|A curious example of the omissions that sometimes come to light during the investigation of common substanpes is found in the fact that from 1785 till 1894 no complete and exhaustive examination of the homogeneity of atmospheric nitrogen was attempted. In the former year Cavendish published his Experiments on Air, in which he investigated this point as minutely as his methods and apparatus would permit; but that he recognised his own limitations is evident from the following sentence written at the conclusion of his paper: "... if there is any part of the phlogisticated air of our atmosphere which differs from the rest, and cannot be reduced to nitrous acid, we may safely conclude that it is not more than 1/120th part of the whole. |
This observation remained unnoticed until Rayleigh, in the course of a very accurate determination of the density of nitrogen from various sources, observed that the density of atmospheric nitrogen, by whatever method obtained, was consistently higher than that of " chemical" nitrogen, obtained from ammonia, oxides of nitrogen, etc. This discrepancy (amounting to 1/120th of the whole) was much too great to be written off as experimental error, as may be seen from the data quoted below, in which the error of the determination does not exceed 0.0002 gm.
|Nitrogen from||Weight of Gas in large Globe.|
|Nitric oxide by red-hot iron (4 expts.)||2.30008 gms|
|Nitrous oxide by red-hot iron (2 expts.)||2.29904 gms|
|Ammonium nitrite; purified at red-heat (2 expts.)||2.29869 gms|
|Ammonium nitrite; purified cold||2.29870 gms|
|Urea and sodium hypobromite||2.2985|
|By means of red-hot copper (1892)||2.31026 gms|
|By means of red-hot iron (1893) (4 expts.)||2.31003 gms|
|By means of cold ferrous hydroxide (1894) (3 expts.)||2.31020 gms|
Mean of Series I. = 2.29927 gms.
Mean of Series II. = 2.31016 gms.
Difference = 0.01089 gm.
Rayleigh proved that the density of none of these specimens was altered by the action of the silent electrical discharge; and demonstrated, moreover, that the lightness of " chemical" nitrogen was not due to admixture with any known gas lighter than itself, such as hydrogen, ammonia, or water vapour.
The only other possible explanation of the high density of " atmospheric " nitrogen was the presence in it of some heavier constituent. It was at this point that a search in the literature revealed the astounding fact that the accepted view as to the homogeneity of the nitrogen of the air rested solely upon the experiments of Cavendish mentioned above. It was at once evident that, if his small residue were a gas having a density double that of nitrogen, its presence would more than suffice to account for the observed differences.
Cavendish had passed sparks from an. electrical machine through a mixture of air with excess of oxygen (dephlogisticated air) over mercury, and had afterwards absorbed the excess of oxygen and oxides of nitrogen by means of liver of sulphur. Repetition of the experiments with improved apparatus demonstrated that a residue was always obtained, and that its volume was proportional to the volume of air used. Spectroscopic examination proved conclusively that it was not nitrogen.
In collaboration with Ramsay, Rayleigh then obtained this new gas from air by absorbing the oxygen and nitrogen by the action of red-hot copper and magnesium respectively, and it was shown by spectroscopic examination to be identical with the gas obtained by Cavendish's method. That the new gas was not combined with the nitrogen was proved by its separation by atmolysis.
This new gas had a density of about 20 (H = 1), whence its molecular weight would be about 40. The high ratio of the specific heats pointed to the conclusion that its molecule was monatomic, i.e., that its atomic weight was also 40. To this new element the name argon (inactive) was given.
At first it did not seem as if there could be any place in the periodic classification for an element of atomic weight 40, but as time went on and other members of the group described- in this volume became known, chemists generally accepted the view, first suggested by Ramsay, that argon should be placed in a new group of the periodic system between bromine and potassium. By allocating this position to argon there arises an anomaly which is even now unexplained, as its atomic weight is greater than that of potassium which it precedes, but the greater part of the available evidence goes to justify the accepted view.