142 



NA rURE 



[June 7, 1894 



the spectrum of another, which I have recently observed, not 

 only strengthens weak lines, but in certain cases brings a new 

 series of lines into view. Thirdly, the spectra of mixed vapours 

 have been shown to be ditVerent from the spectra of the sub- 

 stances by themselves (Liveing and Dewar, " Roy. Soc. Proc," 

 vol. xxxiv. p. 42S) ; and, fourthly, the influence of the strong 

 lines of an element on adjacent weaker lines of another sub- 

 stance is to strengthen the weaker lines in some cases, but 

 almost 10 obliterate them in others. 



Variations in the spectrum of carbon as observed in different 

 circumstances have been carefully examined and described. In 

 order to lest the probability of the carbon and nitrogen speclra 

 being subject to variations when the two elements aie together 

 in the spark or flame, it is necessary to consider the effect of 

 one spectrum on another when the two are produced simul- 

 taneously from quite different materials. 



In the oxyh>drogen flame the water-vapour lines are pro- 

 minent, but only two groups are visible in ihe spectrum under 

 normal conditions, and with an exposure of half an hour. If, 

 however, some sulphur be burnt in the flame, the conditions 

 being otherwise unchanged, then the spectrum, in addition to a 

 band of continuous rays and fluiings characteristic of sulphur 

 vapour, shows the water-vapour lines wonderfully strong, with 

 groups extending beyond those portions of the spectrum usually 

 photographed, and not only are the lines disiinci, but dense, as 

 if their radiating power or the chemical action of their radiations 

 was greatly increased. This does not arise from the continuous 

 spectrum merely overlapping and apparently strengthening the 

 water-vapour lines, since new groups of lines came into view 

 which were too feeble to be visible on the other photographs. 

 Sulphur is not the only substance which affects this spectrum ; 

 for instance, the banded spectrum of magnesia and the spectrum 

 of lime also appear to intensify it. 



It is probable that someihmg similar lakes place with regard 

 to carbon ; we know that the spectrum is modified by the sur- 

 rounding nitrogen of the atmosphere, and the rays of carbon 

 increase the intensity of the nitrogen rays adjacent to the 

 carbon lines, the effect being increased in the case of the spark 

 by a saturated solution of zinc or calcium chloride. 



The facts here set forth cettamly favour the view that the lines 

 in Hartley and Adeney's spectrum of carbon are the lines of the 

 element and not merely the edges of cyanogen bands. Finally, 

 the carbon spectra of tder and Valenia differs from that pub- 

 lished in ihe Journal of the Chemical Society, vol. xli. p. 91 ; 

 the graphite spectrum. No. 10, on plate ii., yields neither the 

 group III. nor group IV. of cyanogen as depicted in spectrum 

 No. 4 of the photogravure plate illustrating Eder and Valenta's 

 paper. 



" Experimental Determination of Poisson's Ratio." By 

 C. E. Stromeyer. 



The experiments with which this paper deals were carried out 

 between the years 1SS3 and iSSb by Prof. Kennedy and the 

 author, with an instrument which the latter had originally de- 

 signed for measuring local strains in metal structures, but which 

 proved itself to be so exceedingly sensitive that it was capable 

 of being applied to the measuring of the cross contraction of 

 test pieces while these were subjected to a longitudinal pull, 

 thus providing the means for measuring Poisson's ratio direct. 



The conclusions drawn from the experiments with nineteen 

 samples are ; — 



(1) That Poisson's ratio is not a constant value for all 

 materials. 



(2) That mechanical treatment (cold rolling and annealing) 

 of the metal alter it. 



(3) TTiat Poisson's ratio is sometimes a function of Ihe stress. 



(4) That Poisson's ratio, as found by direct measurement, is 

 not the same as that found by comparing torsion and tension 

 experiments. 



May 24. — "Some Voltaic Combinations with Fused 

 Electrolytes and Gaseous Depolariscr." IJy. J. W. Swan. 



In this paper arc described several voltaic combinations in 

 which fused electrolytes and a gaseous depolariser were used. 

 The electrodes were the same in all the experiment.', viz. lead 

 in a fused stale as the positive, and carbon as the negative. 

 The electrolyte used in Ihe first experiments was a fused 

 mixture of KCI,Na(JI, but this was changed for one of PbClj. 

