228 MODERN SCIENCE READER 



zelius' error eliminated itself, owing to the fact that the 

 molecules of the ordinary gaseous elements are made up of 

 equal numbers of atoms. Knowing the true atomic weights 

 of the ordinary gaseous elements, Berzelius was able to 

 obtain correct figures for many other elements, with the aid 

 of the principle of isomorphism and certain other principles 

 that need not be explained here. Thus, his figure for mer- 

 cury was 200, that for phosphorus 31, that for sulphur 32 

 figures practically identical with those accepted at pres- 

 ent. In 1827, however, Dumas invented his celebrated 

 method of determining vapor densities, and undertook to 

 apply Berzelius' principle of equal volumes to elements 

 which are not ordinarily gaseous. Finding that the vapor 

 of mercury is 101 times as heavy as an equal volume of 

 hydrogen, the vapor of phosphorus 62.8 times, and the 

 vapor of sulphur 96 times, as heavy as hydrogen, Dumas 

 concluded that the relative weights of their atoms must be, 

 respectively, 101, 62.8 and 96, and not 200, 31 and 32 as 

 Berzelius thought. The error of Berzelius' principle thus 

 emerged in the results of Dumas. But instead of rectifying 

 the error of his principle by introducing the concept of the 

 molecules of elements, Berzelius only concluded that the 

 principle was unreliable. The result was that chemists 

 began to disagree as to the true values of the atomic weights, 

 and many even abandoned the hope of ever knowing atomic 

 weights altogether, and decided to use nothing but equiva- 

 lents. These represented the weights of elements that were 

 capable of combining with, or of being replaced by, unit 

 weight of hydrogen. For example, Berzelius' view that an 

 atom of oxygen was 16 times as heavy as an atom of hydro- 

 gen was abandoned, and as hydrogen combined with 8 times 

 its weight of oxygen the latter was represented by its 

 equivalent 8. But the use of equivalents was not universal, 

 many chemists using systems in which the figures were 

 partly equivalents, partly atomic weights, and thus for 

 years great confusion reigned in chemical notation, the true 

 purpose of which is to avoid confusion by exhibiting the 



