g£4 Parallel of Rome de Visit s and 



pose them are equal in number and proportion. We 

 will next suppose the crystals of this substance can be 

 divided by sections parallel to six planes. In that supposi- 

 tion, nineteen or twenty different speeies of parallelopipe- 

 dons can be produced. Among these species some will 

 be similar, others not ; but none of the species will be ex- 

 actly parallel to each other. We will proeced on two si- 

 milar crystals of the same substance and equal in solidity ; 

 dividing the first into one species, the other into a different 

 species, of parallelopipedons, equal in solidity but not in 

 surfece; and let the division of eaeh be pushed to its last 

 term. But as we are come by smooth sections to parallelo- 

 pipedous of different species, those sections have also pro- 

 duced their differences : but by supposition these parallelo- 

 pipedons are the result of the last possible term of division 

 without destroying the chemical composition, and being 

 equal in solidity, though not in surface, they cannot con- 

 tain each other; therefore if their differences are not inte- 

 grant parts or' both, these differences must cease to be ho- 

 mogeneous, and we come to a sort of chemical decomposi- 

 tion. It is true we cannot execute this excessive division, 

 but we can form a very correct idea of it. If the little pa- 

 rallelopipedons contain two sorts of elements, their differ- 

 ences will also, but in different proportions; and, sir, if 

 you will turn to Bertbollet's Researches on the Laws of Af* 

 jinities, you will see him in all his experiments proving, that 

 "however perfectly a chemical decomposition may have been 

 made, the results will always contain a certain portion of 

 those substances from which it was the object of the ope- 

 ration to separate them. If these reflections, sir, are well 

 grounded, do they not give us hopes, and perhaps show 

 the possibility, of descending from the integrant particles 

 to the constituent particles ? This second research is of the 

 same nature as the first. It is more than probable that the 

 constituent particles themselves are divisible, having no de- 

 termined ligure, but are aggregations, subject to the same 

 laws as the integrant particles. The object of the natural 

 philosopher is not to discover the forms of the ultimate par- 

 ticles, but to determine their respective positions ; which, 

 if ever they could be determined in the integrant particles 

 and their component parts, the grand problem of chemical 

 affinities would be fully solved ; and should such ever be 

 the case, to the Abbe IJaiiy's theory would be due the merit. 

 The Encyclopaedia Dritamuca, under the article Chemistry, 

 in the Supplement, p. 396*, says : 



" This theory, to say no more of ^ is, in point of in- 



gcnuitVj 



