and should it be erroneously supposed that the gram weights used were of brass, the total error in the correction would lie between 0.000012 and +0000187 with a mean value of Correction of the weight of a gram for Aluminium is sometimes employed for weights, but, so far as I know, only for the smaller fractions of a gram. The use of aluminium for milligram weights would diminish the error arising from the neglect of the correction for small fractions. The influence of changes of temperature on the correction for vacuum is readily ascertained. If we call the variation 2 caused by temperature from 21the correction as calculated for 15°, x1; then 23 22 20 19 18 17 16 C being the weight of one 15 vacuum. (Platinum weights.) 1 2 3 4 5 6 7 8 9 10 11 12 13 Mg. A variation of an inch in the height of the barometer would affect the correction still less, and the effects of ordinary elevations and of latitude or the density of the atmosphere are evidently too minute to be taken into consideration in the present state of chemical apparatus. In conclusion, attention may be drawn to the fact, which appears very plainly in the table, that for the purposes of the reduction of weighings to vacuum, no very accurate determination of the specific gravity is necessary. Berkeley, California, May, 1878. ART. XXX. On the composition of the new Meteoric Mineral Daubréelite and its frequent, if not universal, occurrence in Meteoric Irons; by J. LAWRENCE SMITH, Louisville, Ky. WHEN I first announced the discovery of the mineral daubréelite, the amount at my disposal was only sufficient for the determination of its specific characteristics. Since then I have made numerous sections of the first iron, which weighed about 250 kilograms, and also sections of another iron of 200 kilograms from the same locality (Cohahuila), and in this last have found the nodules even more abundant than in the first. Of the second iron, I have a section with two polished surfaces of about 900 square centimeters each, which show twenty-five to thirty nodules, varying from three to sixteen millimeters in diameter, at least ten of which are from one to one and a half centimeters in diameter, and all of them exhibit to the eye daubréelite in angular segregations. The mineral used for my first analysis I obtained by breaking it out from the nodules mixed with troilite and other impurities, depending on the eye to separate the impurities. Since then, I have found that chlorhydric or fluohydric acid will attack the troilite readily and not act on the daubréelite, and thus a method has been adopted by which the mineral is obtained more abundantly and quite pure. The shavings and cuttings procured in making the sections were used (several kilograms of which were at my command); the fragments of iron were separated by a large magnet, and the small particles left behind consisted essentially of troilite and daubréelite; for the former is only feebly magnetic and the latter not at all so. Strong chlorhydric acid is next added to this last portion and gently warmed over a water bath; the troilite is readily attacked; after a time the first acid is poured off, and a fresh portion added, and the digestion continued over a water bath from one to three hours; the residue contains a good deal of light black matter that is easily washed away (this last has not yet been thoroughly examined, but much of it is impalpable daubréelite); the larger black particles are again treated with a little chlorhydric acid, after which the daubréelite is left quite pure and is easily washed, and any foreign particles are readily picked out. In this form it consists, as already described, of shining black fragments more or less scaly in structure, not altogether unlike fine particles of molybdenite. The fracture is uneven, except in one direction where there appears to be a cleavage. It is *This Journal, III, xii, 109, August, 1876. brittle and easily pulverized, the fine particles retaining their brilliancy. It is not magnetic. Before the blowpipe it undergoes but little alteration, losing its luster, but not fusing, and after heating in the reducing flame it is slightly magnetic. With borax it fuses slowly, the smaller particles giving an intense green color to the bead when cold. It is not acted upon in the slightest degree by chlorhydric acid, either cold or hot, but dissolves slowly and completely in nitric acid when heated over a water bath, without, however, any liberation of free sulphur. Its specific gravity is 501. It is needless to give the details of the method of analyses. I will only remark that when a mixture of hydrated oxides of chroinium and iron are separated by the addition of bromine to an alkaline solution holding the oxides in suspension, the operation must be repeated two or three times to insure complete conversion of all the chromium oxide into chromic acid, and consequently to separate it totally from the iron. The following is an average of three analyses giving concor dant results within one half per cent of each constituent: A minute quantity, of what appeared to be carbonaceous matter, was mixed with the residual traces of oxides found in the mother-water. It is very evident from the above proportions that this mineral is a sulphide corresponding in atomic constitution to the well-known oxide, chromite (FeO+€r03), daubréelite being FeS+ErS3, sulphur replacing the oxygen; the calculated percentage is : Sulphur... Iron. Calculated. Found. Calculated. Found. 68.00 30:45 29.75 Sesquisulphide of chromium... 