190 CORRELATION OF PHYSICAL FOECES. 



have shown that the specific heats of certain substances; 

 when multiplied by their chemical equivalents, give a con 

 stant quantity as product or, in other words, that the com* 

 bining weights of such substances are those weights which 

 require equal accessions or abstractions of heat, equally to 

 raise or lower their temperature. To put the proposition 

 more in accordance with the view we have taken of the na 

 ture of heat : each body has a power of communicating or 

 receiving molecular repulsive power, exactly equal, weight 

 for weight, to its chemical or combining power. For in 

 stance, the equivalent of lead is 104, of zinc 33, or, in round 

 numbers, as 3 to 1 : these numbers are therefore inversely 

 the exponents of their chemical power, three times as much 

 lead as zinc being required to saturate the same quan 

 tity of an acid or substance combining with it ; but their 

 power of communicating or abstracting heat or repulsive 

 power is precisely the same, for three times as much lead as 

 zinc is required to produce the same amount of expansion or 

 contraction in a given quantity of a third substance, such as 

 water. 



Again, a greaf, number of bodies chemically combine in 

 equal volumes, i. e. in the ratios of their specific gravities ; 

 but the specific gravities represent the attractive powers of 

 the substance, or are the numerical exponents of the forces 

 tending to produce motion in masses of matter towards each 

 other ; while the chemical equivalents are the exponents of 

 the affinities or tendencies of the molecules of dissimilar sub 

 stances to combine, and saturate each other ; consequently, 

 here we have to some extent an equivalent relation between 

 these two modes of force gravitation and chemical attrac 

 tion. 



Were the above relations extended into an universal law, 

 we should have the same numerical expression for the three 

 forces of heat, gravity, and affinity ; and as electricity and 

 magnetism are quantitatively related to them, we should have 



