A Comparative Study of the Chemical Behavior of Pyrite and Marcasite. THESIS PRESENTED TO THE FACULTY OF THE DEPARTMENT OF PHILOSOPHY OF THE UNIVERSITY OF PENNSYL¬ VANIA FOR THE DEGREE OF DOCTOR OF PHILOSOPHY. ' By Amos Peaslee Brown. ■* Read, before the American Philosophical Society, May 18, 18^4. /) Reprinted June 19, 1894, from Proc. Amer. Philos, Soc., Vol. xxxiii, \ / * 3 A Comparative Study of the Chemical Behavior of Pyrite and Marcasite. By Amos Peaslee Brown. (Bead before the American Philosophical Society, May 18, 189 4 .) "While much has been done in the way of investigating the chemical properties and constitution of the various artificial chemical compounds, comparatively little attention has been paid to the constitution of the naturally occurring chemical compounds. The carbon compounds, for instance, have in an immense number of cases been investigated with sufficient exactness to allow of our expressing their constitution by means of structural formulte, but to how many minerals, aside from the simplest compounds, can we assign structural formulae that are based on any knowledge that we possess of their reactions? It is true that much has been done in the way of the artificial production of minerals, and some knowledge of the constitution of certain minerals has been arrived at by a study of their decomposition products, but very little in comparison to the extent of the field. There are probably several reasons for this neglect of the study of the chemical properties and constitution of minerals, as want of homogeneity in the minerals themselves, difficulty of procuring or separating pure material for investigation, and similar difficulties which do not so frequently occur with artificial compounds. It thus hap¬ pens that mineral chemistry is not as much studied as it deserves to be. Certain groups of minerals have, however, received some attention ; for instance, Prof. F. W. Clarke has been carrying on a very interesting series of investigations on the constitution of certain silicates which have been productive of most valuable results. The natural sulphides, sulpliarsen- ides and sulpho-salts present some very interesting problems in regard to their constitution, and it was with a view of adding to out knowledge of the chemical behavior of two of these sulphides that I undertook the series of investigations about to be described. The compound FeS 2 is found in nature in two well-known forms—the one Pyrite (the isometric mineral) and the other Marcasite (the ortho¬ rhombic mineral). Since the separation of the two names from the gen¬ eral term, pyrites, it has been recognized that the orthorhombic form is lighter in color and also of lower specific gravity than the isometric form. From early times, also, the greater tendency of “white pyrites,” or mar¬ casite, to decompose in the air was well known. Pyrite, the form which resists atmospheric weathering most thoroughly, is of a bright brass-yellow color and metallic lustre, breaking with an uneven conchoidal fracture, but bright on the surface of fracture. It crystallizes in the isometric system in forms showing generally pentagonal hemihedrism. Its specific gravity ranges from 4.8 to 5.2, averaging some what over 5. The brass-yellow crystals are generally quite unaltered in the air. REPRINTED JUNE 19, 1894, FROM rROC. AMER. PHILOS. SOC., VOL. XXXIII. 2 Marcasite, on the other hand, has a pale greenish or grayish yellow color, an uneven lracture, which shows a somewhat fibrous structure, and generally but little lustre on the surface of fracture. It crystallizes in the orthorhombic system, very commonly in twins or radiated fibrous masses. It is not very permanent in moist air, but readily decomposes and largely into FeS0 4 . The chemical formula of either form, calculated from quantitative analyses, is the same, FeS 2 or Fe = 46.67 %, S = 53.33 %. The chemical study of these two minerals has been mainly confined to the formation of one of them artificially and to a few experiments on their relative decomposability. Pyrite has been made artificially in a number of ways ; marcasite has not as yet been formed artificially.* * * § In 1836, Wohlerf produced cubes and octahedra of pyrite by subjecting a mixture of ferric oxide, flowers of sulphur and ammonium chloride to a temperature a little above the volatilizing point of the ammonium chlo¬ ride. The resulting mass was washed to isolate the crystals from the accompanying pulverulent matter. Stanislas Meunier \ modified this method by mixing equal parts of ferrous carbonate and sulphur and heat¬ ing in glass tubes over a moderate flame. When the excess of sulphur has been driven off, there remains a black powder containing a consider¬ able percentage of cubes of pyrite. Dana § states that pyrite may be made “ by slow reduction of ferrous sulphate in presence of some carbon¬ ate.” Baubigny || produced FeS 2 as a crystalline crust by acting on metallic