5f> ATOMIC WEIGHTS. 



refined mode of analysis, will prove hereafter 

 that several of the present atomic numbers do 

 not represent the true weights of the atoms. But 

 it is also probable, that in nearly all cases where 

 this is not so, they represent either multiples, or 

 sub-multiples of them. 



There are other exceptions to the law, that all 

 bodies combine chemically in fixed or definite 

 proportions by weight. Many of the metals 

 unite with each other in all proportions when 

 converted into the liquid state by heat, or elec- 

 tricity, forming alloys. 



Water combines with alcohol, and with the 

 strong acids in unlimited proportions ; and it will 

 be shown further* on, that the ratios in which 

 sugar, numerous salts, oily, resinous, and gela- 

 tinous substances, unite with water by chemical 

 solution, are determined by the temperature of 

 the water, solubility of the bodies, &c. But in 

 all such cases, the affinity is comparatively weak, 

 and the changes of properties consequent on 

 combination inconsiderable. 



Theory of Volumes. 



In the year 1805, Gay-Lussac and Humboldt 

 proved by numerous experiments, that water is 

 always composed of 2 measures by volume of 

 hydrogen to 1 of oxygen, as had been previously 

 shewn by Higgins. By following up the inves- 

 tigation, Gay-Lussac found, that many other 



