56 HISTORY OF CHEMISTRY. [LECTURE IV. 



I have not been able to find any statement as to whether 

 two atoms of one element can combine with three atoms of 

 another, but it appears as if Dalton regarded this assumption 

 as untenable. Compounds which are most simply regarded in 

 this way, consist, according to him, of two composite atoms ; 

 he is obliged, of course, to make the assumption that atoms of 

 the higher orders are capable of combination with one another. 22 



I have pointed out, above, that Dalton's theory was in 

 agreement with the facts ; I shall now explain how, from his 

 experiments, he determined the atomic weights. In order that 

 he might be able to do this, the first thing necessary was to 

 settle the number of atoms in a compound. According to 

 Dalton, this number is to be sought for, in general, in the 

 simplest possible ratios. In estimating it, he starts from the 

 following principles : 23 



1. When only one compound of two elements is known, 

 this is composed of an atom of the second order. 



2. When two compounds are known, the one consists of an 

 atom of the second, and the other of an atom of the third 

 order. 



3. When three compounds are known, one atom of the 

 second and two atoms of the third order must, be assumed. 



How does Dalton now proceed to the determination of the 

 atomic weights, i.e., the. relative weights of the smallest par- 

 ticles? In the first place, he requires to choose a unit for 

 comparison. As unit he assumes hydrogen with the atomic 

 weight = i, and he refers all the other atomic weights to this. 

 To fix the others, he then applies his first principle. At that 

 time, only one compound each of oxygen and of nitrogen with 

 hydrogen was known, viz., water and ammonia respectively ; 

 therefore the atomic weights of oxygen and nitrogen can be de- 

 termined directly from the composition of these compounds. 

 In this way, Dalton finds them to be 5 and 7 respectively. He 

 checks the numbers so obtained by the proportions of the 



22 New System. I, 213-215 ; A.C. R. 2, 30-31. <23 New System. I, 



214; A.C.R. 2, 30. 