iron by a solution of S0 2 in water (H 2 S0 3 ) in closed tubes and at a temperature of 200°. As neither this experiment nor the one imme¬ diately preceding it shows that the crystals were isometric, it is possible that both of them may be marcasite. Henri Saint Claire Devilled pro¬ duced cubes of pyrite by melting a mixture of potassium sulphide (K 2 S) and iron sulphide (FeS) in presence of excess of sulphur. This reaction, if correct as to the cubical product, is a remarkable one, as I should rather expect marcasite to result under such conditions. Senarmont ** produced FeS 2 by decomposing a salt of iron by an alkaline sulphide at an elevated temperature in sealed glass tubes. The product was an amorphous black powder, not altering on exposure to air and not attacked by hydrochloric acid. This may have been pyrite, as marcasite is readily decomposed by moist air. Raminelsberg,tf in 1862, made FeS 2 pseudomorphs after ferric oxide (Fe 2 0 3 ) by passing a current of hydrogen sulphide over it at a tem¬ perature between 100° and a red heat. The product of this reaction would likely be pyrite. In nature it would seem that in most cases the sulphide of iron first *Doelter, Zeit.fiir Kryst., xi, 31, 1885; cf. Dana, Syst. Min., 1892. f Pogg. Ann., xxxvii, p. 238. X Les Methodes de Synthase en Mineralogie, S. Meunier, 1891. § J. D. Dana, System of Mineralogy , edition of 1868, p. 64. H S. Meunier, Synth. Min., p. 279. If Cited in Diet. Chem. of Wartz, by E. Wilm, article “ Iron,” T. i, p. 1414. **S. Meunier, Synth. Min., p. 285. ft Jour, fiir Praktische Cliemie, T. lxxxviii, p. 266. 3 formed is FeS, but probably by action of a ferric salt, or carbonic acid (H.,C0 3 ), causing a loss in iron FeS 2 results, and this is almost always pyrite. On the other hand, where ferrous sulphate has been reduced by slow action of decomposing organic matter, the resulting sulphide seems to be generally marcasite, which if not in crystals may be recognized by its ready weathering to ferrous sulphate (FeS0 4 ). This compound may, however, in many cases, be a mixture of pyrite and marcasite, as much pyrite seems to be.* These several ways in which pyrite may be formed are of importance as indicating the condition of the iron in the compound, and will be referred to later on. Equally important as bearing on the constitution of these sulphides are the observations that have been made as to their decomposition under at¬ mospheric agencies. On exposing crystallized pyrite to atmospheric weathering it is generally found that but little, if any, change takes place even in a year’s time, while crystallized marcasite, under the same condi¬ tions, shows a rapid weathering. The main product of the weathering of pyrite in nature is the compound limonite, Fe 4 0 3 (0H) 6 , which occurs in large quantities in nature, evidently derived from pyrite. Its pseudo- morphs after pyrite are very common. On the other hand, when marca¬ site weathers a very large percentage of ferrous sulphate (FeS0 4 ) is formed with a comparatively small percentage of limonite, unless the marcasite decomposes underground and under pressure, when limonite is largely produced.f Much of the excess of sulphur with marcasite is changed to sulphuric acid, but with pyrite much remains behind as sul¬ phur. Some comparison of the rate of oxidation of these two minerals in the air is afforded by an examination of specimens in a collection. Here it will be found that most of the pyrite is unchanged, but nearly every specimen of marcasite will show tarnish if no other sign of oxida¬ tion, and often a considerable coating of oxide can be seen, or a white efflorescence of FeS0 4 . Chemical investigations of these two minerals have been mainly in the way of analysis, but some experiments have been made in the way of studying their relative behavior towards certain reagents. Before de¬ scribing my experiments, it will be necessary to briefly mention some of these. Both minerals are decomposed by a solution of silver nitrate and gold chloride, the decomposition taking place quite rapidly.:}: My experiments in this connection are mentioned later. A. A. Julien \ has shown that different samples of pyrite show a differ¬ ence in their reaction with bromine vapor. His experiments consisted in exposing finely ground pyrite to the action of bromine vapor at the tem- * Compare A. A. Julien, “Decomposition of Pyrite,” Ann. N. Y. Acad. Sci., Yols. iii and iv. t Blum, Pseudomorphosen, 1843, pp. 197-199. JS. Meunier, Synth. Min., p. 309. I A. A. Julien, Ann. N. Y. Acad. Sci., Vol. iv, pp. 151,155. 