288 



NATURE 



[May 6, 1909 



Dr. W. \V. Coulentz, of the United States Bureau of 

 Standards, recently completed an investigation of the 

 radiation constants of metals with a view to account for 

 the high efficiency of the new metallic filament incandescent 

 electric lamps, and his results are published in part iii. 

 of vol. V. of the Bulletin. The radiation from the filament 

 of an incandescent lamp provided with a fiuorite window, 

 after passing through a fiuorite prism, was measured by 

 the bolometer while the temperature of the filament was 

 kept constant. From the curve of distribution of energy 

 throughout the range of wave-lengths examined, the radia- 

 tion constant a of the formula dE = C\~''«"'^/^'rfA is found 

 for the material of the filament at various temperatures. 

 Its value for a " cavity " black body is known to be 4, 

 while for platinum it is 6. Dr. Coblentz finds it to be 

 about 6 for " flashed " and for untreated carbon, to be 

 between 7 and 8 for silica-coated carbon, between 6 and 

 8 for platinum and tungsten, between 6 and 7 for tan- 

 talum, and to be about 6-8 at all temperatures for osmium. 

 As the temperature at which the lamp is run increases, 

 the radiation constant decreases in the case of the metallic 

 filaments with the exception of osmium. At the normal 

 voltage the constants have the following values : — 

 melallised carbon, 6-i ; tantalum, 6-3; tungsten, 6-6; 

 osmium, 69. The high value of a explains the superiority 

 of the osmium lamp. 



The Proceedings of the American Academy for March 

 contain two papers from the Harvard Laboratory on the 

 atomic weight of chromium. Since the early determina- 

 tion of Berzelius in 1818 (Cr = 5s-95), thirty-three values 

 have been placed on record, the earliest being those of 

 Peligot (1844) and the latest those of Meineke (1890). 

 Rejecting one high and one low value, the eleven figures 

 given by Meineke ranged from 52-03 to 52-27, mean 52-12, 

 in good agreement with the earlier values of Siewert (1861), 

 5207; Baubigny (1884), 52-13; and Rawson (iS8g), 52-09. 

 The- methods now adopted by Baxter and his colleagues 

 consisted in converting silver chromate and dichromate 

 into the chloride and bromide, and thus deducing the per- 

 centage of silver in the chromium compounds. Conversion 

 of chromate into chloride gave Ag = 65-0345 per cent., and 

 into, bromide Ag = 65-032i per cent., mean 650333-per cent., 

 whence if Ag= 107-88, Cr=52-oo8; conversion of dichro- 

 mate into bromide gave Ag = 49-9692 per , cent., whence 

 Cr=52-oi3. -It is noteworthy that the final value, 

 Cr = 52-of, differs from the whole number by only one- 

 hundredth of a unit, whereas the figure adopted by the 

 International Committee for the present ■ and previous 

 years,. Cr=S2-i, differed by a tenth of a unit.- It will be 

 remembered that the recent revision of- the atomic weight 

 of nitrogen also resulted in bringing the value within o-oi 

 of the integer. 



Mr. J. H. Siiaxby asks us to state that, by an un- 

 fortunate mistake,' '- he . wrote ' " Faraday " instead of 

 " Tyndall "in his' letter upon the fluorescence of Lignum 

 Nephriticum, published last week (p. 249). 



OUR ASTRONOMICAL COLUMN. 



' . Development of M.-irtian Canals. — Through the agency 



: of the Kiel Centralstelle (Circular No. 107) we have re- 

 ceived a message from Prof. Lowell saying that the 



" development of the Martian canals corroborates the pre- 

 diction that they would be seen leaving the south polar 



-■cap of the planet. 



' Colours and Magnitudes of Stars. — In a note appear- 



?ing in these columns on February 4 [p. 410, No. 2049, 



vol. Ixxix.) we directed attention to Mr. Franks 's con- 



NO. 2062, VOL. 80] 



elusions regarding the relation of star colours to star 

 magnitudes in galactic and non-galactic regions. 



A note by Miss Bell, appearing in No. 5, vol. Ixi.x., of 

 the Monthly Notices, confirms Mr. Franks 's in showing 

 that there appears to be a slightly more intense relation 

 between luminosity and colour in the galactic regions. 

