86 PROGRESS OF SCIENCE IN THE CENTURY. 



It has also been noticed that in estimating these 

 numbers, hydrogen is taken as a unit, because it 

 enters into compounds in relatively the smallest 

 weight. The other elements and compounds are 

 tabulated according to the relative amounts of their 

 weights in forming compounds with hydrogen, or 

 with some other element whose equivalent with hy- 

 drogen has been already estimated. When one and 

 the same substance combines in several proportions 

 with another, as nitrogen, for instance, does with 

 oxygen, the smallest number according to which the 

 substance forms combinations is taken, the other 

 numbers relating to the same substance being found 

 to be exact multiples of the smaller. So far the Dal- 

 tonian rules. 



What Dalton began was continued by Berzelius, 

 Turner, and others; but we cannot enter into the 

 record of toil. Only two or three points of interest 

 can be indicated. The process of determining the 

 atomic weight of an element involves: (1) finding 

 the combining proportion or equivalent, and (2) 

 multiplying this by a factor (1 4) decided by the 

 measurement of the vapour density (Avogadro's 

 Law), or by finding the specific heat whose product 

 by the atomic weight is practically constant (Law of 

 Dulong and Petit), or by some other consideration. 



Berzelius in his determinations utilised Gay-Lus- 

 sac's law of volumes (1808) (that two gases always 

 combine in simple proportions by volume), the law 

 of Dulong and Petit (1819), and furthermore the 

 aid furnished by Mitscherlich's discovery of isomor- 

 phism (1820). " Mitscherlich established the fact 

 that the corresponding phosphates and arseniates, 

 with the same number of atoms of water, possess the 

 same crystalline form, so that even the secondary 