4 perature of the air for twelve hours. The residue was extracted with dilute H ? S0 4 , which removed the iron rendered soluble (bromide), and the iron was then determined in this solution. The percentage of iron that had dissolved varied from 2.43 to 15.20 per cent., although all sam¬ ples tested are described as pyrite. He also tried the action of bromine in aqueous solution, but the reaction was too rapid to give any compara¬ tive results. Much more important are the results obtained in tbe oxidation of these minerals by the electric current as conducted by Prof. Edgar F. Smith,* and it was the remarkable results that were then obtained that induced me to continue the study of the comparative reactions of these two min¬ erals Smith found that a current which would completely oxidize the sulphur in marcasite in a given time would oxidize less than half of the sulphur in pyrite in the same time. This remaining sulphur was held very tenaciously, though the mineral was subjected to more powerful cur¬ rents and longer continued action than in the case of marcasite or pyrrho- tite. Finally, by adding an equal quantity of CuO, and using a more powerful current, all ot the contained sulphur was oxidized. Previous to the addition of CuO but 21 or 22 per cent of the sulphur was oxidized. In concluding the article above referred to the author questions whether the crystalline form alone can make this difference in the action of the two minerals when under the influence of the current. The two samples of pyrite and marcasite that I selected for the follow¬ ing study were chosen after considerable examination of material as being typical of the two forms of FeS 2 . The pyrite was from the hematite mines of Elba. It is exceptionally pure and free from decomposition or tarnish. Before deciding on it finally pieces were ground and polished and then examined under the microscope with powers ranging from 50 to 200 diameters, in order to see if it contained any enclosures or varied in texture. It was perfectly homogeneous and showed no enclosures. It took a high polish. The crystals showed the combination of octahedron and pentagonal dodecahedron O -{- £§-. Some of the crystals were coated in places with scales of hematite, but this was all carefully removed in breaking up material for experiment. The color was bright brass-yellow, the specific gravity was determined as 5.179. The marcasite was from the zinc mines of the Subcarboniferous of Joplin, Jasper county, Mo., finely crystallized in polysynthetic twinnings. The freshly broken crystals show a greenish yellow color, almost white, but they tarnish readily with bluish or brownish colors. No gangue was present, everything dissolving com¬ pletely in nitric acid. This marcasite was examined with the microscope in the same way as the pyrite ; it did not take such a high polish on account of a fibrous structure, but no foreign matter was found with a power of 200 diameters. Its color was uniform throughout, showing that no pyrite was present. The specific gravity as determined was 4.844. In preparing material for experiment only sufficient was ground forirn- *Jour. Franklin Inst., Vol. cxxx, pp. 152-154. 5 mediate use to avoid any cliance of oxidation of the ground material; the stock samples of the two minerals broken to nut size were kept in stoppered bottles. The grinding of material was continued as long as grit appeared, but no bolting was resorted to. As the experiments of Prof. Smith on oxidation by the electric current showed such remarkable results, my first experiments were on oxidation. As an oxidizing agent potassium permanganate (KMn0 4 ) was used, sev¬ eral strengths of which were tried for varying intervals of time with each mineral, and the amount of sulphur oxidized to sulphuric acid determined in the liquid by precipitating as barium sulphate. The object was to secure a complete series of results which would show the comparative rates of oxidation of the sulphur in the two minerals. Neutral aqueous solutions of the potassium permanganate were used, and the strengths of solution employed were normal, one per cent., three per cent, and five per cent.; the periods of oxidation extending over one, two, three, four and five hours, and the entire series being performed at ordinary temperatures and at 100°. As duplicate determinations were made in the majority of cases (I made about 130 determinations of su pliur as barium sulphate), this work consumed a large amount of time and prevented as fall a study of some other reactions of the two minerals as had been orig¬ inally intended. The following are the detailed descriptions of my pro¬ cesses and results : Action of Normal Potassium Permanganate Solution at Ordinary Temperature. These oxidations were performed as follows : Two-tentlis of a gram of the finely powdered mineral was placed in a stoppered bottle of about 100 c.c. capacity, then 50 c.c. of the permanganate solution added and the contents of the bottle violently shaken to break up lumps. This shaking was repeated about every fifteen minutes while the oxidation lasted. The temperature of the room was at the same time recorded. As stated, the oxidation was continued for one, two, three, four and five hours with each mineral, making at least ten experiments necessary