NATURE 



[July 4. 189; 



ARGON AND HELIUM IX METEORIC IRON. 



T N the light of the new discoveries of argon and helium, 

 ■»■ it appeared that the reinvestigation of the gas 

 evolved on heating meteoric iron might promise interest- 

 ing results. This anticipation has been fulfilled. Meteoric 

 iron, heated in vafuo, yields a small amount of both argon 

 and helium, besides a comparatively large quantity of 

 hydrogen. 



The investigation of gases occluded in meteoric iron 

 was undertaken by Craham in 1867 (Proc. R. S.. xv. 502). 

 From 452 grams of a specimen of iron from Lenarto, 

 in Hungary-, Graham obtained, by heating it //; 

 vacuo, i6"53 c.c. of gas, consisting of 8568 per cent, 

 of hydrogen, 4"46 of carbonic o.xide, and 9S6 per cent, 

 of " nitrogen.'' .And eight years later. Prof. .Mallet 

 investigated the gases from a specimen of meteoric iron 

 from .Augusta County, \'irginia, and found 3583 per cent, 

 of hydrogen, 38"33 per cent, of carbonic oxide, 975 per 

 cent, of carbonic anhydride, and ib'oq per cent, of 

 " nitrogen." iProc. R. .S"., xx. 365.) 



In the same year. Prof .A. \V. Wright examined 

 spectroscopically the gases evolved from two meteorites, 

 one the " great Texas meteorite " in the museum of Vale ' 

 College, which weighs 742 kilograms ; another a specimen ; 

 of meteoric iron from Tazewell County, Tennessee ; and 

 a third set of experiments was made with fragments of a 

 meteorite from .Arva, in Hungar)-. The gases obtained 

 were examined spectroscopically, and were found to show 

 the usual spectra of hydrogen, carbon compounds, 

 oxygen, and nitrogen. He was searching for lines present j 

 in the spectra of stars, but found none ; and he con- 

 cludes that the spectrum of the solar corona is to be 

 ascribed merely to atmospheric gases. .A preliminary 

 account of the examination of a fourth I'a stony) meteorite 

 is given in the same journal iAiner. foiirnal of Science 

 [3]. ix. pp. 294 and 459,1, and the full account in vol. x. 44. 

 Suffice it to say that the last fractions of gas evolved con- 

 tained 6'gi per cent, of "nitrogen." On p. 257 of the 

 next volume 'xi.j. Prof Wright gives analyses of the gases 

 from various samples of meteorites, which contain from 

 I '54 to 5-38 per cent, of " nitrogen." .And lastly, in vol. xii. 

 p. 165, he gives further details, including descriptions of 

 spectra, in none of which he noticed anything unusual. 



Prof. Wright's interesting papers are instructive, 

 inasmuch as they show how little reliance is to be placed 

 on the evidence of the spectroscope as to the presence of 

 any one gas in a gaseous mixture consisting of a large 

 proportion of other gases. There is no doubt that in 

 future, much attention should be paid to the relative con- 

 ductivity of gases. The characteristic spectrum of argon 

 is almost completely masked by the presence of a few 

 parts per cent, of nitrogen or of hydrogen ; and that of 

 helium is similarly affected, although to a less degree. 

 Though no quantitative experiments have been made 

 on the subject, yet I should judge that the presence 

 of from 5 to 10 per cent, of nitrogen entirely obscures the 

 characteristic yellow line ; the other strong lines still 

 remain visible. I hope soon to be able to communicate 

 further information on this interesting subject. 



The presence of both argon and helium has been 

 demonstrated in the meteorite from .\ugusta County, 

 Virginia, a sample of which was purchased from Mr. 

 Gregor)'. Two ounces of turnings of this meteorite were 

 heated to bright redness in a hard glass tube, all air 

 having been first removed in the cold by a Tiipler's pump. 

 From this iron, 45 c.c. of gas were obtained. It was 

 mixed with oxygen in a gas burette, and exploded. It 

 appeared to consist for the most part of hydrogen. After 

 absorption of any carbon dioxide and the excess of 

 oxygen with alkaline pyrogallate. the residue amounted to 

 atxjut half a cubic centimetre. It was transferred to a 

 small tube and dried with a morsel of solid caustic 

 potash, and with it several vacuum tubes xvere filled. The 



NO. 1340, VOL. 52] 



spectrum showed that it consisted for the most part of 

 argon ; the trace of nitrogen which appeared at first 

 rapidly disappeared under the influence of the discharge. 

