THE

AMERICAN

JOURNAL OF SCIENCE AND ARTS.

[THIRD SERIES.]

ART. XXXI.-On the Gases contained in Meteorites; by ARTHUR W. WRIGHT, Professor of Molecular Physics and Chemistry, in Yale College.

IN an article published by the writer in this Journal, for July, 1875, an account was given of an examination of the gases obtained from the meteorite of Iowa County, Iowa, which fell on February 12, 1875. This meteorite is of the ordinary stony kind, containing 12.54* per cent of nickeliferous iron, and the investigation was undertaken chiefly with a view to ascertain whether the spectrum of the gases evolved from such a body, by the application of heat, would afford any information respecting the recent theories connecting such meteorites with the comets. An analysis of the gases obtained at moderate temperatures developed the unexpected fact that their chief constituent was carbon dioxide, with a small proportion of carbonic oxide, these two gases constituting more than nine tenths of the product evolved at a temperature of 250°, and nearly one half of that given off when the heat was just below redness. As was to be expected from such a composi tion, the spectrum obtained from the earlier portions of gas given off was chiefly that of the carbon compounds, and showed a very close resemblance to those of several of the

comets.

Among the conclusions drawn from the investigation, it was stated, that the nature of their gaseous contents establishes a marked distinction between the stony meteorites and the irons * Analysis of Prof. J. L. Smith, this Journal, III, x, p. 362. AM. JOUR. SCI.-THIRD SERIES, VOL. XI, No. 64.-APRIL, 1876.

"hitherto examined," provided the Iowa meteorite could “be taken as a representative of its class."* With a view to obtain data for a more extended comparison, the investigation was continued, and a number of meteorites of both classes examined. The results of this work are given below, and it will be seen that they tend to justify completely the conclusions in my former paper, so far as any limited number of determinations could do so.

The method of experiment was the same as that described in the former paper, except in some of the minor details, and need be but briefly described here. The specimen to be examined was placed in a tube of very hard and refractory glass, which was merely softened at a red heat, and which, when filled with the meteoritic substance, could be maintained for a long time at this temperature without yielding more than so much as merely to deform the tube. In no instance was air admitted by the cracking or drawing in of the hot glass. The air was exhausted and the gas collected by means of a Sprengel pump of such perfection that it would produce a vacuum of but a fraction of a millimeter, and maintain it for days unchanged. The specimen tube having been attached to the pump, the latter was set in action and kept running until the air was thoroughly removed, as could be seen by the state of the gauge. The meteorite was then heated cautiously and the gas pumped out into the tube in which it was to be examined. Further details of the mode of procedure, where varied in the different cases, will be given in their appropriate places.

The problem of determining the exact nature and relative proportion of the gases in a meteorite is less simple than it might at first sight appear. For not only, as Grüner has shown,† is metallic iron attacked by carbon di-oxide, but it also, in the presence of this gas, or other oxidizing agents, determines the reduction of carbonic oxide, and its disappearance therefore from the gaseous products. In the case of the stony meteorites the question is still more complicated, as there is always present a greater or less quantity of oxide of iron, which at an elevated temperature must exert no inconsiderable influence upon the constitution of the gaseous mixture obtained from the mass. Grüner's very careful experiments showed that pure carbonic oxide progressively reduces the oxide of iron, at a temperature of 400° C. On the other hand it is itself reduced by metallic

*This conclusion has been criticized as hasty by Prof. J. W. Mallet (this Journal, III, x, 206), and a second one by M. M. Delafontaine, (Bibliothèque Universelle, Oct., 1875, 188), both of whom have overlooked or ignored the fact that they were given as merely provisional, conditioned upon the assumed general agreement of other iron and stony meteorites, as respects the gases derived from them, with those to which the statement referred.

+ Comptes Rendus, lxxxiii, p. 28, et al., lxxiv, p. 231, etc.

iron, with a deposition of pulverulent carbon, though the action is very slight at temperatures less than 400° C. The commission who reported upon his memoir, in repeating some of his experiments, found that the temperature must exceed 350° in order that this effect may be produced at all. At higher temperatures the action is very marked. More recently Sir I. Lowthian Bell, in his work containing the results of a very elaborate and admirable series of researches upon the mutual action of the two oxides of carbon in the presence of metallic iron and oxide of iron,* has, in the main, confirmed Grüner's conclusions, but has shown that the results vary, not only with the temperatures, but also with the relative proportion of these substances present. He found that pure carbonic oxide begins to reduce Fe,0, at from 140° to 200° C., according to the substance used, while at a moderate red heat the oxygen is rapidly removed, with the formation of carbon di-oxide. On the other hand the latter gas was partially reduced by spongy iron at a low red heat, with the formation of carbonic oxide. We have further to consider the action of the hygroscopic moisture upon the metallic iron, as well as the mutual action of hydrogen and oxide of iron, at elevated temperatures.

