SUBSTANCES WITHOUT CHEMISTRY 243
the gases which we now call oxygen and nitrogen, and he said that if what remained hehind was not nitrogen its volnme was not more than one one htmdred and twentieth that of the nitrogen. He did not carry his experiments any farther, a course of action not to be wondered at when we learn that he and his assistant in carrying the investigation to this point had already kept turning the handle of the frictional electric machine for the not inconsiderable period of three weeks. Most people would count that a long enough time to keep the nose to the grindstone. It is now possible by simply switching on an electric current to obtain the same result in a much shorter time; but, although for the last fifty years or more Cavendishes experiment could have been re- peated with ease, no one thought of attempting it; and text-books in chemistry continued calmly to assert that air consists on the average of 20.96 per cent, of oxygen and 79.04 per cent, of nitrogen.
Attention was drawn to the subject owing to an investigation car- ried on by Lord Eayleigh, with no thought of the wonderful outcome of his work, which was started for an entirely different purpose. He set out to determine the densities of various gases, in the first place the relative densities of oxygen and hydrogen, at which he worked from time to time during ten years 1882-1892. After that he determined the density of oxygen and nitrogen and of air with a view to determin- ing the percentage of the two gases in the atmosphere. Lord Bayleigh prepared the gases in different ways. Oxygen was prepared in three different ways, but, no matter in what way it was prepared, its density was always the same. Such was not the case with nitrogen, how- ever. Of the nitrogen obtained from five different chemical compounds which Lord Bayleigh employed, the amount contained in the globe that he used weighed on the average 2.29900 grams, while the nitrogen ob- tained in three ways from the air weighed on the average 2.31049 grams. Translated into English measures, this means that approxi- mately three pints of nitrogen got from air weighed about one seventh of a grain more than the same volume of nitrogen from the chemical compounds.
The ratio between the two weights was not far different from that between an ordinary letter, before and after the stamp is put upon it; but the actual difference in weight was only about one tenth the weight of a postage stamp. But the greatest difference in the weight of the nitrogen obtained from the different chemical com- pounds was not more than one seventieth the weight of a stamp, while in most of the experiments the difference was much less. It was evident then that the difference noticed between atmospheric nitrogen and what might be called chemical nitrogen could not be due to Bayleigh^s errors in weighing.
Bayleigh at first inclined to the opinion that atmospheric nitrogen
�� �
