• ISi^,] Dr. Thomson^ First Principles of Chemistry, 147 



1 atom of base ; or, which comes to the same thing, 3 atoms of 

 acid unite with 2 atoms of base. In the subsesquisalts, 1^ atom 

 of the base unite with 1 atom of the acid ; for example, the se5- 

 quicolumbate of h^xytes is composed of 



3 atoms columbic acid, containing 3 atoms oxygei^, 

 ,fi 2 atoms barytes ^ • h^*- '^ '• 



fiere we see, that the oxygen of the acid is not al^muifipfe of thai 

 in the base. 



" When the acid contains 2 atoms of oxygen, and the base 1 

 atom, it is plain that the sesquisalts must all come under Berze- 

 lius' law ; because 1^ atom of acid will contain 3 atoms of oxy- 



fen, and 3 is, of course, a multiple of 1 ; but in acids containing 

 or 3 atoms of oxygen, the law of Berzelius cannot hold. 

 , " With respect to the subsesquisalts they will all come under 

 Ber2:elius' law when the acid happens to contain 3 atomsoxygen, 

 and the base only 1 atom ; but they will deviate from it when- 

 ever the acid c6ntains I'or 2 atoms of oxygen. *^"'^ n *> 



*' Upon the whole, though the subsalts and sesquisalts have 

 not been sufficiently investigated to enable us to decide upon 

 the point with perfect certainty ; yet from what we do know, 

 there appears sufficient evidence that Berzelius' rule cannot be 

 considered as a general chemical law ; and that we run the risk 

 of falling into most egregious mistakes, if we make use of such 

 a law in calculating th^ atomic weight and chemical constitu- 

 tion of the acids or bases. . I pointed out some remarkable 

 fexamples of this error when treating of uranium, to which it is 

 merely necessary to refer the reader. ^^ "^ " '^ "*" *^" 



In concluding our remarks, we may observe, thalt'we have 

 freely expressed our differences of opinion with the author ori 

 certain subjects 5 to this we are sure he will not object, morQ 

 especially as they are mostly matters of opinion, from which we 

 have withheld our assent. His method of experimenting ap- 

 pears to us liable to exception in very few cases ; the work must 

 form a part of every chemical library, and will be referred to as a 

 standard by those who wish to acquire information as to the 

 atomic weights of bodies, or a knowledge of the experimental 

 means oit^acm^^M^Zom';"^ ^aiBi-^o'blao^ 3l»' ■ 

 ; ! I ^^imomoo lo sliuq^Qni qiow abizoioiq y«^ ^c f?'"ar«r;i 

 lifi'i^i^avfid I a^^gil §0ib3O9Tr edi tii sbios 81 eri^ fo yajs'^o fjn^^ Uj 

 mii bsbfiuaiaq IBB has ^ gaoiiBnidraoo rloiia io 8elirr.*^Z9 le . >; 

 il^ti'.^t^dolo noUfl9.^Jta pds j^vsnedw bsiavooaib ec^ MJw -^-^x) avF.; 

 i . . . .aiLegdug sdi oi L^nijjt -^litduoiitB' 



*X9 ibiiiw rii floiififiidmoD snilsa lo bald wdloac ai m^d I' ** 

 ofidfii I ; i6i baiool sd oalfi ^sm guilsyiad lo wbI sdi oi ^aoiiq^t} 

 ^i^m^'^pt MAilq» odi ^d bed^w^ai^Bih syrsd I ib^riw aiba aacdi- 

 liiw atimi bias to m^in. frl ,^jfe£iak->g9* adi «i! ».\\^?;'wj»W?iA«?. ?r 



