On reducing the pressure to 10 mm. of mercury the nitrogen solidified. Prof. Dewar has stated that liquid air solidifies as such, the solid product containing a slightly smaller percentage of nitrogen than is present in the atmosphere. My experiments have proved this statement to be incorrect; liquid oxygen does not solidify even when boiling under a pressure of 2 mm. of mercury.

After carrying these experiments to a successful conclusion, I found that it was yet necessary to prove that, on reducing the vapor pressure of boiling oxygen, to a minimum, no corresponding fall of temperature takes place. The vessel, e, was partially filled with liquid oxygen, and, by means of a small siphon, a small quantity of the liquid was allowed to flow into the tube, a. The inner vessel, a, was then connected with the air pump and manometer, and the pressure was reduced to 2 mm. of mercury. The oxygen remained liquid and quite clear. In a second experiment the temperature of the liquid oxygen, boiling under 2 mm. of mercury pressure, was measured by means of a thermometer. The temperature indicated lay above -220° C., a temperature easily arrived at by means of liquid air. I, therefore, concluded that liquid air was a much more efficient cooling agent than liquid oxygen, and that it would be quite unnecessary to make further experiments on the liquefaction of helium.

In every single instance I have obtained negative results, and, as far as my experiments go, helium remains a permanent gas, and apparently much more difficult to liquefy than even hydrogen. The small quantity of the gas at my disposal, and, indeed, the extreme rarity of the minerals from which it is obtained, compelled me to carry out my investigation on a very small scale. Using a larger apparatus, and working at a much higher pressure, I could have submitted the gas to greater expansion. Further, I should have been able to measure the temperature of the gas at the moment of expansion by means of a platinum thermometer, as I did when working with hydrogen; but to make such experiments I should have required 10, if not 100, liters of the gas. As I was unable to determine the temperatures to which I cooled the gas, by any experimental means, I have been obliged to calculate them from Laplace's and Poisson's formula for the change of temperature in a gas during adiabatic expansion.

T/T = (p/p)k – 1/k.

Where:

T, p are the initial temperature and pressure of the gas.

T, p are the final temperature and pressure of the gas.

k is the ratio (cp/cv) which, for a monatomic gas, is 1.66.

In the first series of experiments the gas, under a pressure of 128 atmospheres, was cooled down to -210° C.

The results of these calculations tend to show that the boiling point of helium lies below -264° C., at least 20° lower than the value I have found for the boiling point of hydrogen. If the boiling point of a gas be taken as a simple function of its density, helium, which, according to Prof. Ramsay's determination, has a density of 2.133, more than double that of hydrogen, should liquefy at a much higher temperature. Both argon and helium have much lower boiling points than might be expected, judging from their densities. This anomalous condition may be accounted for by the fact that in each case the molecular structure is monatomic, as shown by the values obtained for the ratios of their specific heats.

The permanent character of helium might be taken advantage of in its application to the gas thermometer. The helium thermometer could be used to advantage in the determination of the critical temperature and boiling point of hydrogen. To determine whether the hydrogen thermometer is of any value at temperatures below -198° C. I carried out a series of experiments, in which I measured the temperature of liquid oxygen boiling under reduced pressure. I made use of the identical thermometer tube employed by T. Estreicher (Phil. Mag. [5] 40, 54, 1898) as a hydrogen thermometer for the same purpose, and applied the same corrections as were made in his experiments.

The results of these experiments prove that the coefficient of expansion of hydrogen does not change between these limits of temperature, and that the hydrogen thermometer is a perfectly trustworthy instrument even when employed to measure the very lowest temperatures.

I have already pointed out (Wied. Ann., Bd. xxxi, 869, 1887) that the gas thermometer can be used to measure temperatures which lie even below the critical point of the gas with which the instrument is filled. For instance, the critical temperature of hydrogen, which I have found to be -234.5° C. (Wied. Ann., 56, 133; Phil. Mag. [5] 40, 202, 1898) can be determined by means of a hydrogen thermometer. The helium thermometer could be used at much lower temperatures, and would probably give a more exact value for the boiling point of hydrogen than it is possible to obtain by means of a platinum thermometer.

[1] Translated from the original paper, by Prof. K. Olszewski, in the Bulletin de l'Academie des Sciences de Cracovie for June, 1896, "Ein Versuch, das Helium zu verflunigen," by Morris Travers, and published in Nature.