APPENDIX. 443 



analysis, and reduced to a simple standard of comparison. 

 Its truth is independent of all theory. 



RELATIONS OF ATOMIC WEIGHT AND DENSITY. The com- 

 parative weight of equal measures or masses of different sub- 

 stances is not necessarily the same as the comparative weight 

 of their atoms. The mass of iron, for example, is heavier, 

 while the atom of iron is lighter than that of potassium. To 

 account for the fact, we must suppose the lighter atoms of 

 iron so closely arranged that they thus more than make up 

 by their larger number, for their inferior weight. In solids 

 generally, there is no correspondence between atomic weight 

 and specific gravity ; but in the case of many elements which 

 exist in the gaseous state, or are capable of assuming it, the 

 correspondence is complete, as shown in the following para- 

 graph. 



COMBINING MEASURES OR EQUIVALENT VOLUMES. A cubic 

 foot of nitrogen, weighs just fourteen times as much as the 

 same measure of hydrogen, and the relation of the atomic 

 weight is the same. In combining by atomic weights or 

 equivalents, they therefore combine in equal measures. 

 Chlorine, and the vapors of bromine, and iodine, belong 

 to the same class. Taking hydrogen 1 as the standard, 

 their combining measures are all 1. In the case of oxygen 

 the correspondence referred to does not exist. It is sixteen 

 times as Iieavy as hydrogen, while its atom weighs but eight 

 times as much ; here again we are under the necessity of 

 supposing a closer arrangement of the atoms. Those of 

 oxygen are not only heavier, but twice as closely approxi- 

 mated. Taking hydrogen as the standard, the combining 

 measure of oxygen is therefore . That of phosphorus arid 

 arsenic is the same, and that of sulphur^-. In the case of 

 most other substances the ratio is not so simple. 



In the comparison of combining measures it is more 

 customary to adopt oxygen as the standard of unity. The 



