SCIENCE 



NEW YORK, SEPTEMBER 23. 1892. 



LIQUID OXYGEN. 



BY Q. D. UTEING, F.R.S., CAMBRIDGE, ENGLAND. 



It is now fifteen years since Pictet and Cailletet first liquefied 

 oxygen. Since then liquid oxygen has been tlie object of inves- 

 tigation by Olsewski, Wrobleaski, and more particularly by 

 Dewar. At a lecture delivered at the Royal Institution in Lon- 

 don, in June last, Dewar exhibited a litre of liquid oxygen in an 

 open vessel; and he has prepared from time to time many litres 

 of the liquid for the purpose of examining its properties. The 

 method he uses is the same in principle as Pictet's, but he has 

 much larger and better pumps for exhausting and compressing 

 the gases. The essential thing is to cool the oxygen well below 

 its critical temperature, or absolute boiling point, — 119° C. This 

 is effected by means of some other gas, such as ethylene, which 

 has a much higher critical temperature, namely + 10° C, and 

 still a very low boiling point under atmospheric pressure, namely 

 — 103°. Nitrous oxide, which has a critical temperature of 

 -4- 53° C, and boils under atmospheric pressure at — 93° C, may 

 also be employed. If the liquid ethylene be first cooled to — 80° 

 C. by immersion in a mixture of solid carbonic acid and ether, it 

 can then be easily reduced to — 103° by allowing it to evaporate 

 at the pressure of the air; and by pumping away the vapor as 

 fast as formed the temperature of the remaining liquid can be re- 

 duced as low as — 140°, twenty-one degrees below the critical 

 temperature of oxygen. At this temperature oxygen is liquid if 

 condensed until its pressure is equal to 30 atmosfiheres or there- 

 abouts. On removing the pressure the liquid boils and is cooled 

 by its ovs^n evaporation until under a pressure of one atmosphere 

 it falls to — 183° C. By pumping away the gas as it is formed, 

 the temperature is easily reduced to — 300° C, and the liquid 

 then remains quite tranquil, and has the appearance of so much 

 water. 



Air may be liquefied in the same way, but the boiling point of 

 nitrogen is somewhat lower than that of oxygen, namely — 193° 

 C, so that when liquid air is allowed to boil away gradually, the 

 residue becomes richer and richer in oxygen until nearly pure 

 liquid oxygen is left. 



The compressed oxygen met with in commerce always contains 

 a little air and some carbonic acid, and both jjass into the liquid 

 oxygen. The carbonic acid crystallizes out in the solid state and 

 renders the liquid milky. It may, however, be filtered through 

 paper, and is then perfectly limpid. 



To prevent the rapid deposition of hoar frost on the vessel con- 

 taining the cold liquid, it has to be protected by an outward ves- 

 sel, and the intervening space well dried. A beaker glass may 

 be fitted with a varnished wooden cover and a smaller beaker to 

 contain the liquid inserted through a hole in the cover, the space 

 between the two being dried by a layer of phosphoric anhydride. 



Oxygen does not show any increased chemical activity in conse- 

 quence of liquefaction. As already mentioned it may be filtered 

 through paper without affecting the paper. It is powerfully 

 magnetic. Poured into a saucer of rock salt it at once assumes 

 the spheroidal state, exaporating from its surface, but quite tran- 

 quil. If now it be brought near the pole of an electro-magnet, it 

 will jump up, through half an inch or more, and adhere to the 

 pole, looking like a blob of transparent ice. Of course it is not 

 really solid, and as soon as the current of the electro-magnet is 

 broken it falls down. Like iron it is attracted by either pole in- 

 differently. 



As it is the only transparent element which is magnetic, its 

 behavior to light is of great interest with reference to the electro- 



magnetic theory of light. According to that theory it would be 

 expected that light reflected from a plane surface of a transparent 

 magnetic body, when the reflected and transmitted rays are at 

 right angles, would not be polarized in the plane of incidence as 

 it is when the reflecting body is diamagnetic. Dewar has found, 

 however, that light incident at the proper angle and reflected by 

 liquid oxygen at — 200° C. is very completely polarized in the 

 plane of incidence. 



Seen by transmitted light liquid oxygen, in a thickness of three 

 or more inches, has a faint, but decided, blue tint. On examin- 

 ing the transmitted light through a prism, the cause Is plain. 

 There are several absorption bands, of which the strongest is in 

 the yellow. These bands, as observed by Olsewski, and by Live- 

 ing and Dewar who extended their observations into the ultra- 

 violet, are identical in position with, but much darker than, the 

 diffuse bands produced by oxygen gas. They coincide with cer- 

 tain diffuse dark bands noticed by Brewster in the solar spectrum, 

 and ascribed by him to atmospheric absorption because they were 

 stronger when the sun was near the horizon than when he was 

 high in the sky. The persistence of these bands indicates con- 

 tinuity in the physical state of oxygen when passing from the 

 gaseous to the liquid state. 



It was observed by Jannsen. and Liveing and Dewar' s observa- 

 tions tend to the same conclusion, that the intensity of these dif- 

 fuse bands, for a given thickness of the gas, increases as the 

 square of the density. On the kinetic theory such a result would 

 follow if the molecules of oxygen absorb the corresponding rays 

 when they are under the influence of other molecules but not 

 when in free path. For both the number of the molecules in a 

 given thickness, and the frequency of their collisions, increase 

 directly as the density. 



Furthermore, oxygen gas produces, besides these diffuse bands, 

 certain absorptions consisting of rhythmical groups of tine lines. 

 These, in the solar spectrum, are known as A, B, and a; and it 

 was Egoroff who identified these lines with the absorptions of 

 oxygen gas. Their intensities appear to increase directly as the 

 density, and may therefore fairly be ascribed to the action of the 

 free molecules of oxygen gas. What becomes of these absorptions 

 when the gas is liquefied? Olsewski. looking through 30 milli- 

 meters of the liquid, observed an absorption at the place of A. 

 Liveing and Dewar, looking through six inches of the liquid, saw 

 absorptions corresponding to both A and B, but somewhat differ- 

 ent from those due to the gas. As produced by the gas, A is a 

 group of lines which are very close together on the less refrangi- 

 ble side and are set farther and farther apart as they get more 

 refrangible; so that the group when seen with low dispersion has 

 the appearance of a shaded band with a strong, sharp edge on the 

 less refrangible side, gradually fading away on the more refrangi- 

 ble side. B is on the blue side of A, and is precisely similar to 

 it, but not so intense, a is still more refrangible, and still weaker. 

 The absorption of the liquid at the place of A is also a shaded 

 band, but its shading is turned the other way. Its sharp and 

 strongest edge is on its more refrangible side, and it fades away 

 on the less refrangible side. The strong edge does not correspond 

 exactly with the strong edge of A, but is a little more refrangible, 

 though still falling within the group. The band of the liquid 

 could not be resolved, like A, into lines. The band correspond- 

 ing to B is precisely similar, but fainter. It overlaps B, and has 

 its strong edge a little more refrangible than that of B. It seems 

 that we have in these bands the absorptions due to the individual 

 molecules of oxygen, only modified in the way described by the 

 change from gas to liquid ; and we should infer that the molecules 

 of the liquid are still the same as those of the gas. 



There is more diSiculty in determining the physical characters 

 of oxygen than the facility with which it can be manipulated 



