1826.] M. Berzelius's Hypothesis of the Atomic Theory. 35l 



not only occur wlien a solution contains merely the salts of the 

 above-mentioned groupe., but if a great number of salts be dis- 

 solved in the same'liqmd, they will form by prefei;ence so that 

 it may be said that salts belonging to the same order ol compo- 

 %Zl seek each other, as it were, to crystallize together, and 



" MnX'cire'ofte mixture of different salts of the same 

 formula it is observed that the crystalline forms are not sensibly 

 Sfected, fo such salts have, if not identical forms, at least forms 

 nf tbP same kind and very nearly allied with respect to their 

 'aigks a" wai; fir^obserLd by^M-Mitscherlich Hence we 

 cat imagine that at the moment of their becoming solid a certain 

 Tmber^orthe molecules of one salt may be substituted tor those 

 of another without occasioning any irregularity in the crystalli- 

 zation This identity of formulae is not only observed between 

 sats with the same acid, and having different bases of the same 

 . degree of oxidation, but also between salts of the same base or 

 bales of similar degrees of oxidation, that have different acids 

 of the same order of composition. Whence it results that not 

 llvsaltrof different basis have analogous forms, more or less 

 Sly allied, but also that salts having different acids are simi- 



'"?/MSr;s'"T:ublnces belonging to the same formula of 

 composition are also extremely frequent m nature both in simple 

 and multiple compounds ; but as we cannot in this case, any 

 Sore than in that of artificial salts, separate at will the^^^J?^' 

 dia e principles of these bodies, it is only by the considemtion 

 of the^r analyses that we can arrive at a knowledge of those 

 nixtmes Now by this consideration, we find in the simple 

 comPds, that sLh or such an oxide is replaced by such or 

 such another belonging to the same order of composition. Foi- 

 ?nstance?in stoneslccidentally coloured by a combined oxide 

 we Shatthe colouring principle - --\-f -^^^it^of 

 either that which serves as base, or that which plays the part ot 

 an acid Thus in the silicates with base of lime, or the bioxide 

 Scalctm, the colouring matter is frequently the b-xide o iron 

 and its quantity is such, that its oxygen is precisely equal to 

 that of t^e lime^hat is waging It follows that the smn o the 

 oxvP-en of the lime, plus that of the bioxide of iion, is exactly 

 ^q^ll to die quantity of oxygen, which the lime would con am 

 fn the pure colourless silicate. In silicates with base of dlu- 

 mina Jr he trioxide of aluminium, the colouring matters are 

 r tnoxide of iron, the trioxide of manganese, 8cc. sometimes 

 both ; and their quantity is such that their oxygen is equal to 

 that of the deficient alumina. . j- .^ .-.rin 



- In multiple compounds, one or other of the immediate prin- 



• Traite Elementaire, &.c. p. 244. 



