5^4 



NATURE 



[February i6, 191 i 



1904-7 show that ^ Gruis is a binary with a range of 

 velocity from —8-7 to +17 km. 



Variations in the radial velocities of the following stars 

 have also been detected from Lick and Santiago observa- 

 tions, and are rejx>rted by Prof. Campbell : — 16 Aurigae, 

 Oj Canis Maj., 12 Comae Berenices, 4 Ursae Min., t and 

 36 Ophiuchi, / Draconis, A Sagittarii, and o Cygni. In 

 the case of i Ophiuchi, a plate taken on April 28, 19 10, 

 shows that the line at X 4481 distinctly double, giving 

 radial velocities of —77 km. and +9-2 km. for the two 

 components. Fifteen plates of a Cygni, taken between 

 August, 1896, and December, 1909, show that the 

 variability of the velocity is not great, the range being 

 from o to 7-9 km. 



Observations of Jupiter's Galilean Satellites. — In 

 No. 5 of the Transvaal Observatory Circulars, Mr. Innes 

 gives an account of the observations of Jupiter's satellites 

 made at the observatory during December, 1909, to August, 

 19 10. The observations were made with the 9-inch re- 

 fractor, and, in addition to the times of occultations and 

 transits, remarks are added as to the appearance of the 

 satellite, the phenomena of its disappearance or reappear- 

 ance, and the appearance of various belts on the planet 

 itself. Mr. Innes records that on February 16, 19 10, the 

 final occultation of J iii was long drawn out ; whereas 

 five-sixths of the satellite was occulted in 6^ minutes, the 

 remaining one-sixth took another 6m. los. When half the 

 satellite was occulted, the remaining half had the appear- 

 ance of a close double star alongside Jupiter's edge. 

 Satellites i and in were occasionally remarked to be oval 

 rather than round, and several spots and markings were 

 seen on their discs. An unpredicted partial transit of iv 

 across the N. pole of Jupiter occurred on August 14, 1910. 



A CONFIRMATION OF THE DISINTEGRATION 

 THEORY.^ 



TT is probable that the transition from radium through 

 ■^ the emanation to radium D involves the loss of four 

 a particles, that is, four atoms of helium. The atomic 

 weight of radium may now be taken to be 226-4, ^"<i if> 

 on changing into niton, one a particle is lost, it is to be 

 expected that the atomic weight of niton should be 222-4, 

 for 226-4 — 4 = 222-4. But attempts to estimate the density 

 of niton by determinations of its rate of diffusion have in 

 most cases yielded the value 176 to 180, though Perkins, 

 comparing the diffusion-rate with that of mercury vapour, 

 obtained the value 235; and Debierne, using Bunsen's 

 method of causing the gas to issue through a minute hole, 

 arrived at the value 220. Undoubtedly the emanation 

 belongs to the series of the inactive gases, and to com- 

 plete the series — helium, 4 ; neon, 20 ; argon, 40 ; krypton, 

 83 ; and xenon, 130 — there is room for two higher members 

 with atomic weights 178 and 222-4. 



It might happen that, in the disintegration of radium 

 to niton, a non-radio-active substance might be produced 

 of atomic weight 44 ; the change would then be : — 

 radium (226-4) = helium (4)-f- (say) scandium (44)-|-niton (178-4). 



The only certain method of ascertaining the molecular 

 weight of a gas is the determination of its density ; and 

 in this case it is almost certain that the gas is monatomic, 

 and that its molecular and atomic weights are identical. 

 This constant has now been determined by the help of a 

 balance closely resembling one recently described by Steele 

 and Grant in the Proceedings of the Royal Society. 



For details of the construction and use of the balance, 

 the original paper must be referred to ; suffice it to say 

 here that its sensibility is about two or three millionths 

 of a milligram. The weight is ascertained by the altera- 

 tion of the pressure in the balance-case, thus altering the 

 buoyancy of a small bulb of silica containing about 20 

 cubic millimetres of air, the weight of which is 0-027 

 milligrani, or 27,000 millionths of a milligram. 



A preliminary experiment, in which 0-0977 cubic milli- 

 metre of xenon was weighed, gave its weight as 578 

 millionths of a milligram instead of the calculated 577 ; it 

 was thus shown that fairly good results might be expected 

 in determining the density of the emanation. 



