54 Account of a numeneal Talle 



manner of relating every case as a different experiment, 

 above two tho-usand separate articles would be requind. 



Having been engaged in the collection of a few of the 

 principal facts relating to cheniistry and pharmacy, I was 

 induced to attempt the investigation of a series of these 

 numbers; and I have succeeded, not without some diffi- 

 culty, in obtaining such as appear to agree sufficiently well 

 with all the cases of double decompositions which are fully 

 established, the exceptions not exceeding twenty, out of 

 about twelv'e hundred cases enumerated by Fourcroy. The 

 same numbers agree in general with the order of simple 

 elective attractions, as usually laid down by chemical au- 

 thors ; but it was of so much less importance to accommo- 

 date them to these, that I have not been very solicitous to 

 avoid a few inconsistencies in this respect, especially" 'as 

 many of the bases of the calculation remain uncertain, and 

 as the common tables of simple elective atlractions'are cer- 

 tainly imperfect, if they are considered as indicating the 

 order of the independent attractive forces of the substances 

 concerned. Although it cannot be expected that these num- 

 bers should be accurate measures of the forces which they 

 represent, yet they may be supposed to be tolerable approxi- 

 mations to such measures; at least, if any two of them are 

 near'v in the true proportion, it is probable that the rest 

 cannot deviate verv far from it: thus, if the attractive force 

 of the phosphoric acid for potash is about eight tenths of 

 that of the sulfuric acid of barita, that of the phosphoric 

 acid for barita must be about nine tenths as great ; but they 

 are calculated only to agree with a certain number of phae- 

 nomena, and will probably require many alterations, as well 

 as additions, when all other similar phaenomena shall have 

 been accurately investigated. 



There is, however, a method of representing the facts, 

 which have served as the bases of the determination, inde- 

 pendently of any hypothesis, and without being liable to the 

 contingent necessity of any future alteration, in order to 

 make room for the introduction of the affections of other 

 si2)>3tances ; and this method enables us also to compare, 



upon 



