July 21, 1921] 



NATURE 



663 



intensity of light. It was not until 1893, after the 

 subject had been considered for twenty-two years, that 

 he gave details of his investigation of it. He made 

 many successful experiments on photography in natural 

 colours, but his greatest self-contained achievement 

 was his photography of the infra-red. The normal 

 sf>ectrum as photographed on his plates was more than 

 five times the length of the visible spectrum, for they 

 were sensitive to* the ultra-violet right away through 

 the visible spectrum to a wave-length of 2200/i/z. 

 Abney was accustomed to quantitative work from the 

 first, and perhaps the most irnf>ortant service he ren- 

 dered was the introduction of methods of measurement 

 into scientific photography. 



The Comptes rendus of the Paris Academy of 

 Sciences for June 20 contains a note by M. Baille- 

 Barrelle on the production of coke from Sarre coal. 

 By the usual method of coking, this coal is well known 

 to give a poor coke, but M. Baille-Barrelle shows that, 

 by a special mode of heating, Sarre coal can be made 

 to yield a coke comparable with the finest cokes from 

 Ruhr coal. The experiments were made on a semi- 

 industrial scale (charge of 500 kg.), and preliminary 

 work on the extension to a full commercial scale has 

 been commenced. The coal is first maintained at a 

 temperature of 320° C. for some time ; then the tem- 



I perature is slowly raised uniformly to a final tem- 



I perature of 750° C, or about 200° C. below the usual 



coking temperature. Figures for the resistance to 



' crushing and shaking are given. It is also claimed 



that the bj'-products obtained are superior to those 



given by the ordinary coke oven, and an investigation 



into their nature is in progress. The yield of am- 



: monia was unexpectedly high, about double that ob- 



I tained when the coal is coked in the ordinary way ; 



1 owing to the lower temperature a reduced quantity of 



I ammonia was anticipated. It is probable that the 



I actual quantity of ammonia produced was less, and 



that the increased yield was due to the lessened 



i amount decomposed into nitrogen and hydrogen. If 



I the process is successful on the large scale, the Lor- 



j raine iron industry will be freed from the necessity 



I of using Ruhr coke. 



I The National Physical Laboratory has issued a 

 i pamphlet dealing with "Tests on Volumetric Glass- 

 j ware Used in Dairy Chemistry," single copies of 

 which may be obtained free of charge on application 

 to the Director, Metrology (Glass Testing) Depart- 

 ment, National Physical Laboratory, Teddington. 

 The pamphlet contains specifications as to size and 

 construction of butyrometers, test-bottles, and pipettes 

 which can be accepted for test by the Laboratory. 



Our Astronomical Column. 



AuROR,^ AT A Height of 500 km. — The careful 

 auroral observations made in Norway and Sweden 

 liave established the remarkable fact that some of the 

 streamers extend to the height of 500 km. above the 

 earth's surface. This presumably implies that there 

 is a certain amount of atmosphere at that height, 

 which is a conclusion of cosmical importance. 



Geofysiske Publikationer, vol. ii.. No. 2, contains an 

 investigation by Dr. Carl St0rmer of the height of 

 streamers during the brilliant aurora of March 22-23, 

 1920. There were seven photographic stations at 

 work in Norway on this occasion, and telephonic com- 

 munication enabled simultaneous exposures to be 

 made, the cameras being directed to the same stars. 

 The investigation is based on simultaneous photo- 

 graphs taken at Christiania and Kongsberg, which 

 are 657 km. apart. The streamers photographed had 

 well-defined edges, and crossed the constellation 

 Cassiopeia, the brighter stars being visible on the 

 plates. The heights of seven points in the" streamers 

 are determined as 597, 550, 607, 562, 528, 485, and 

 519 km. respectively. Two pairs of plates are repro- 

 duced, on which the streamers and the stars are 

 clearly visible. The author notes that it is only the 

 extremities of the long rays that attain these great 

 heights. The bases may be as low as 85 to 90 km. 



The Minor Planet Eros. — This planet will make 

 one of its near approaches to the earth early in 193 1, 

 when there will doubtless be another solar parallax 

 campaign. A parallax still more accurate, however, 

 than that obtained by direct measures will probably 

 be determined by the very large perturbations pro- 

 duced by the earth on the planet's motion. For this 

 purpose it is desirable to obtain accurate observations 

 at every opposition. The planet will next be in 

 opposition in mid-September in N. decl. 14°, magni- 

 tude about 105. Mr. F. E. Seagrave has computed 

 an ephemeris for Greenwich midnight, a portion of 

 which is given below. Corrections, due to G. Stracke, j 



NO. 2699, VOL. 107] 



of —22s. and —2' y" have been applied 

 ascension and declination : — 



July 



Aug. 



23 

 27 

 31 

 4 

 8 

 12 

 16 



K.A. 

 h. m. s. 

 23 41 24 

 23 41 23 

 23 40 42 



23 39 17 

 23 37 6 

 2334 8 

 23 30 22 



6581 

 7506 

 8418 



9311 



10 i8-i 



11 2-1 



II 430 



Log r 



0-23754 

 0-23600 

 023438 

 023266 

 023088 

 0-22900 

 022700 



23 25 49 12 198 022494 



to the right 



Log .i 



998409 

 996672 



994938 

 9-93211 

 991512 

 9-89861 

 988265 

 986759 



Jupiter's Four Great Satellites. — ^The Annals of 

 Leyden Observatory (vol. xii., parts i and 2) consist 

 of researches on these satellites by Prof. VV. de Sitter 

 and Dr. A. J. Leckie respectively. These parts were 

 published in 1918 and 1919; they are therefore quite 

 independent of Prof. Sampson's theory, which only 

 appeared in print in 192 1. One point of Prof, de 

 Sitter's method is the use of a new intermediary 

 orbit ; instead of using the Keplerian ellipse, he sub- 

 stitutes for the eccentricity the great periodic in- 

 equalities. This is analogous to the use by Drs. Hill 

 and Brown of the variation oval as intermediary orbit 

 in the lunar theory instead of the Keplerian ellipse. 



Prof, de Sitter finds for the masses of the satel- 

 lites in terms of that of Jupiter 3796, 2541, 8201, and 

 4523 (units of the 8th decimal). In terms of the 

 moon's mass these are 0-985, 0-659, 2-128, and 

 I- 173. Using the diameters of the satellites found 

 by the interferometer (mean of Hamy's and Michel- 

 son's results), viz. i-oo', 0-905", 1-325", and 1-31" at 

 distance 5 units, the densities become 0-853, 0-788, 

 o-8ii, and 0-462 in terms of that of the moon. 



Prof, de Sitter's final values of the mean daily 

 motions of I., II., and III. referred to First Point of 

 Aries are 203-48899280°, 101-37476180°, and 

 5031764630°. These have been adjusted to fit the 

 relation n^ — ^ni+ztij^o. 



