M. Stas's New Determination of the Atomic Weights. 143 



If, for instance, we suppose that nitrate of silver does not con- 

 tain its elements exactly in the proportion of their atomic weights, 

 even the best methods for its analysis or synthesis will give with 

 the same inaccuracy the relation of this weight. 



In developing his ideas on this point, M. Marignac refers to 

 his own experiments on monohydrated sulphuric acid. He has 

 shown that this compound, which was always considered very 

 stable, is really very unstable; it is only when it contains a 

 slight excess of water (1 per cent.) that it is quite stable, other- 

 wise the least increase of temperature causes it to give off va- 

 pours of anhydrous sulphuric acid. 



Who could say a priori that the sulphuret and nitrate of silver 

 are not capable of retaining at high temperatures a trace of sul- 

 phur or nitric acid, seeing that sulphuric acid can retain a slight 

 excess of water far above 100°? 



Causes of error of this kind, to which others might be added, 

 lead M. Marignac to doubt whether the differences between ex- 

 periment and Prout's law do not arise from the imperfection of 

 experimental methods. 



M. Marignac has another objection against Stas's conclusion. 

 "If the numbers of M. Stas do not absolutely coincide with 

 those of Prout, they approximate to such an extent that they in 

 fact cannot be considered accidental. What has been said of 

 Mariotte's and Gay-Lussac's laws may be applied to Prout's 

 laws. These laws, long considered absolute, were found to be in- 

 exact when the experiments were made with the accuracy attained 

 by M. Regnault and by M. Magnus. Nevertheless they will 

 always be considered as expressing natural laws, whether from 

 a practical point of view (for by their means the changes of 

 volume in gases may be calculated with sufficient accuracy) or 

 from the point of view of theory; for they probably express the 

 normal law of these changes of volume abstracted from some dis- 

 turbing influences, the. effects of which may some day be calcu- 

 lated. The same may be believed of Proufs law." 



M. Marignac terminates his remarks by an observation due 

 to Dumas : — The fundamental principle which led Prout to pro- 

 pound his law, that is to say, the idea of the unity of matter 

 and all the conceptions which have been based on this principle, 

 is quite independent of the magnitude of the unit which might 

 serve as common divisor of the atomic weights, Whether this 

 weight be that of an atom of hydrogen, of half or a quarter of 

 an atom, or whether it be any infinitely small fraction, all these 

 considerations would nevertheless retain the same degree of pro- 

 bability. The relations between the constitution merely become 

 somewhat less simple between the different elements. 



