118 On the Hexatomicity ofFerricum and Aluminium. 



rine and oxygenants in general is always to effect the breaking 

 up, never, so far as I know, the building up of compounds, save 

 of intermediate or diameric compounds such as green hydroki- 

 none. That the first action of chlorine upon iron should be to 

 produce a dichloride, Fe CI 2 , and that its continued action should 

 have the effect of uniting two molecules of this chloride into one 

 molecule of a non-diameric higher chloride according to the 

 equation, 2Fe Cl 2 -fCr 2 = Fe 2 Cl 6 or Ffe CI 6 *, seems most impro- 

 bable ; as does also the oxidation of two molecules of yellow 

 prussiate of potash, 2K 4 Fe"Cy 6 , into one molecule of red prus- 

 siate of potash, K 6 Ffe vi Cy 12 , instead of 2 K 3 Fe'" Cy 6 . 



8. It appears that the atoms of one class of elements unite 

 respectively with 1, 3, 5, and 7 atoms of chlorine. These ele- 

 ments I call perissads. They yield oxides having the general 

 formula? M 2 0, M 2 3 , M 2 5 , and M 2 7 . The atoms of another 

 class of elements unite with 2, 4, and 6 atoms of chlorine 

 respectively. These elements I call artiads. They form oxides 

 having the general formulas M 2 2 , M 2 4 , and M 2 6 , or per- 

 haps M O, M O 2 , and M O 3 . But some few metals which I call 

 periss-artiads, copper and manganese, for instance, both belonging 

 to the same group as iron, which indeed constitutes their inter- 

 mediate term, are indisputably both perissad and artiad. Thus 

 we have both cuprous chloride, Cu CI, and cupric chloride, CuCl 2 , 

 and both manganic and permanganic acids corresponding respec- 

 tively to a hexachloride of manganese, Mn CI 6 , and a heptachlo- 

 ride of manganese, Mn CF. The analogy of cuprous to silver 

 compounds, and of permanganates to perchlorates, seems to me 

 conclusive of the perissad character of copper and manganese in 

 some of their combinations. Now taking copper, iron, and man- 

 ganese, with their congeners, as forming between them a com- 

 plete series of chlorides, we should have produced by the action 

 of chlorine upon the typical metal, first M CI, represented by 

 cuprous chloride, sub-nick elous chloride, and sub-ferrous iodide ; 

 then M CI 2 , represented by the cupric, ferrous, and manganous 

 chlorides ; then M CI 3 , represented by the so-called sesquichlo- 

 rides of copper (?), iron, and manganese ; then M CI 4 , represented 

 by tetrachloride of manganese (?), pyrolusite, and iron pyrites; 

 then M CI 5 , corresponding to a well-defined pentoxide of chro- 

 mium ; then M CI 6 , represented by the ferric and manganic acids, 

 and by oxychloride of chromium ; and lastly M CF, represented 

 by heptachloride of manganese and permanganic acid, thus : — 

 Perissads. Artiads. 



MCI MCI 2 



MCI 3 MCI 4 



MCI 5 MCI 6 

 MCF 



