On Melting-point and Boiling-point. 173 



and to what extent we may expect information on capillary 

 action at a distance, will depend also on the question whether 

 the same phenomena by which chemical attractions are accom- 

 panied play a part also in capillary forces. If this be the 

 case, we shall not be justified in assuming the same force of 

 tension with the layer of liquid carbonic acid on the glass as 

 occurs with the carbonic acid not exposed to such molecular 

 attractions, any more than it would be permissible to take the 

 vapour-tension of pure water as that of an aqueous solution 

 of a salt. It may be mentioned that atmospheric air behaves 

 similarly towards smooth glass surfaces as carbonic acid does; 

 that, on the contrary, the accumulations of gas become sta- 

 tionary on carbon, and such substances as are traversed in all 

 directions by molecular interstices in a relatively short time. 



XXIII. On Melting-point and Boiling-point as related to Che- 

 mical Composition. By Edmund J. Mills, D.Sc, F.R.S* 



1. TN considering the relation between melting-point and 

 JL boiling-point as related to chemical composition, two 

 lines of investigation lie before us. We may conceive a sub- 

 stance to consist of parts definitely arranged and in definite 

 acts of motion, and then infer on mechanical principles under 

 what contingencies the crises referred to must necessarily 

 make their appearance. This is the internal method; but it 

 is so encumbered with hypotheses, in themselves most difficult 

 to verify, that a long time must elapse before any hope can 

 be entertained of its successful employment. The external 

 method, on the other hand, proceeding by way of tentative 

 effort and analogy, is more simple and direct, and must be 

 judged mainly by its results. It appears to me that the time 

 has arrived for an application of this latter method, to which 

 I have been specially induced to have recourse by the fact 

 that a number of series of organic bodies are now known 

 whose boiling-points and melting-points are sufficiently well 

 established to supply data for calculation. 



2. It has been long customary to regard what w r e term heat 

 as a mode of motion. In a memoir communicated to this 

 Magazine! I took occasion to show that matter as known to 

 us is also a mode of motion, differing from energy only by its 

 more concrete or isolable character. The suggestion naturally 

 presents itself, that heat may be a chemical reagent. This 

 suggestion, in fact, lies at the basis of the greater part of 



* Communicated by the Author. 



t u On the First Principles of Chemistry," Phil. Mag. [5] i. p. 1. 