for each strength of solution. After the solution had acted for the required time it was rapidly filtered through asbestos with aid of the filter pump, the filtrate transferred to a beaker, 20 c.c. of concentrated hydrochloric acid added, and the whole heated until all manganese was reduced to manganous chloride. If not too acid the solution was then diluted to about 300 c.c. and the sulphuric acid precipitated as barium sulphate. When very acid, excess of hydrochloric acid was removed by evaporation or by adding ammonia, the ammonium chloride seeming to facilitate the precipitation. The precipitate was washed with hot water and then weighed. All precipitations were made at boiling temperature and di¬ gested hot for at least two hours, and then cold for at least twelve hours more before filtering. The filtrates from most of the cold tests were re- 6 duced with metallic zinc and titrated with permanganate, but no iron was found in the solution. The two minerals did not present the same appearance when acted on by the oxidant. Pyrite retained its color and seemed as pulverulent as when the permanganate was added, but marcasite immediately on the addition of the reagent became coated with manganese dioxide, took on a brownish color, and showed a tendency to cake together and stick to the sides of the bottle, so that it was with difficulty dislodged. This tendency of the marcasite was more marked with stronger solutions of the perman¬ ganate and was doubtless the cause of much of the irregularity that will be noticed in the results. The reason for this difference in action of the reagent on the two minerals will be discussed later on. The percentages of sulphur oxidized in the two minerals by this method are shown in the following table, where all results that were obtained are recorded. The figures show the percentages of sulphur oxidized, calcu¬ lated on the basis of FeS 2 equal to one hundred per cent. It will be noted that the four-hour oxidation of marcasite shows a result that is less than the two-hour. This was due to caking of the mineral against the w alls of the bottle, which prevented much of it from coming in contact with the solution. On the whole, this series was about the most satisfactory of the cold experiments with KMn0 4 , the action of this dilute solution being less rapid and hence more even than that of the more concentrated solu¬ tions ; naturally the action ceases with a certain dilution, and hence the four- and five-hour oxidations of pyrite are about equal. Table Showing the Relative Oxidation of Sulphur in Pyrite and Marcasite by a N. Solution of KMn 0 i at 22°. Mineral. 1-Hour. 2-Hour. 3-Hour. 4-Hour. 5-Hour. Pyrite with N. solution KMn0 4 .. .78 1.17 1.88 1.74 1.72 Marcasite with N. solution KMn0 4 . 1 1.07 1.86 2.04 (1.25) 2.38 The curves formed by plotting these results on rectangular coordinates are shown in Pis. i and ii. They are marked 22° M x for the marcasite and 22° P x for the pyrite. Action of 1 Per Cent. Potassium Permanganate Solution at Ordinary Temperature. This and also the two following series were performed as described under normal solution above. At least two experiments were tried with each mineral in this and the two following cold oxidations, and whenever a result was notably higher or lower than its dupli- cate a third or fourth was tiied. The tendency of the maroasite to cake, noted in the previous series, became still more marked here, and is doubt¬ less the cause of one of the four-hour oxidations (marked by parenthesis) being notably lower than the three-hour. Such a result is obviously incorrect. On the other hand, the result in the tliree-hour column which is placed in parenthesis is the highest obtained. This experiment was made at the same time as the one showing 1.93 per cent., but the room was very warm (25°), which may in part account for this high result. It will be noticed that the oxidation of the pyrite seems to stop at the three- hour trial, those following showing no appreciable increase. This is well seen in the graphic representation of these oxidations (Pis. i and ii). Table Showing the Relative Oxidation of Sulphur in Pyrite and Marcante by a 1 Per Cent. Solution of KMnO 4 at 22°. Mineral. 