356 PROCEEDINGS OF THE AMERICAN ACADEMY § 2 ' 



lished, and this accounts, undoubtedly, for the fact that the 

 laws of cohesion have hitherto escaped a systematic in- ■ 

 vestigation. These considerations will be met later on. ; 

 It would seem, moreover, at first sight, that a multitude of 

 solutions might be possible, each equally capable of ex- 

 plaining the various phenomena by means of a sufficient 

 number of arbitrary assumptions — a number continually ' 

 increasing, as new facts are discovered, until, like the cor- I 

 puscular theory of light, the whole structure would fall as j 

 if by its own weight, under the blows of some simpler 

 hypothesis. j 



Such is not the nature of the present investigation, j 

 which adds nothing to the assumptions already required 

 by the molecular theory save one, which is shown to be 

 the necessary consequence of various phenomena. 



The object of this paper will be to investigate more 

 fully the nature of the law of cohesion and to apply it 

 to the solution of the relations existins: between certain ! 

 physical constants well known in thermod3'namics. It , 

 may be stated that these have been generally viewed, 

 hitherto, from a purely empirical standpoint. 



f 



] 



§ 2. There are evidently three distinct pressures which 

 exist in a liquid or solid: first, the external pressure (/^) ; , 

 second, the kinetic pressure (/^'), due to the vibration of j 

 the molecules; and third, the cohesive pressure (/^"), due j 

 to the attraction of these molecules for each other. If j 

 these pressures are all expressed in the same unit (dynes 1 

 per square centimetre), and are reckoned positively out- 

 wards, we have for a body in equilibrium the expression, 



direction is inwards. % 



P+ P' -\- P" = o, 

 the line drawn over P and P" indicating that their true 