 This result was obtained by the statistical method of con- 

 tingency, and a further calculation shows that, as a chance 

 coincidence, the chances are 500 to i against there being 

 a group of stars so divergent from stars as a whole as 

 are the galaxy stars, whilst they are 2500 to i against 

 any random sample showing the divergency from the whole 

 that the non-galactic stars display. 



A Group of Red Stars in Sagittarius. — Whilst 

 examining the Draper memorial photographs of stellar 

 spectra, Mrs. Fleming has found that a plate covering the 

 area R.A. i8h. 48m. to iqh. 29m., dec. 130° S. to 23-1° S. 

 (1900), shows an abnormal number of red stars having 

 peculiar, spectra. The positions and spectral types of these 

 stars are given in Circular No. 149 of the Harvard College 

 Observatory. The area includes the n.f. portion of the 

 constellation Sagittarius, and is in the southern border 

 of the Milky Way. Besides twenty-one stars having spectra 

 of the third type, there are six of the same type with the 

 addition, of bright hydrogen lines (class Md) and one of 

 the sixth type (class R). 



In contradistinction to the above, Mrs. Fleming found 

 that a similar plate of a neighbouring region (R.A. 

 i7h. 24m. to iSh. iim., dec. 27° 8' S. to 38° o' S.) shows 

 a deficiency of red stars, but contains several stars having 

 peculiar spectra. These include stars of the third type, 

 a variable of the fourth type, two of the fifth type with 

 bright lines, and two gaseous nebula;. 



The Calcul.ition of Cometary Orbits. — It frequently 

 occurs that the definitive calculation of a comet's orbit is 

 carried out by two or more calculators working independ- 

 ently, and without the whole of the available observational 

 data. This leads to varying results and confusion, which 

 Prof. Kobold is trying to obviate. For this purpose he 

 publishes in No. 4319 of the Astronomische Nachrichten 

 a list of comets since 1757, with the names of the workers 

 by whom the definitive orbits have been, or are being, 

 worked out. The present list accounts for forty-six comets, 

 and Prof. Kobold will welcome any additions thereto. 



Photometric Observations at Catania. — A paper by 

 Signor A. Bemporad, in which he describes and discusses 

 the photometric observations made at Catania during the 

 three years 1904-6, appears as an abstract from vol. xxxvii. 

 of the Meniorie della Societd degli Spettroscopisti Italiani. 



The programme of work included (i) the determination 

 of the wedge constant ; (2) the study of the extinction 

 curve, for Catania and then for the Etna Observatory: 

 (3) th'- determination of the atmospheric absorption at both 

 plac^j ; and (4) the observations of variable stars. The 

 results obtained under each heading are fully discussed in 

 the memoir, and the light-changes of a number of variables 

 are compared with previously published ephemerides. 



Recent Solar Research. — As a reprint from vol. xxxvii. 

 of the Memorie dclIa Societd degli Spettroscopisti Italiani 

 we have received an interesting paper, in which Prof. 

 Ricco discusses the recent work done in the field of solar 

 research. Most of the subjects have already been discussed 

 in these columns, e.g. Hale's vortices, Deslandres's fila- 

 ments, Belopolsky's anomalous forms of the K line in 

 1906-7, &c., but readers of Italian will find Prof. Ricco's 

 review to be a useful resume of all these researches. 



Occultations of Planets.^ — In a brief note appearing in 

 No. 5, vol. Ixix., of the Monthly Notices (p. 431), Dr. 

 Downing gives the data for two occultations of planets 

 by the moon, during the present year, observable at British 

 observatories. Times and position-angles of immersion and 

 emersion are given for occultations of Mars and Venus 

 visible at Ottawa and Sydney respectively, the former on 

 September i, the latter on November 17. 



SS AuRiG.^ (31,1907) AN Irregular Variable. — A note 

 from Prof. Hartwig, appearing in No. 4319 of the Astro- 

 nomische Nachrichten, announces that the Bamberg 

 observations show the star SS .Aurigae to be an irregular 

 variable of the SS, Cygni type. 



