ATOMIC WEIGHTS. If one should weigh out a number of volatile oxygen compounds in such quantities that the volumes occupied by them (under some given pressure and tempera ture) should be equal to one another as well as to the volume occupied by 2 parts by weight of hydrogen gas, he would find, by analysis. that some of these compounds contain 16 parts of oxy gen, others 32 parts of this element, still others 48 parts of it, etc. The number of parts of oxy gen in any one of the compounds would be either 16 or some multiple of this number; but in no ease less. A similar comparison of the com pounds of chlorine would show that they contain either 35.5 parts of this element, or some multi ple of 35.5, but in no case less. By extending in vestigations of this kind to volatile compounds in general, chemists have been able to demonstrate that for every element there is a characteristic number representing the smallest amount of that element capable of existing in a certain fixed quantity of a chemical compound, viz., in a quan in the state of vapor, the same volume as 2 parts by weight of hydrogen gas un der the same conditions of temperature and pres sure. This elm racteristic number is called the atomic weight of the element, and is generally implied in its chemical symbol; thus the symbol CI stands for 35.5 parts of chlorine, and the sym bol 0 for 16 parts of oxygen. The formula C10,, representing an explosive compound of chlo rine and oxygen, shows therefore that that com pound contains once 35.5 parts of chlorine and twice 16 parts of oxygen: or. briefly, one 'atom' of chlorine and two 'atoms' of oxygen. The great advantage of using atomic weights as units of measurement for the several elements consists mainly in the fact that it permits of representing by integral numbers (usually small ones) the composition of any chemical compound, and that it therefore greatly simplifies inquiry concerning the similarities and dissimilarities in the chemi cal nature of material substances.
In analyzing equal volumes of compounds, it was seen above that the amounts of the elements found would represent either their atomic weights or nmItiples of these. Water would yield 2 X 1 parts of hydrogen and 16 parts of oxygen; hydrochloric acid, 1 part of hydrogen and 35.5
parts of chlorine; caustic soda. 23 parts of the metal sodium (Na), 16 parts of oxygen, and 1 part of hydrogen; common salt, 23 parts of the metal sodium and 35.5 parts of chlorine, etc. If, in studying these compounds, we analyzed. not equal volumes, but equal weights—say, 100 parts by weight of each—we would find their composi tion in the form of percentages as follows: Wa ter, 11.11 per cent. of hydrogen and 88.89 per cent. of oxygen; hydrochloric acid. 2.74 per cent. of hydrogen and 97.26 per cent. of chlorine; cans tic soda, 57.5 per cent. of sodium. 40 per cent. of oxygen. and 2.5 per cent. of hydrogen; common salt, 39.32 per cent. of sodium and 60.68 per cent. of chlorine, etc. Of course, to say that water con tains 11.11 per cent. of hydrogen and 88.89 per cent. of oxygen is the same as to say that it con tains 2 parts of hydrogen and 16 parts of oxy gen. But by comparing the two sets of results just cited it is easy to see that, while the per centage figures have apparently nothing in com mon with one another, those derived from com paring equal volumes indicate plainly the simple laws according to which the chemical elements are distributed in their compounds.
The reason for choosing the volume occupied by two parts of hydrogen gas as the standard vol ume for the investigation of compounds is simply this, that by doing so no compound will be found to contain less than 1 part by weight of hydro gen; so that the atomic weight of hydrogen will be 1. I f, instead of two, one part of hydrogen gas were chosen as the standard, the atomic weight of hydrogen in its compounds would be found to he one-half, the atomic weight of oxygen would he 8, that of chlorine 17.75, etc. The rela tive magnitude of the atomic weights would, of course, be the same, hut the atomic weight of hy drogen would be one-half. Now, the atomic weight of hydrogen being less than that of any other element, chemists prefer to assign to it the value 1, thus making it the unit with which the atomic weights of the other elements are com pared. That by doing so they do not render their results any less general may be readily seen from the fact that the 'two parts by weight of hydro gen gas' taken may be 2 grains, 2 grams, 2 pounds, 2 anything.