0 

 220-4 



14-6 



1035-0 



3-6 



3-4 



11-5 

 211-0 



3-0 



65-8 



1-2 



0-9 



4-0 

 395 



4-0 



136-2 



1-3 



1-8 



5-7 

 30-6 



1-5 



27-6 



1-2 



1-5 



6-3 

 12-9 



Vapour-pressures at corre- 1 

 sponding temperatures J 



Molecular volumes of liquid "1 

 at corresponding pressures J 



Molecular volumes of liquid at 1 

 corresponding temperatures J 



Molecular volumes of satu- 1 

 rated vapour at correspond- I 

 ing pressures j 



Molecular volumes of satu- ] 

 rated vapour at correspoud- > 

 ing temperatures J 



When the alcohols are compared with one another, the devia- 

 tions of the ratios from constancy are much smaller than when 

 fluorbenzene is taken as the standard substance, but they are 

 still far outside the limits of experimental error. It may, 

 perhaps, be said — as with the members of Group II. in the 

 comparisons with fluorbenzene — that the generalizations offer 

 a rough approximation to the truth. 



Here, again, the comparison of the molecular volumes of 

 saturated vapour at corresponding pressures is very much 

 more satisfactory than at corresponding temperatures. 



In the Philosophical Magazine for November 1890, p. 417, 

 it was pointed out by Prof. Orme Masson that the ratio of the 

 molecular volumes (in the liquid state) of any two members 

 of certain groups of nearly related carbon compounds, at their 

 boiling-points under equal pressure, is equal to the ratio of 

 those boiling-points, expressed on the absolute scale of tem- 



V T 

 perature, or y' = V' 



In the same number of the Philosophical Magazine (p. 423) 

 I showed that Masson's relation is a special case of a more 

 general one which should hold good if the generalizations of 



