128 VOLUMES OF ATOMS IN THE GASEOUS STATE. 



obviously inapplicable to compound gases, as their volume has 

 a variable relation to that of their elements; and its adoption 

 would require grave alterations to be made in the atomic 

 weights of several of the elements themselves, to accommodate 

 these weights to the observed densities of the bodies in the 

 gaseous state. This will be seen from the following table, in 

 which the volume or fractional part of a volume placed against 

 each element always contains the same number of its presently 

 received atoms. These volumes are, therefore, the equivalent 

 volumes of the elements, and may be viewed as representing 

 the bulk of their atoms in the gaseous state, the combining vo- 

 lume of hydrogen being here taken as one. 



ATOMS 



Volume. Weight, 



Hydrogen . . . 1 . . 12.5 

 Nitrogen . . . 1 . . 177 

 Chlorine 1 442.6 



Bromine 1 978 



Iodine . . . 1 . .1580 



Oxygen . . . i .100 



Phosphorus \ 392 



Arsenic \ 940 



Sulphur . . . ^ . .201 

 Mercury 2 1266 



Of the first five bodies enumerated, equivalent weights oc- 

 cupy equal volumes. It was indeed the observation of this 

 equality between the atom and volume in these gases, that led 

 to the supposition of that relation being general. But the 

 atoms of oxygen, phosphorus and arsenic occupy only half 

 a volume, and would require to be doubled to fill the same 

 volume as the preceding class. The present atom of sul- 

 phur affords only one-sixth of a volume of vapour, and must, 

 therefore, be multiplied by six to afford a whole volume ; while 

 the atom of mercury supplies two volumes of vapour, and 

 would, therefore, require to be divided by two, or reduced 

 to one-half of its present number. Of these changes the re- 

 quired modification of the atoms of phosphorus, arsenic and 

 sulphur is incompatible with their chemical relations to other 

 bodies which are best established, and is quite inadmissible. 

 The densities of the vapours of these bodies must, therefore, 



