340 Dr. Thomson's Observations on the Weights of [Nor. 



Weight of an atom. 



5 Carbon 0-750 



6 Boron 0-875 b 



7 Silicon 1-000 c 



formed by the union of two volumes of hydrogen gas and one 

 volume of oxygen gas. I consider it, with Dalton, as a com- 

 pound of one atom of hydrogen and one atom of oxygen. Hence 

 the weight of an atom of hydrogen will depend upon its specific 

 gravity. Biot and Arrago found the specific gravity of hydrogen 

 gas 0-074; that of air being 1. Now if the specific gravity of 

 oxygen gas be 1*111, and if two volumes of hydrogen gas are 

 equivalent to one atom, it would follow that an atom of hydrogen 

 weighs 0-133. 1 have been at much pains in endeavouring to 

 determine the specific gravity of hydrogen gas, but never could 

 find it lower than 0-073, which differs but little from the deter- 

 mination of Biot and Arrago. But I do not think that abso- 

 lutely pure hydrogen gas has ever been weighed. It usually 

 contains traces of phosphorus, iron, zinc, &c. which must mate- 

 rially affect its weight. On that account I think that Dr. Prout's 

 method of determining the specific gravity of this gas from that 

 of ammoniacal gas is more likely to be correct. In ammonia we 

 are sure that the hydrogen is pure. It is likewise twice as 

 dense in ammonia as it is in hydrogen gas. This doubles the 

 chance of precision. When the weight of hydrogen gas is 

 deduced from ammonia, that of oxygen gas is found to be 16 

 times heavier. Hence the number in the table for the weight of 

 the atom. 



b The data upon which this number is founded are not quite 

 satisfactory. Common boracic acid, according to Davy's expe- 

 riments, is composed of acid 57 + 43 water. If it be a 

 compound of one atom acid + two atoms water, the weight of 

 an atom of boracic acid will be 2-998. From the analysis of 

 borate of ammonia by Berzelius, the weight of an atom of the 

 acid is 2-66. The mean of these two is 2-829. From the expe- 

 riments on the combustion of boron, there is reason to believe 

 that the acid contains at least two atoms of oxygen. If so, it 

 must be a compound of 2 oxygen + 0-829 boron. The number 

 0*875 was chosen in preference to this number ; because it is a 

 multiple of 0-125, which all the atoms seem to be. 



c Neither are the data for this number to be depended on. 

 Every thing shows us that silica in stony bodies acts the part of 

 an acid. Table-spar is a compound of 50 silica + 45 lime ; if it 

 be a bisilicate of lime, the weight of an atom of silica will be 

 2-015, for 45 : 50 :: 3*025 : 4*03. Again, nepheline is a com- 

 pound of 49 alumina -f 4G silica. Suppose it a silicate of 

 alumina, silica will weigh 1-994 ; for 49 : 46 :: 2-125 : 1*994. 

 Now the mean of these two numbers is 2*0045. Hence we may 



