xiv THE ATOMIC THEORY 301 



may or may not be the same as that which makes a definite 

 quantity of charcoal and of hydrogen dissolve in another 

 quantity of oxygen, to form the sugar of our plants ; but we 

 can see clearly, that the two sorts of attractions are so 

 different in their results, that it is impossible to confound 

 them." 



" Thus the dissolution of nitre in water is for me quite 

 distinct from that of nitrogen in oxygen which produces 

 nitric acid, or from that of nitric acid in potash which 

 produces saltpetre." 



" The dissolution of ammonia in water is in my eyes not 

 at all like that of hydrogen in nitrogen, which gives birth to 

 ammonia." 



" The dissolution of silver sulphide in antimony sulphide, 

 which produces red silver ore, 1 is not at all like that of 

 silver in sulphur which produces the sulphide of this metal." 



" Finally, the dissolution of antimony sulphide in the 

 lower oxide of antimony is, to my feeling, not at all like that 

 of antimony in sulphur, and the reason of these distinctions 

 seems to me to be evident, for the dissolution of sugar, of 

 nitre, of ammonia in water, can be obtained in a range of 

 proportions of which the extremes are infinitely separated ; 

 but the dissolution, in one another, of the elements of nitric 

 acid, of saltpetre, or of ammonia has only been permitted to 

 us under the rigorous condition of a proportion of the two 

 or more constituents " (Journ. de Physique, 1806, 63, 369). 



The distinction which Proust first made between a 

 DISSOLUTION and a COMBINATION (Journ. de Physique, 1806, 

 63, 370), and his description of the characteristics of a 

 CHEMICAL COMPOUND, have become regular features of the 

 chemical text-books of the hundred years which have 

 elapsed since his memoirs were published. 



Stas (1865) tests the accuracy of the law of fixed 

 proportions. The experiments of Proust were carried out 

 before the introduction into chemistry of the methods of 

 exact analysis which we owe to Berzelius ; they were there- 

 fore subject to errors varying from i per cent, to as much 

 1 Proust, Journ. de Physique, 1804, 59, 403. 



