April 20, 1899] 



NATURE 



595 



nickel and cobalt. Both papers bear the stamp of highly 

 accurate and searching work, and, in addition to a discus- 

 sion of the main question, contain many subsidiary features 

 of interest. Complete analyses of nickelous bromide and 

 cobaltous bromide were undertaken, the salts being reduced 

 by moist hydrogen, and the metals weighed as such. The bro- 

 mine was determined as silver bromide. The final result for 

 nickel was 58706, and for cobalt 5S'995. As the result of a 

 critical examination of earlier determinations in the case of 

 nickel, the authors select as the most trustworthy the numbers 

 of Zimmermann, 5S'694, and of Winkler, corrected by them to 

 58 '69, so that the value 5870 may be taken as final. As the 

 authors' results for cobalt show less concordance than those for 

 nickel, the extreme values being 58 '955 and 59 '021, fresh 

 experiments involving different methods are being undertaken. 

 The evidence of all the work is said to strongly support 

 " Winkler's contention that nickel and cobalt, as we knew 

 them of old, cannot contain more than an infinitesimal amount 

 of any unknown element." Several radically different methods 

 of preparation and many fractionations led to atomic weights, 

 constant within a reasonable limit of experimental error. 

 Cobalt has a higher atomic weight than nickel, although this 

 conflicts with the inference to be drawn from the position of 

 rhodium and palladium in the periodic classification. Amongst 

 the minor matters involved in the above investigations the 

 following may be mentioned : — Nickelous bromide and cobaltous 

 bromide, when sublimed in porcelain tubes, are slightly con- 

 taminated with sodium bromide. The bromides are more 

 easily reduced by moist than by dry hydrogen ; the metals so 

 prepared do not occlude an appreciable amount of hydrogen, 

 possibly on account of the trace of sodium bromide present. 

 Both nickel and cobalt are acted upon appreciably by water, 

 giving colloidal solutions of the hydrates. This circumstance 

 and also the use of glass apparatus, leading to siliceous residues, 

 account mainly for Kriiss's supposed discovery of " gnomium." 

 The colloidal solution of cobaltous oxide absorbs cxygen from 

 the air, and deposits cobaltic hydrate : no such action was 

 observed in the case of nickel. 



The additions to the Zoological Society's Gardens during the 

 past week include a Common Badger [Meles /axiis), British, 

 presented by Mr. Geo. M. Margon-Wilson ; a Silver-backed 

 Fox {Cam's chama) from South Africa, presented by Mr. C. R. 

 Rennie ; two Black Rats (Mus ratliis, var.), British, presented 

 by Mr. W. J. Smith ; a Purple-faced Monkey {Seniiiopitheciis 

 cephalopterus, 9 ) from Ceylon ; a Common Camel ( Camelus 

 dromedarius, i ) from Arabia, deposited ; two Black -headed 

 Buntings (Einberha melanoccphala), a Puffin (Fratercula 

 arctica), European ; two Canada Geese (Bernicla canadensis) 

 from North America ; fourteen Golden Carp ( Carassius 

 aitratus) from China, purchased. 



OUR ASTRONOMICAL COLUMN. 



Tuttle's Comet (1899 !>)■ — The following positions are 

 given by Herr J. Rahts in Ast. Nach. (Bd. 149, No. 3555) : — 

 Ephcmeris for I2i. Berlin Mean Time. 

 1899. R..A. Decl. Br. 



April 20 



4 8 26-6 



21 ... II 58-1 



22 ... 15 28'9 



23 ... 18 590 



24 ... 22 285 



25 ... 25 57-4 



26 ... 29 257 



27 ■•■ 4 32 53'4 



I -So 



82 



+ 16 37 42 



16 II 14 



IS 44 31 



IS 17 34 •• 



14 50 24 ... 



14 23 I ... 



13 55 24 .■■ I 86 

 + 13 27 35 ■■• 



The comet should be looked for immediately after sunset ; it 

 travels during the week from between 7 and 8 Tauri to a posi- 

 tion about 2° south of a Tauri (.Mdebaran). 



NO. I 538, VOL. 59] 



Temfel's Comet 11873 ILL — M. L. Schulhof, of the Paris 

 Observatory, gives, in a communication to Asl. Nach. (Bd. 149, 

 No. 3554) ths calculated elements and ephemeris of this comet, 

 which will arrive at perihelion about the middle of June next. 



Elements. 

 T = 1S99. June iS o Paris Mean Time. 



27 ... 18 36 50-4 ... - 5 13 22 



The brightness in 1878 was 0-113, while during the appar- 

 ition of 1894 it was 0-190, so that the present return should be 

 much easier of observation. The comet will be travelling Irom 

 the south-eastern border of Ophiuchus, through Scutum Sobieski, 

 into Aquila, and search should be made in the early morning. 



VARi.iBLE Star Notes. — The fourth pamphlet issued from 

 the Rousdon Observatory of Sir C. E. Peek, at Lyme Regis. 

 South Devon, contains the individual observations of the vari- 

 ables R and x Cygni, extending over the period 1887 January 

 20 to 1S96 December 23. The resulting estimates of magnitude 

 are also plotted out at the end of the pamphlet showing the 

 light-curves, from an examination of which it ^ippears that the 

 period of x Cygni is something over 14 months, and that of 

 R Cygni about 12-5 months. 



Spectrvm of Sati'Rx's Rings. — Prof. Vogel many years 

 ago stated that he had observed a strong absorption band in the 

 red region of the spectrum of Saturn, at A 6183, which was 

 extremely faint or absent in the spectrum of the rings. Prof. 

 Keeler in 1S89 could detect no trace of it in the ring spectrum, 

 using the 36-inch Lick refractor. During the evening of 

 August 18, 189S, Mr. Ellerman, of the Verkes Observatory, 

 photographed the spectrum of Saturn on an isochromatic plate 

 very sensitive to the red region, using the 40-inch telescope. 

 The spectroscope used had one flint prism of 60' angle, the 

 collimator and camera lenses being i -4 inches aperture and 19 

 inches and loj inches focus respectively. The planet being far 

 south the exposure had to be short, so the slit had to be used 

 fairly wide. A reproduction of an enlargement (7-5 times) 

 from this negative is given in Astrophysical yoiirnal for 

 March (vol. ix. p. 186). The absorption band referred to is 

 very readily seen in the spectrum of the ball of the planet, but 

 no trace of it is at all visible in the ring spectra. The conclu- 

 sion drawn by Prof. Hale from this fact is that the rings possess 

 little or no atmosphere', thus confirming the result formerly 

 obtained from visual observations. 



The Sun's Mean Temperature. — In our last issue we 

 called attention to a criticism of Dr. See's article on " The 

 Sun's Heat, " by Dr. A. S. Chessin, in the Astronomical Journal, 

 No. 456. In the current issue. No. 458, there are some further 

 remarks by Prof. S. Newcomb and Dr. Chessin, which we give 

 below. 



Prof. Newcomb writes: "Dr. Chessin's remark in A. J., 

 456, does not seem to me well founded. The problem is this : 

 The parts of a spherical gaseous mass A are kept in equilibriuin 

 between the force of their mutual gravitation, and of their 

 elasticity due to temperature. To preserve this equilibrium let 

 there be an absolute temperature T,„ which may increase from 

 the surface to the centre. Now, by the radiation of heat, let 

 the radius of the mass A contract from R„ to R. What is the 

 temperature T necessary to maintain the equilibrium of the 

 mass after contraction ? The formula given by See, 

 T = V..Jj, 

 R ' 



