HISTOEY OF THE ATOMIC THEORY. 143 



had no idea of the fineness of her touch, and her absolute 

 refusal to make any allowance for inaccuracy in the construc- 

 tion of instruments. It was not even known that all bodies 

 could be compared by their weights ; why should they not as 

 well be known by their lightness? This plan had its fair 

 trial. By a curious circle of reasoning, it was decided, that 

 what we call oxygen, which makes an oxide, or calx of a 

 metal, was sulphur ; afterwards it was the principle of com- 

 bustion ; not such an erroneous idea. Now oxides or earths 

 were, of course, simple bodies ; when they were reduced to 

 metals in the fire, they combined with phlogiston ; they be- 

 came lighter. Therefore phlogiston had the principle of 

 lightness in it. The rule generally is, that we should begin 

 wrong. We now say the metal is simple, and by uniting with 

 oxygen, it becomes a compound, and is heavier. As the metal 

 burns and gives out heat, they said it gives out its phlogiston, 

 and loses its principle of lightness.* Stahl calls it sulphur. 

 This would scarcely come under our view had it not been the 

 cause of so many inquiries in the same direction, as to bring 

 about a result, derived from an analysis of all the oxides, and 

 a careful comparison of the weight of the metals, with the 

 weight of the oxides, whether produced by combustion or 

 oxidation in the fire, or precipitated from their acid solutions. 

 Even this strange theory tended in the right direction, although 

 at first threatening to take a mystical course. We could 

 scarcely have anticipated this difficulty of proving that all 

 bodies have weight and not lightness, but our forefathers 

 encountered it, and it may yet come to the struggle again, 

 renewed in a higher form, when we have to deal with those 

 physical existences, now called imponderables. 



I am not aware that any one went into the subject with 

 care before Bergman. He may be said to have introduced 

 modern analysis. Before him analyses were not superior io 



* p. 277. •• Traile de Soufre" Traduit de I'Allcmar.d de Stahl. Paris, 1767. 