It is very evident then that the composition of the gases obtained at or above the temperature of red heat cannot be considered to represent accurately the true constitution of the gaseous contents of a meteorite, and especially is this true in the case of the stony ones. On the other hand we can hardly assert with confidence that the different gases are expelled in exactly their proportionate amounts at all temperatures. In fact the experiments show that the proportions of the gases vary with the temperatures of their evolution in a manner not satisfactorily explainable on the assumption that such an effect is due to chemical action alone. It is important therefore that the experiments should be conducted in such a way as to facilitate as much as possible the evolution of the gases, while at the same time they are exposed for as short time as possible to the action of high temperatures. The first of these conditions is attained in a good degree by reducing the material examined to a state of minute subdivision. The second is approximated by continuing the application of the high temperatures for the shortest time consistent with a satisfactory effect in driving off the gases sought.

In the case of the iron meteorites the material was generally prepared by boring out the solid iron with a steel drill upon a lathe, the substance being rendered as fine as possible. In * Chemical Phenomena of Iron Smelting.

some instances this was not practicable from deficiency of material, and chips produced by a planing machine were used. The stony meteorites were reduced to powder in a diamond. mortar. The iron contained in them being for the most part in very minute particles no further operation was necessary in this case. The powder from the irons, when the tube containing it was deprived of air, gave off a small quantity of gas from the mere diminution of pressure, without the application of heat, in one instance enough having been evolved to allow of its collection in a tube. A qualitative examination of it showed that hydrogen and the oxides of carbon were present, leaving no doubt that the mere pulverization of the iron was sufficient to cause it to part with a portion of its gaseous contents at ordinary temperatures, and greatly to facilitate the process at higher temperatures.

The heat was applied by means of a Bunsen burner, carried slowly back and forth beneath the tube, which was wrapped with wire gauze. For the irons the temperature was carried, in the first instance, to a point below redness, in order that the action of the iron upon the gases should be as little as possible. It was about 500° C. The gauge was watched during the heating and, as soon as it ceased to rise perceptibly, the flame was slowly withdrawn, and the gas at once pumped out. The evolution of gas, at this temperature, generally ceased very nearly in twenty or thirty minutes. After the gas was thoroughly removed, the iron was heated to redness with a cluster of four Bunsen burners, the heat being continued as long as any considerable amount of gas appeared to come away. This required usually but thirty or forty minutes, though in one or two instances it was continued somewhat longer. It will be seen from the results given below that the larger portion of the gas was obtained at the lowest temperature, in every instance but one. The iron meteorites examined were the following: First, that from Tazewell Co., Tennessee, described by Professor J. L. Smith, in this Journal, II, xix, 153. Its composition is Fe, 83-02; Ni, 14-62; other substances, 1.93. No carbon was found. Specific gravity 7·9.

Second, that of Shingle Springs, Eldorado Co., California, described by Professor B. Silliman, this Journal, III, vi, 1. It contains Fe, 81-48; Ni, 17·17; C, 0·07, other substances, 1.27. Sp. gr. 7-875.

Third, the meteorite of Arva, in Hungary, noticed in this Journal, II, viii, 439. The analysis of A. Löwe gives Fe, 90-471; Ni, 7.321; residuum of carbon, silica, and cobalt, 1.404. Sp. gr. 7-814. Another analysis by Bergemann, Pogg. Ann., c. 256, gives for its composition, exclusive of the sulphide of iron contained in it, Fe, 82-25; Ni, 8-12; Co, 0.364; P, 0·74; C, 1.54; Graphite, 2.00.

Fourth, the great Texas meteorite in the cabinet of Yale College, described by Professor C. U. Shepard, this Journal, I, xvi, 216, also, with an analysis, by Professors B. Silliman, and T. Sterry Hunt, this Journal II, ii, 370. It contains Fe, 90-91; Ni, 8·46; residue containing carbon, 0:50. Sp. gr. 7.543.

Fifth, that from Dickson Co., Tennessee, described by Professor J. L. Smith, this Journal, III, x, 349, and examined at his request. It contained Fe, 91·15; Ni, 8·01; Co, 0·72; Cu, 0-06. Sp. gr. 7-717.

The following table gives the results obtained, the numbers expressing parts in one hundred. The numbers in the third line in each case give the percentage of each gas in the total amount obtained. They are not the simple averages of the numbers above them, but the means reduced according to the volumes in each case. The totals in the last column are the sums of the volumes given off at the different temperatures.

The small quantity of the iron available in the examination of the Dickson Co. meteorite rendered it necessary to be content with a single heating to redness. The iron was in the form of coarse chips which were cut by a planing tool. The same was true of the Shingle Springs iron, and this accounts in part for the smaller volume of gases obtained in these two

cases.

We may add to this list the Lenarto iron examined by Professor Graham, and the meteorite of Augusta Co., Virginia, the gases from which were analyzed by Professor J. W. Mallet.+ The former yielded CO, 4:46; H, 85.68; N, 9.86, the whole amount of gas being 2.85 times the volume of the iron. The + Ibid., xx, 365.

*Proc. Royal Soc., xv, 502.