(c) The iron flask, 3 litres in capacity, containing liquid ethylene, under high pressure.

(e) The condenser, for cooling ethylene by means of ether and solid carbon dioxide.

(f) The cock connecting the condenser with the small pump.

(g) The cock serving to let liquid ethylene into the vessel b.

(h) Iron flasks, 10 litres in capacity, containing oxygen or air under a pressure of 100 atm., connected by means of a tube with the manometer i and the cylinder a.

(k) The glass vessel with triple walls, serving to receive. liquid oxygen or air under atmospheric pressure.

(1) The glass tube, plunged in liquid oxygen or air, serving to liquefy hydrogen, connected with the manometer m and iron cylinders n n', containing hydrogen under a pressure of

150 atm.

(0) Iron flask containing hydrogen under the pressure of several atmospheres, which passes through the liquid oxygen in the vessel k, when its pressure is diminished.

(pp) Mercury manometers, serving to measure the pressure of ethylene and liquid oxygen, contained in the vessels b

and k.

(r) Cocks serving to connect the vessels b and k with the large pump.

(8) Iron cylinder, containing liquid carbon dioxide.

As my experiments on the liquefaction of hydrogen do not at all confirm those of M. Pictet, made at Geneva in 1879, I requested Dr. Krzyżanowski to examine whether the glaring discrepancies between my experiments and M. Pictet's might not be explained by impurities contained in the hydrogen he used. And indeed Dr. Krzyżanowski (24) proved that, if potassium formate be heated with potassium hydroxide, hydrogen cannot be obtained free from moisture, even in the most advantageous case, which occurs when the latter is in excess; and that if these substances are employed in the proportion given by M. Pictet, that is with potassium formate in excess of what is required by the calculation of molecular weights, we get by heating them a sample of hydrogen containing not only water but also considerable quantities of carbon monoxide and dioxide. Now these substances, interfering with the purity of the hydrogen experimented on by M. Pictet, were doubtless the ground of the extraordinary results M. Pictet described, and which, though glaringly improbable, are cited in nearly all chemical manuals. Perhaps this remark of mine will contribute to a proper appreciation of M. Pictet's experiments respecting the liquefaction and solidification of hydrogen; perhaps it will suggest to the author (who has now established a laboratory in Berlin with the special purpose of obtaining very low temperatures) that he might, with advantage for science, control and rectify the results of the work he performed in 1879.

On the Optic Properties of Liquid Oxygen.

[These investigations were made in collaboration with my friend Prof. Witkowski, and were published in the Reports of the Cracow Academy and in the Bulletin International, October 1891 and June 1894.]

As far back as 1887 I discovered a very remarkable and powerful selective absorption of light in the liquefied oxygen.

It was desirable to obtain some numerical data as regards the absorbing power for the several bands. For this purpose we made use of the spectrophotometer of Glan. The liquid oxygen was contained in the innermost tube, provided with a flat bottom, of the apparatus shown in the annexel figure (fig. 4, taken from the Bulletin International, October 1891). Fig. 4.

The thickness of the liquid could be varied to a known degree by screwing up and down the tube ef, closed at both ends by plane glass plates, and provided with a millimetre division.

So far as the accuracy of photometric measurements under such difficult conditions could be relied upon, we found a proportion of light varying between 84 and 89 per cent., transmitted by a layer of oxygen 1 millim. in thickness in the most intense part of the yellow-greenish band (λ=577 to λ=570). The corresponding average number for the red band (X=630 to λ=638) was 88 per cent.

In the same pamphlet we described a method of determining the coefficient of refraction of liquid oxygen and gave results for sodium light. The most suitable method for the purpose proved to be that of total reflexion. The liquid is contained in a thin iron parallelepipedon A (see fig. 5), provided with plane glass windows, and protected by several varnished pasteboard boxes of similar form containing some phosphoric anhydride. A double glass plate, composed of two carefully selected plane bits of glass, separated at the corners by small pieces of mica (thickness about 0.006 millim.) and cemented at their obliquely-ground borders by means of fish-glue, is immersed in the oxygen. The double plate is rigidly conected with the axis of a divided circle. A similar arrangement has been employed by Ketteler in some of his investigations on the refraction and dispersion of water (Wiedemann's Annalen, vol. xxxiii.). It is therefore unnecessary to enter into particulars as regards the mode of observation and calculation of results. A first series of observations gave the value 1.2235 of the coefficient for sodium light, in very close agreement with a result found by Messrs. Liveing and Dewar by means of the prism method. Later determinations of the same constant by myself and Prof. Witkowski, made with the view of ascertaining the dispersion of liquid oxygen (Bulletin International, July 1894), yielded a rather smaller result (1.2226) as a mean of five distinct measurements. At the same time we found n=1.2213 for λ=6705, and 1-2236 for the wave-length 535 μμ. Anomalous dispersion, specially sought for, could

not be detected.

Besides these researches I have made (partly by myself, partly in collaboration with Prof. Witkowski) certain hitherto unpublished experiments concerning the intensity of chemical energy at low temperatures; I mention them here to complete the list of my investigations. They refer chiefly to two substances, viz. liquid ethylene and oxygen, under the influence of agents which combine with these substances at the usual or at a higher temperature. Ethylene, boiling under atmospheric pressure (-102°.5 C.) was submitted to the

action of chlorine and bromine. The action was feeble, but in both cases I obtained products of combination (C2HCl, and C2HBr2) in considerable quantity.

A

The possibility of preparing relatively large quantities of liquid oxygen in open vessels at atmospheric pressure gave us an opportunity to examine the chemical properties of this interesting substance more closely. We ascertained first the exceedingly feeble chemical affinity of liquid oxygen. piece of metallic sodium in contact with it showed no change in its metallic brilliancy; a hardening of the substance, in consequence of its very low temperature, being apparently the only effect produced. This might have been anticipated, considering that every trace of moisture had been frozen away. Potassium acted similarly to sodium.

But when the reaction of oxidation with light and heat phenomena had already begun, the low temperature (—181°·4) · is not able to cool the burning substance to such a degree as to interrupt the reaction. For instance, a piece of ignited wood immersed in liquid oxygen takes fire just as in gaseous oxygen; a steel spring burns and spreads sparks of burning iron, which in this experiment burst the glass vessel of oxygen, and the liquid oxygen was consequently spilt on the table, giving thus the interesting sight of liquid drops rolling and jumping about in a perfectly spheroidal state.

In this connexion, as I had (in 1891) large quantities of liquid; oxygen, my friend Prof. Kreutz performed a series of experiments on the behaviour of coloured substances at very low temperatures, and showed that many of them (HgS, HgI2, I, &c.) become much brighter at -181°4.

This

From this summary of researches, as well as of dates, it follows that the first apparatus serving to produce large quantities of the liquefied so-called permanent gases, with the solitary exception of hydrogen, was constructed by me. apparatus can be enlarged at will by increasing its parts, but without changing anything in its construction, so that it might be used to obtain liquefied gases in factories should they at any time prove of practical utility. By means of this apparatus I obtained as large quantities of liquid gases as I wanted; and they were used for the first time on a large scale as cooling agents (for instance, in my attempts to liquefy hydrogen), or as an object of scientific researches (the) absorption spectrum of liquefied oxygen, its coefficient of refraction, &c.)

The experiments of Prof. Dewar are merely the repetition Phil. Mag. S. 5. Vol. 39. No. 237. Feb. 1895.

Р