19"^ 



ON CHEMICAL ATTRACTION. 



*' which a substance can be resolved without decouiposition. 

 ** The integral parts are united by the force of aggregation, 

 " the constituent parts by chemical affinity *," Berthollet 

 describes the force of cohesion of a compound, as that by 

 which the integral parts are held together f. 

 This has' led (o It is the object of the following observations, to point out 

 errours in the ^y|,{^j. J ^-onceive to be an inaccuracy, in the opinion of com- 

 p:iieral theory <. ■• • i n t • i i-n 



of chemical pound integrant particles, and ol the attraction bj"^ which 

 attractioa. ^i^^y jj,,g supposed to be united; and especially to notice 

 some errours, which have been introduced into the general 

 theory of chemical attraction, by the adoption of this opi- 

 nion. 

 Opinion of It Is proper to premise, that the opinion itself of corn- 



compound in- ponnd integral p'afticles thtist be admiftfedHo be hvpotheti- 

 te<^rani parti- , ,,. . , , , , • • t 



f.les hypotheti- cal. VV e mix two substances together^ .and their particles 



^^^- unite in that manner, which constitutes cFiemical combina- 



tion ; but to say in what precise manner they unite, I apprer 

 hehd. to be ittipossibleij that they first collect together to 

 form parti.cles of a uew kind, andof a^superior order, which 

 unite by homogeneous attraction, is sorely not very mani- 

 fest. It isperhaps more probable, that chemical union is A 

 less complicated operation. If a number of heterogeneous 

 particles be mixed, together, they assume respectively that 

 and ajr-^retTa- situation, which their mutual attraction allots to them ; every 

 tiou effects of particle is probably attracted by every other ; and of this 

 cause*'"^ attraction, coaibinxition and aggregation are equally the e£- 



. . fects. ' 



Nor can the cohesion of a compound substance be attri- 

 buted more to the agency of houiogeneons, than of hetero- 

 geneous attraction; ,for if, in a compound, the particles be 

 drawn tpwards each other, it is of no importance whether 

 these particles be similar or dissimilar ; the same effect, in 

 point of cohesion, will be produced. 

 The contrary The account therefore usually given of the formation of 

 Drotmble*'^ '" *^^^ integrant particles of a compound, which unite by ho- 

 mogeneous attraction, or cohesion, is not only without proof, 

 but, as I humbly conceive, without probability. We shall 

 • however admit the opinion, and jyoceed to consider how it 



* Murray, ed, 2, vol. I, p, 63r f Researches, p. 38. 



Records 