 The spectrum was carefully compared with that given 

 by a tube of atmospheric argon, provided with magnesium 

 electrodes. This sample of argon always shows the U 

 lines of sodium, owing to the magnesium electrodes, and 

 proves especially convenient for the detection of helium, 

 the yellow line of which is not coincident with the lines of 

 sodium. Both spectra were thrown into a two-prism 

 spectroscope at the same time, and on circful comparison 

 it was evident that all the argon lines were present. 

 Besides these, the yellow line Dj of helium was faintly 

 visible, not coincident with the sodium lines; and on 

 cornparing the spectrum of the gas directly with that of 

 helium from cleveitc, it was possible to recognise the 

 identity of the red, blue-green, blue, and violet lines of 

 helium in the meteoric gas. No other lines were visible 

 than those of argon and helium. It may thus be con- 

 cluded, on spectroscopic evidence, that both argon and 

 helium are contained in meteoric iron, the former in much 

 larger amount than the latter. This conclusion was veri- 

 fied by mixing about 90 per cent, of argon with 10 per 

 cent, of helium. The spectrum of helium, under these 

 circumstances, was much more brilliant than that of 

 argon ; hence it may be concluded that less than 10 per 

 cent, of this gaseous residue consisted of helium. 



It appeared likely that metallic iron might be induced 

 to absorb argon. It had been noticed, last tictober, that 

 in attempting to prepare argon by passing atmospheric 

 nitrogen through iron tubes filled with magnesium turn- 

 ings, and heated to redness, a smaller quantity of argon 

 than usual was collected. This rendered it not improb- 

 able that iron at a red heat is permeable to argon. If 

 permeable, then it might be permanently absorbed. .An 

 experiment was therefore undertaken by .Mr. Kellas, to 

 whom 1 have to express my indebtedness, to ascertain 

 whether finely divided iron, obtained by the reduction of 

 ferric oxide in hydrogen, would occlude argon, 



-About 14 grams of the finely divided iron was placed 

 in a combustion-tube, the capacity of which was 53'6 

 c.c. The tube was connected by a three-way stop- 

 cock to a Sprcngel's pump and to a water-jacketed 

 reservoir containing argon over mercury. .After exhaust- 

 ing the tube, argon was allowed to enter, and the tem- 

 perature was slowly raised to 600 and maintained for 

 three hours. Until equilibrium of temperature had been 

 established, no perceptible change of volume could be 

 noted. The tube was allowed to cool, connection with 

 the argon reservoir was closed, and the gas was pumped 

 off. The volume, corrected for temper.iiure and pressure, 

 was 54'2 c.c. On heating the tube, about 59 c.c. of gas 

 was given off; it was collected in three fractions, (c/), (b\ 

 and (f), the heating having been continued for twelve 

 hours. 



[a) The volume of this gas was 30 c.c. It was i:ol- 

 lected at about 200" C. This was exploded with oxygen ; 

 and a residue was obtained, of which the greater p.irt dis- 

 solved in caustic jjotash, showing that the gas had con- 

 sisted of hydrogen and hydrocarbons. The final residue 

 was 17 c.c. 



{b) The second fraction, collected at 450', amounted 

 to 15 c.c, and after treatment as above, the residue was 

 o"25 c.c. This residue was united with that from C(/), 

 and a vacuum tube was filled. The flutings of carbon 

 were visible, and also a trace of hydrogen, but no argon, 

 This gas was sparked with l c.c. of oxygen, antl on absorb- 

 ing the excess of oxygen with alkaline pyrog.illate, 045 c.c. 

 remained. On transferring this residue to a vacuum 

 tube, the banded spec trum of nitrogen was alone visil)le. 

 ic) The third fraction, collected at a red heat, also 

 showed only the spectrum of nitrogen, when purified 

 and transferred to a v.icuum lube, and on continuing 

 the discharge it also disappeared and the tube became 



