February 6, 1896] 



NA 'JURE 



329 



appeal is made for their general adoption. Rowland's scale of 

 wave-lengths, as represented by the tables in course of publi- 

 cation in the journal above named, is to be employed, and the 

 unit of wave-length is to be the ten-millionth of a millimetre, 

 or " tenth-metre." For measurements of velocity in the line 

 of sight, the kilometre is to be taken as the unit. To distin- 

 guish the lines of hydrogen, the nomenclature starting with 

 H„ in the red and continuing in alphabetical order through the 

 entire series is agreed upon. Maps of spectra are to be drawn 

 with the red end to the right, and tables of wave-length are to 

 t)e printed with the shorter wave-lengths at the lop. 



Although some of the leading workers in astrophysics have 

 not been consulted, it is probable that these arrangements, so 

 far as they go, will meet with general approbation. It is to be 

 regretted, however, that the representation of intensities of 

 spectrum lines was not considered, as a scale which every one 

 might be willing to adopt is, perhaps, even more urgently re- 

 quired than any general agreement on the points to which 

 reference is made above. 



Reproduction of Astronomical Photograi'hs. — The 

 Council of the Royal Astronomical Society has lately under- 

 taken the reproduction (by paper prints and lantern slides) of a 

 selection of the instructive astronomical photographs in the 

 possession of the Society. The prints and lantern slides are sold 

 to Fellows of the Society at approximately cost price, and full 

 <letails as regard subject, instrumental data, exposure. &c., are 

 given upon each. Among the celestial pictures which have been 

 thus rendered available to a wider circle of astronomers, are 



observaition, while the body itself expanded; six days afterwards, 

 the nucleus had almost disappeared again. A photograph taken 

 on November lo, 1892, is chiefly remarkable as showing a large 

 irregular mass of nebulosity covering an area of at least a square 

 degree, and connected with the comet by a short hazy tail. This 

 curious appendage, which certainly belonged to the comet, seems 

 to have been overlooked by most observers, but its recognition may 

 possibly at some time or other prove to be of importance. The 

 facts seem to be in favour of the comet having suddenly become 

 bright just before the time of its discovery. It differed from the 

 average comet in having a nearly circular orbit, and unless there 

 had been some great change in its path, or some internal change, 

 it should have been discovered long before. As the comet could 

 not be seen with the Lick telescope during the .succeeding 

 opposition. Prof. Barnard thinks that it no longer exists in the 

 cometary form, and will never be seen again. 



photographs of total solar eclipses of 1886, 1889, and 1893, Dr. 

 Roberts' photographs of the IMeiades and the (jreat Nebula in 

 Orion, Prof. Barnard's photographs of the Milky Way, and of 

 Brooks' and Swift's comets. Dr. Gill's photograph of the nebula 

 about Tj Argus, and MM. Loewy and Puiseux's lunar photo- 

 graphs. The accompanying illustration of the eclipse of April 

 16, 1893, *i3.s been reduced by one-third from a print sold by the 

 Society. The original was taken by Sergt. -Major Kearney, 

 R.K., at Fiundium, West Africa, with a Dallmeyer photo- 

 heliograph, the exposure being twenty seconds. 



Holmes' Comet. — Prof. Barnard has just pulilished an 

 account of his observations and photographs of this comet, made 

 during its appearance in 1892 and 1893 ^Astrophysical Journal, 

 vol. iii. No. I ). Some of the telescopic features appear to have 

 been quite unique. On January 4, 1893, only a feeble glow was 

 visible ; twelve days later it seemed like a hazy star, and the 

 nucleus was actually seen to brighten in the few hours of 



NO. 1371, VOL. 53] 



THE LIi2UEFACTI0N OF AIR AND RESEARCH 



AT LOW TEMPERA TURES. 1 

 T^HE best and most economical plant for the production of 

 -'■ liquid air or oxygen is one based on the general 

 principle of that used by Pictet in 1878, for liquefying oxygen ; 

 instead, however, of using Pictet's combined circuits of liquid 

 sulphur dioxide and carbon dioxide kept in circulation by com- 

 pression, liquefaction and exhaustion, it is better to employ 

 ethylene in one circuit, as Cailletet and Wroblewski did, and to 

 use nitrous oxide, or preferably carbon dioxide, in another. 

 Further, instead of causing the oxygen to compress itself during 

 its formation from potassium chlorate heated in an iron bomb 

 connected with the refrigerator, it is found convenient to use gas 

 previously compressed in steel cylinders. 



A very convenient laboratory apparatus, the arrangements of 

 the circuits of which will be easily understood from the sectional 

 view shown in Fig. i, has been devised for the liquefaction of 

 small quantities of oxygen or other gases ; with this simple 

 arrangement, 100 c.c. of liquid oxygen can readily be obtained, 

 using liquid carbon dioxide at - 79° C. for cooling and employ- 

 ing no exhaustion. The gaseous oxygen, cooled before 

 expansion by passing through a spiral of copper tube immersed 

 in solid carbon dioxide, passes through a fine screw stopcock 

 under a pressure of 100 atmos., and thence backwards over the 

 coils of pipe. The liquid oxygen begins to drop in about a 

 quarter of an hour from starting. The pressure in the oxygen 

 cylinders at starting is generally about 150 atmos., and the best 

 results are got by working down to about 100. This little 

 apparatus will enable liquid oxygen to be used for demonstration 

 and research in all laboratories. 



By employing jacketed glass vessels, of which the annular 

 space is highly exhausted, for storing liquefied gases, the influx 

 of heat is reduced to one-fifth of that which occurs when the 

 jacket contains air ; if the interior walls are silvered, or excess 

 of mercury vapour is left in the jacket, the influx of heat is 

 again reduced to one-sixth, so that the total effect of the high 

 vacuum and the silvering is to reduce the ingoing heat to about 

 3i per cent, of that which enters when these precautions are 

 neglected. The suggestion that the metallic coating is useless, 

 because Pictet has found that all kinds of matter are transparent 

 to heat at low temperatures, is thus disposed of; further, no 

 increase in the transparency of glass to thermal radiation occurs 

 on cooling to the boiling point of air. 



In order to test Olszewski's statement that air cannot be 

 solidified at the lowest pressures (/'-^zY. Mag., February 1895), 

 the author's former experiments have been repeated on a larger 

 scale. If a litre of liquid air be exhausted in a silvered vacuum 

 ves-sel, half a litre of solid air may be obtained and kept 

 solid for half an hour. The solid is at first a stiff transparent 

 jelly, which, when placed in a magnetic field, has the still lic|uid 

 oxygen drawn out to the poles, showing that solid air is a 

 nitrogen-jelly containing liquid oxygen. Solid air can only be 

 examined in a vacuum or an atmosphere of hydrogen, because it 

 instantly melts on expcjsure to the air, causing an additional 

 quantity of air to liquefy ; it is strange to see a mass of .solid air 

 melting in contact with the atmosphere, and all the time welling 

 up like a fountain. 



On causing dry air, contained in sealed flasks, to solidify by 



1 A paper read Ijefore the Chemical .Society on December 10, 1895, by 

 Prof. J. Dewar, F.R..S. (Abridged from the Proceedings of the Society 

 issued J.tnuary 14.) ^. 



