250 Chloride of Aluminium and its Analysis. 



Aluminium, - _. - - 46,7 

 Oxygen, _ _ _ 53.3 



100.0 



• If the atomic weight of ahiminium be calculated, supposing Ber- 

 zelius' numbers in the composition of Alumina, to have been acciden- 

 tally transposed as suggested above, we find it to be 0.876172 ; for 

 . C Oxygen. Aluminium. Oxygen. Aluminium. 

 ^l 100 : cc :: 53.3 : 46.7 whence a;=0.876172 



This number is a little more than one half that of Berzelius. 2 x= 

 1.752344, and Berzehus' number calculated from a sesquioxide com- 

 posed of 46.7 ox. and 53.3 al. is 1.71166. The correspondence 

 of the oxygen in the water of the hydrates of alumina with that in 

 the base, and a similar principle in reference to some other compounds 

 appear to have weighed with that distinguished chemist in confirm- 

 ing him in the composition of alumina, and the atomic weight of al- 

 uminium as given by him, but the compositions by transposing his 

 numbers in alumina still correspond with definite proportions. 

 Tlie composition of alumina from my experiments is, 

 Aluminium, - - 46.7859 



Oxygen, - - - 53.2141 



100.0000 



Alumina. Oxygen. 



.2975* : .1583118* a:==53.2141 



Alumina. Aluminium. 



.2975* : .139188* y=46.7859 

 The atomic weight of aluminium deduced from my experiments 

 independently is 0.87920118 for 



C Oxygen. Aluminium. Oxygen. Aluminium 

 ^l 1.00 : 'X :: 53.2141 : 46.7859 whence a;=.87920118 

 To ascertain whether alumina be an oxide or a sesquioxide, we 

 will deduce the atomic numbers for alumina, 1st, upon the supposi- 

 tion that the chloride analyzed was a neutral chloride, (vide k.) and 

 2d, that it was a sesquichloride, (vide n.) The correspondence of 

 the atomic numbers, will also demonstrate whether the chloride be 

 neutral in composition or a sesquichloride. 



* Vide table a. 



