1825.] M. Berzelius's llypothesh of the Atomic Theory. 349 



of their author's hypothesis, but do they therefore prove its 

 accuracy? It seems to us to be reasoning in a circle. Ihe for- 

 mula are made for the hypothesis, and the hypothesis supports 

 the formute; but what arguments can be deduced from both 

 together to render it more probable that alumina and silica con- 

 tafnSatoms of oxygen, and lime, baryta, &c 2, than that each 

 of those substances is composed of 1 atom <;[ ^ase and 1 of oxy- 

 een'' In point of fact, both views come to the same thing tor 

 if we assume lime to contain 2 atoms of oxygen, the weight ot 

 the atom of the base (as stated above) must necessarily be dou- 

 bled, so that in reality whether we represent sulphate ot lime 



by the formula C S, or Ca S% we equally express a. triple ratio 

 oF the oxygen of the acid to that of the base, and so in all other 



cases • for Al S = Al s\ The adoption of the latter formula, 

 therefore, is as if one should expect to approach nearer to the 

 truTofa proportion by writing 999 : 666 instead of 3 : 2. 



But perhaps it may be argued that the hypothesis presents a 

 correct view of the analogies subsisting between all oxidated 

 bases and enables us to arrange them in separate orders accord- 

 hSto certain characteristic properties by whic^i '"^'^^1^1 

 one order may be distinguishect from those of another The 

 observations o^f M. MitscherUch seem to demonstrate that such 

 dSnct orders actually exist, and as the subject is both curious 

 andTmpo tant, and because we would not willing y suppress any 

 argument that may appear favourable to the hypothesis, we 

 Sdwell a little^ upVn it, although this paper has already 

 exceeded the limits we had originally prescribed tor it. 



We have another motive also for domg so. We know that one 

 of the first crystallographers of the present day* thinks favoura- 

 bly of M. Mitscherhch's theory, and our respect tor his opinion 

 would alone induce us to treat it with attention. It would give 

 us sreat pleasure if that gentleman would take up the subject, 

 anf correct any errors thit either ourselves or others may have 



fallen into concerning it. , ^ j wi, tK„ 



M. MitcherUchobl'erved that certain bases, saturated with the 

 same acid to the same degree, affect the same crystalline forms, 

 and that lime, magnesia, and the protoxides o iron and manga- 

 nese compose in this manner one class oi what he has called 

 ^somZhom bases ; whilst alumina and the peroxides of iron and 

 mangaVese form Another. He showed also that isomorphous 

 Shave the property of crystallizing together, concurring in 

 an uniform manner in the formation of one and the same crystal 

 M. MitscherUch supposed that the P"l^^^y j^^^^^ E^.^^^^^^^^^^ 

 i8omorphous bases kre really identical and that this dentity 

 necessarily results from a similarity in their atomic constitution. 



M. Levy. 



