The behaviour of argon at low temperatures was investigated by K.S. Olszewski (Phil. Trans., 1895, p. 253). The following results are extracted from the table given by him:—
| Name. | Critical Temperature, Cent. | Critical Pressure, Atmos. | Boiling Point Cent. | Freezing Point, Cent. |
| Nitrogen | −146.0 | 35.0 | −194.4 | −214.0 |
| Argon | −121.0 | 50.6 | −187.0 | −189.6 |
| Oxygen | −118.8 | 50.8 | −182.7 | ? |
The smallness of the interval between the boiling and freezing points is noteworthy.
From the manner of its preparation it was clear at an early stage that argon would not combine with magnesium or calcium at a red heat, nor under the influence of the electric discharge with oxygen, hydrogen or nitrogen. Numerous other, attempts to induce combination also failed. Nor does it appear that any well-defined compound of argon has yet been prepared. It was found, however, by M.P.E. Berthelot that under the influence of the silent electric discharge, a mixture of benzene vapour and argon underwent contraction, with formation of a gummy product from which the argon could be recovered.
The facts detailed in the original memoir led to the conclusion that argon was an element or a mixture of elements, but the question between these alternatives was left open. The behaviour on liquefaction, however, seemed to prove that in the latter case either the proportion of the subordinate constituents was small, or else that the various constituents were but little contrasted. An attempt, somewhat later, by Ramsay and J. Norman Collie to separate argon by diffusion into two parts, which should have different densities or refractivities, led to no distinct effect. More recently Ramsay and M.W. Travers have obtained evidence of the existence in the atmosphere of three new gases, besides helium, to which have been assigned the names of neon, krypton and xenon. These gases agree with argon in respect of the ratio of the specific heats and in being non-oxidizable under the electric spark. As originally defined, argon included small proportions of these gases, but it is now preferable to limit the name to the principal constituent and to regard the newer gases as “companions of argon.” The physical constants associated with the name will scarcely be changed, since the proportion of the “companions” is so small. Sir William Ramsay considers that probably the volume of all of them taken together does not exceed 1⁄400th part of that of the argon. The physical properties of these gases are given in the following table (Proc. Roy. Soc. lxvii. p. 331, 1900):—
| Helium. | Neon. | Argon. | Krypton. | Xenon. | |
| Refractivities (air = 1) | .1238 | .2345 | .968 | 1.449 | 2.364 |
| Densities (O = 16) | 1.98 | 9.97 | 19.96 | 40.88 | 64 |
| Boiling points at 760 mm. | c. 6°[1] abs. | ? | 86.9° abs. | 121.33° abs. | 163.9° abs. |
| Critical temperatures | ? | below 68° abs. | 155.6° abs. | 210.5° abs. | 287.7° abs. |
| Critical pressures | ? | ? | 40.2 metres. | 41.24 metres. | 43.5 metres. |
| Weight of 1 c.c. of liquid | ? | ? | 1.212 gm. | 2.155 gm. | 3.52 gm. |
The glow obtained in vacuum tubes is highly characteristic, whether as seen directly or as analysed by the spectroscope.
Now that liquid air is available in many laboratories, it forms an advantageous starting-point in the preparation of argon. Being less volatile than nitrogen, argon accumulates relatively as liquid air evaporates. That the proportion of oxygen increases at the same time is little or no drawback. The following analyses (Rayleigh, Phil. Mag., June 1903) of the vapour arising from liquid air at various stages of the evaporation will give an idea of the course of events:—
| Percentage of Oxygen. | Percentage of Argon. | Argon as a Percentage of the Nitrogen and Argon. |
| 30 | 1.3 | 1.9 |
| 43 | 2.0 | 3.5 |
| 64 | 2.0 | 5.6 |
| 75 | 2.1 | 8.4 |
| 90 | 2.0 | 20.0 |
(R.)