1821.] of the Atoms of Copper y Zinc y Iron J Manganese, &^c. 251 



water of crystallization, while they still continued completely 

 soluble in water, and of course had lost no sensible quantity of 

 their acid. After various trials I found that 9*25 grs. of this dry 

 salt, when dissolved in water and mixed with a solution of 13-25 

 grs. of dry chloride of barium, left a liquid which was neither 

 affected by muriate of barytes, nor sulphate of soda, and of 

 course contained neither sulphuric acid nor barytes ; therefore 

 9*25 grs. of dry sulphate of cobalt contain 5 grs. of sulphuric 

 acid ; consequently, the anhydrous sulphate must be a com- 

 pound of 



1 atom sulphuric acid = 5*0 



1 atom protoxide of cobalt == 4*25 



9-25 



Protoxide of cobalt then weighs 4*25. Hence an atom of 

 . cobalt must weigh 3-25. 



We see that cobalt and nickel have exactly the same weights 

 From the table at the beginning of this paper, we see that Ber- 

 zelius gives the weights of these two metals almost the same, y'vl. 



Nickel 3-69755 



Cobalt 3-69000 



But these atomic weights are considerably higher than the 

 truth. 



From the preceding experiments, I consider myself warranted 

 to conclude, that the atomic weights of the six metals which 

 have occupied our consideration in this paper are as follows : 



Oxygen = 1. Hydrogen = 1. 



Copper 8-0 64 



Zinc 4-25 34 



Iron 3-5 28 



Manganese . 3-5 28 



Nickel 3-25 26 



Cobalt 3-25 26 



Thus it appears that these six metals, like tlie 14 other bodies 

 previously subjected to examination, have atomic weights which 

 are multiples of hydrogen : so that if we reckon hydrogen by 

 unity, they are all whole numbers. If we inspect the table of atomic 

 weights when oxygen is unity, we may observe that the atomic 

 weight is either a whole number, or a whole number together 

 with one or other of the three following fractions, ^, {■, f. The 

 atomic weights would still continue multiples of the atom of 

 hydrogen (0-125), though they consisted of numbers terminating 

 in the fractions -i, -|, f , -J : but these fractions seem to be 

 excluded, at least as far as our observations have yet gone. The 

 consequence of this is, that when the atom of hydrogen is unity, 

 the atomic weights of all the other simple bodies hitherto exa- 



