S48 SIXTH REPORT — 1836. 



radius of activity, is reducible to a very converging series, the 

 terms of which depend on the density. In this manner the 

 following equation, containing two terms of the series, and ap- 

 plicable to solids not crystallized, to liquids, and to gases, is ob- 

 tained : 



p = a p'^ + b p §, 



in which p is the pressure, equal in all directions, p the density, 

 and a and b constant coefficients depending only on the nature 

 of the body and the quantity of heat. This equation applies 

 without any consideration of latent heat. Laplace in his specu- 

 lations arrived at an equation of the form p = a p'^. But as it 

 appeared from the phaenomenon of the propagation of sound that 

 for a given quantity of caloric, and consequently a constant 

 value of a, the pressure varied nearly as p ^, to account for this 

 difference the supposition of latent heat was introduced, which 

 is avoided by the more general formula consisting of two terms. 

 M. Poisson shows how his formula indicates tliat in solids and 

 liquids the mutual attraction of the molecules extends further 

 than their repulsion, and may be sensible at distances where the 

 latter has altogether disappeared. In this and other of his 

 writings M. Poisson considers the characteristic property of 

 fluids, or the condition of fluidity to be, that the molecules ar- 

 range themselves alike in all directions from any fixed point, 

 and with this property, that of pressing equally in all directions 

 to be intimately connected. Probably few will be disposed to 

 dissent from this view. But when he proceeds to assign as an 

 a priori reason for this property the perfect mobility of the par- 

 ticles, and considers this mobility to result from their spherical 

 form, or from their being so remote from each other that their 

 form has no sensible influence on their mutual action, we cannot 

 but feel that the cause assigned is not such that we can judge 

 of it by any previous knowledge or experience. It would be 

 more in conformity M'ith the rule Newton laid down of referring 

 cftects to ultimate mechanical causes, if the mobility of the par- 

 ticles of fluids and the property of similar distribution in all 

 directions about a given point, were ascribed to a particular ac- 

 tion of the molecular forces resulting from a particular laAV of 

 variation. If, for instance, the molecular repulsion from a single 

 particle, or rather the resultant of the repulsions from an aggre- 

 gate of particles, decreased very rapidly at a certain small di- 

 stance from the centre to which it is directed, and then after 

 becoming attractive, extended to a much larger distance without 

 ever becoming of large magnitude, it seems demonstrable that 