69.55 The calculation of the daubréelite is based upon the sulphur found in the analyses (43.26). As yet we do not know of any terrestrial mineral corresponding to this, and it is an interesting fact that we are already enabled to establish so clearly its true composition, and also to obtain good characteristic specimens that will find their way into the principal cabinets of meteorites. The occurrence of this mineral in so marked a manner in the Butcher meteoric irons of Cohahuila, when it does not show itself in the troilite of other meteoric irons, induced me to investigate this matter carefully, as I now had chemical methods to aid me. Thus far I have examined the troilite from only three meteoric irons, viz., those from Toluca, Mexico; Sevier, Tennessee; and Cranbourne, Australia. In the first two specimens it was found in marked quantities, about 2.5 grams of troilite being employed, but in the case of the Cranbourne, where the quantity did not amount to one gram, the daubréelite was proportionally less than in the other two. The Toluca troilite furnished the largest quantity; the residue from 2:800 grams of it, after thorough treatment with chlorhydric acid, which dissolves nearly the whole of it, was dissolved in part by nitric acid, and on analysis the solution was found to contain chromium and iron representing about sixty milligrams of daubréelite; the mineral obtained from these troilites was of the pulverulent variety.* There is reason to believe that further research will show the constant presence of daubréelite in meteorites. I am now prosecuting a series of experiments on the mineral segregations in meteoric irons, both those visible and invisible to the naked eye and those only discernible by chemical means, the results of which will tend to a satisfactory solution of this hypothesis. ART. XXXI.-On the Artificial Mounds of Northeastern Iowa, and the evidence of the employment of a Unit of Measurement in their erection; by W. J. MCGEE. WITH very rare exceptions, the artificial mounds of northeastern Iowa and contiguous parts of Wisconsin and Minnesota may be divided into four classes, viz: (1) tumuli; (2) conical mounds, similar to the tumuli in appearance, but smaller, and bearing no evidence of having been used for inhumation, usually found in rows or series, and in such cases sometimes connected by narrow ridges: (3) embankments; and (4) animal mounds. Isolated mounds are sometimes found, but they occur much more frequently in groups; and where the topography of the country is favorable, many groups may be connected, forming extensive systems. All four classes of mounds often appear in a group, and usually in a system. When exceptions occur, it is most frequently the tumul which are found to be absent. Rarely an embankment or a collection of mounds is so situated *The undissolved portion after the nitric acid treatment is principally graphite and schreibersite. as to lead to the inference that it was designed-partially, at least for defensive purposes. True fortifications, like those of Ohio and Kentucky, have not, however, been discovered. Tumuli are never, so far as the experience of the writer extends, regularly and methodically arranged. A few burial mounds occupy prominent spurs and bluffs overlooking watercourses, or other natural elevations. Such mounds usually contain the remains of but one, or at most a few, bodies, and seem to have been used only in exceptional cases. The ordinary grave mounds occur in valleys or on plains, irregularly disposed, and each usually contains the remains of several bodies. Implements, arms, etc., are not always found associated with the human remains in the tumuli. In an extensive collection of burial mounds opposite Clayton, Iowa, arrow-heads and spear-points are often found in such positions as to indicate that they were buried within the bodies. A cranium from the same locality has a horizontal indented fracture about two inches above and parallel with the supraciliary ridges, corresponding in shape to the edge of a stone axe. Here, as in other parts of the United States, one of the strata of the material forming the tumulus consists of a hard light-colored earth, so indurated as to almost if not quite prevent the percolation of water or the permeation of air. No earth of a like nature can be found in the vicinity of the mounds. Hence it must have been either transported a long distance, or artificially prepared; and the occurrence of a like stratum in the mounds of widely separated localities in which there is the greatest diversity in the superficial geological formations (the stratum being in all cases of different material from any of the natural formations), would seem to indicate that the latter supposition may be correct. The preservation of organic remains within some of the mounds is due to the impermeable nature of this stratum. Where it is not found, all organic matter has been completely decomposed. The smaller conical mounds (which are, properly speaking, spherical segments) never exhibit a stratified structure, nor do they contain relics of any kind. In height they vary from one to three feet, and in diameter from fifteen to forty feet. They are rarely-perhaps never-isolated, usually occurring in straight or slightly sinuous lines. The object in erecting them is not obvious. Interest attaches to them only from the fact that they are undoubtedly separated by measured distances. The embankments are straight, about equal in width and height to the small conical mounds, and their length is ordinarily from one hundred to three hundred feet, though they are sometimes much longer. Their dimensions, some of which are given below, are as constant, however, as the distances separating the conical mounds. Except in the very rare instances in |