^ " The Density of Niton (Radium Emanation) aud the Disintegration 

 Theory." Bv R. Whytlaw Gray and Sir William Ramsay, F.R.S. Abstract 

 of paper read before the Royal Society on January 12. 



NO. 2155, VOL. 85] 



In a month, the emanation may be taken as having 

 wholly changed into its degradation products, the chief ot 

 which is radium D ; and an experiment was made in which 

 a minute density-tube was left on the balance for three 

 months before it was opened, evacuated, and reweighed. 

 The loss was helium, and its weight was 27 millionths of 

 a milligram ; the calculated weight, on the assumption 

 that the density of niton is 222-4/2 = 111-2, and that each 

 volume of the emanation yields three volumes of niton 

 on disintegrating, should have been 38 millionths. This 

 helium, judging from previous experience, had probably 

 penetrated the glass of the density-tube and been retained 

 there. The tube was therefore heated in vacuo, and the 

 evolved helium washed out with a cubic centimetre of 

 oxygen ; the gases were transferred to a measuring 

 apparatus, and after absorbing the oxygen by charcoal 

 cooled with liquid air, the helium was measured. Calcu- 

 lating the volume to weight, its weight must have been 

 8 millionths ; and the sum of 8 and 27 gives 35, instead 

 of the calculated 38 millionths of a milligram. A further 

 proof is thus given of the conclusion drawn by Ramsay 

 and Soddy from the measurement of the volume of niton, 

 and of the hefium into which it changes, that the latter 

 is three times the former. 



Five determinations of the density of niton were made;; 

 stated as atomic weights, the figures are : — 227, 226, 225, 

 220, and 218; the mean is 223. This number is the one; 

 calculated on the assumption that when radium disinte- ' 

 grates, the only immediate products are niton and helium, 

 226-4 = 222-4-1-4. 



In suggesting the name niton for the cumbrous ex- 

 pression " radium emanation," the authors point out that 

 it is advisable to indicate by a similar name the fact that 

 this gas belongs to the argon series ; were its radio-active 

 relations to be emphasised, as in the term " radium 

 emanation," it would be necessary to rename radium as 

 a derivative of uranium by some such name as woul<f 

 introduce the word uranium. 



The authors regard the work as a further proof, if anj 

 were needed, of the beautiful disintegration theory 

 Rutherford and Soddy. 



SAFETY LAMPS AND THE DETECTION OF 

 FIRE-DAMP. 



Wl^ have received from the Home Office a leaflet and 

 '* a card in a convenient form for carrying about in 

 the pocket, upon which are shown, reproduced in colour, 

 the appearances presented by the miner's lamp in the pre- 

 scence of fire-damp. The difficulty of reproducing the appear- 

 ances presented by a fire-damp " cap" in the safety lamp 

 is very great, but it must be admitted that the illustrations 

 issued by the Home Office are of a very high standard of 

 excellence, whether considered from the artistic or from 

 the technical point of view. Necessarily, these illustra- 

 tions suffer from various defects : the Home Office does 

 not state what class of lamp was employed or the nature 

 of the oil burnt in it, and it is a well-known fact that j 

 these conditions influence greatly the nature and appear- ; 

 ance of the cap. It is, for example, very well known that f 

 the Wolf lamp, burning benzene, is more sensitive than an 

 ordinary Massant lamp burning, say, colza, or a mixture 

 of colza and mineral oil. 



We very much doubt whether one man in ten would b 

 able to see I5 per cent, of fire-damp, as indicated on tl; 

 card, the lower limit of visibility with most men bein. 

 about 2 per cent. It is, of course, well known that men 

 eyes differ very considerably in the power of seeing the- 

 faint caps, and the representations here given are of cap 

 as they appear to a man whose eyesight is well develop- 

 by training and well fitted by nature for seeing the? 

 delicate phenomena. It is a pity that the Home Offic 

 has not directed the attention of miners more strong!} 

 upon the card, in the same way as it has done in it> 

 leaflet, to the danger attending far smaller proportions of 

 fire-damp than the lamp can detect in the presence of 

 coal-dust. 



It is to be feared that the issue of the card without sue' 

 a caution as we have referred to, will induce among miner 

 the fixed opinion that they are perfectly safe so long a> 

 their lamp shows no cap. But it is well recognised that a 



1 