 The depolarising gas used in all the cxpcnmenis was chlorine, 

 and was so applied as to chemically act on the electrolytic 

 products formed at the carbon pole. Several methods of 



applying the gas were employed : by forcing the gas through 

 ixirous carbon, by making the carbon pole tubular and feedini; 

 the chlorine through it, and by nearly wholly surrounding the 

 carbon pole by an atmosphere of chlorine. 



The condition found to be most necessary for successful 

 depolarisation was /.> altertuittly expose the carbon pole to the 

 action of Ihe gas ami electrolyte in rapid succession. During the 

 electrolytic action, the lead dissolves as chl iride of lead, and 

 lead is deposited on the carbon pole, but is immediately 

 reconverted to chloride by the action of the chlorine gas. The 

 theoretical E. M.F. corresponding to the union of Pb and Cl.j is 

 I 7942 volts, the highest obtained was 1 40 volts, this lower 

 E.M.F. was probably due to the incomplete solution of the 

 lead at the carbon electrode. 



A noticeable feature of this kind of cell is the very low 

 internal resistance, which makes it possible to obtain a large 

 current density with comparatively small electrodes. In one of 

 the experiments a current of 10 ampiic was given with an 

 area of the carbon of 10 to 12 sq. cm. It was also observed that 

 the internal resistance, and at the same time Ihe polarisation, 

 decrease, when the electrical output increases, and that the cell 

 gave an almost constant effect. The best results were obtained 

 with small cells, the action of the chlorine being then more 

 effective than when larger cells were employed. Experiments 

 were also made with oxygen as a depolariser, but a description 

 of them is left for a future paper. 



Physical Society, May 25. --Prof. A. \V. Riicker, F.R.S., 

 President, in the chair. — Prof. \V. Ramsay, F.R.S., read a 

 paper on the passage of hydrogen through a palladium 

 septum, and the pressure which it produces. After referring 

 to the analogy between osmotic pressure of solutions, and the 

 behaviour of hydrogen and palladium, ihe author described the 

 apparatus he had used in his experiments, and showed it in 

 operation. A vertical platinum tube provided with a palladium 

 cap is enclosed within a glass vessel through which hydrogen or 

 other gases may be passed, and outside the glass vessel is a 

 vapour jacket, by means of which a constant temperature can be 

 maintained. The lower end of the platinum tube communi- 

 cates through a graduated capillary tul)e, with adjustable 

 manometer, which enables the volume of the enclosed gas to be 

 kept constant. Great precautions were taken for ensuring 

 purity and dryness of the g.ases used. After filling the 

 palladium and tilatinum lube with dry nitrogen at atmospheric 

 pressure and the desired temperature, hydrogen was passed 

 through the glass vessel. .Some of the hydrogen permeated the 

 palladium walls, thus increasing Ihe pressure inside. Afier 

 some lime (usually an hour or so) the pressure attained a steady 

 value, and the total increase was then observed. Experiments 

 were made with air, nitrogen, nitric oxide, nitrous oxide, carbon 

 dioxide, carbon monoxide, and cyanogen in the palladium lube, 

 and in some cases the hydrogen was diluted with nitrogen. In 

 all cases the maximum pressure of the hydrogen within the tube 

 was less than that of the hydrogen outside the tube, as will be 

 seen from the following table, which shows the ratio of these 

 pressures under various conditions : — 



Gas originally Gas passed outside 

 inside tube. tube. 



Nitrogen 



t^rbonnioxidc 

 ,, Monoxid< 

 Cyanogen 



I InlcrnnI tiyilrojrcn pressure 



Temp.! Ratio ji„„„j| nyj^o^en pressure 



Hydrogen 980 C.| 



• 1 |3>5 ; 



,. 5o7o('estN) „ 

 ■ • >S 1. .1 ■■ 



Hydrogen 280 



OVOS3 

 08984 

 0*9362 

 o«344 

 ©■962 c 

 09545 

 og6i):) 



NO. I 284, VOL. 50] 



After the palladium had been used once or twice it became 

 coated with mercury (vapourised from Ihe manometer), and lost 

 its permeable properties. It was found necessary to heat the 

 tube to remove the mercury, and then dissolve off the oxide 

 of palladium thus produced, after each experiment. The 

 perinealile nature of palladium was found to depend greatly on 

 the temperature, for at 232' C. the passage of hydrogen was so 

 slow that the internal pressure was still rising afier ten days ; at 

 330 the passage was very rapid. An attempt to use the 

 apparatus for giving a continuous indication of the amount of 

 hydrogen in coal-gas failed because the palladium did not 

 retain its activity sufficiently long. Uther experiments showed 