1-Hour. 2-Hour. 3-Hour. 4-Hour. 5-Hour. Pyrite with 1 per cent, solution KMn0 4 cold. 1.72 1.71 1.38 1.47 1.85 1.87 1.79 1.90 1.70 1.89 Marcasite with 1 per cent, solu¬ tion KMn0 4 cold. 1.16 1.28 1.29 1.13 1.93 2.19 (3.92) 1.95 (1.56) 2.69 2.01 2.15 2.55 Action op 3 Per Cent. Solution of Potassium Permanganate at Ordinary Temperature. The conditions of this series of experiments were the same as those of llie last. The tendency of the results to fluctuate instead of show¬ ing a gradual progression is now very marked. One of the one-liour pyrite oxidations shows more sulphur oxidized than is shown by any other individual result of the series. No explanation can be offered for such a discrepancy as this. On the other hand, the high result shown in the three-hour oxidation is quite easily explained by the marcasite hav¬ ing been little, if any, caked in this experiment. The two low results of pyrite three hour and marcasite four-hour oxidations are readily explica¬ ble on the ground of caking of the material. As the barium sulphate was often determined several days after the oxidation was completed, it is obvious that no reliable notes could be made concerning the caking or non caking of the mineral in the permanganate. With this strength of solution it is evident, too, that the main action of the permanganate is complete at the end of one hour, notably in the case of the marcasite, and it is only when very vigorous agitation exposes fresh surfaces of the min¬ eral to the action of the KMn0 4 that any further action can take place. We therefore see that marcasite in one hour gives up as much sulphur as in live hours, and this is very graphically shown on PI. i. 8 Table Showing the Relative Oxidation of Sulphur in Pyrite and Marcasite by a 3 Per Cent. Solution of KMnO i at 22°. Mineral. 1-Hoor. 2-Hour. 3-Hour. 4-Hour. 5-Hour. Pyrite with 3 percent, solution KMn0 4 cold. 1.65 (3.55) 2.23 2.31 2.80 (1.58) 2.47 2.62 2.81 2.28 Marcasite with 3 per cent, solu¬ tion KMn0 4 cold. 2.72 2.87 2.17 2.33 2.87 (3.31) 2.88 (1.89) 2.83 2.77 Action of 5 Per Cent. Solution of Potassium Permanganate at Ordinary Temperature. In this series, as in the last, the action, as far as pyrite was concerned, was practically complete at the expiration of the first hour, but in the case of the marcasite this point was not reached until probably the end of the second hour, and, in fact, in one case was progressive to the end. But one very great discrepancy is to be noted here in the three-hour col¬ umn with marcasite. The low result in the next column is explained by caking. Table Showing the Relative Oxidation of Sulphur in Pyrite and Marcasite by a 5 Per Cent. Solution of KMn O i at 22°. Mineral. 1-Hour. • 2-Hour. 3-Hour. 4-Hour. 5-Hour. Pyrite with 5 per cent, solution KMn0 4 cold. 2.39 3.15 3.03 3.15 3.22 2.32 2.89 2.79 3.24 Marcasite with 5 per cent, solu¬ tion KMn0 4 cold. 2.10 2.52 3.06 3.76 3.82 (5.83) 2.77 3.16 (2.44) 3.39 4.17 This series finishes the experiments at ordinary temperatures. In all of them the action was comparatively slight, not exceeding at most 10 per cent, of the contained sulphur in the mineral which would not be sufficient to show any marked difference between the two minerals as bearing on their constitution, if the constitution which seems to be indicated by subse¬ quent experiments (to be presently described) is the true one. The oxidations with potassium permanganate at a temperature of 100° were conducted by suspending the vessel containing the mineral and so¬ lution in boiling water. Both stoppered bottles and thin glass flasks closed with perforated corks were used for this series of experiments. The water was kept continually boiling and the bottles or flasks were im¬ mersed deep enough to cover that portion of them containing the per¬ manganate. Six or eignt oxidations were made at one operation. The 9 permanganate solution, after it had acted the required time, was treated as in the experiments conducted at ordinary temperatures described above. Much more active oxidation took place at this temperature (100°), but the tendency of the mineral to cake together was much more marked, and now this took place with pyrite as well as with marcasite. Moreover, the deposition of manganese dioxide in every case was now very great, causing often a stoppage of the oxidation until it could be dislodged. As these oxidation experiments had already occupied much time only one trial was now made at each concentration of solution for each hour from one to five, unless, as before, marked discrepancies occurred, when two or more trials were made. The series of results are hence not so regular as they would have been had more trials been made, these irregularities arising from the difficulties that have been mentioned, as well as from the fact that the dilute solutions soon became exhausted, and that all solutions suffered some evaporation, but some more than others, causing irregular strength with the same solution. Nevertheless, the results agree in kind with those obtained at ordinary temperature, but differ widely in degree. Whereas at ordinary temperature the greatest amount of sulphur oxidized in marcasite by the five-hour trial with o per cent, permanganate was 4.17 per cent., at 100° this became 16.36 per cent, or about four times as much. Action of Normal Potassium Permanganate Solution at 100°. The results given in the following table show perhaps more strongly than either of the other series of experiments at 100° the effect of the different disturbing causes that have been mentioned. It especially shows the effect of caking of the pyrite, which now came in as an important dis¬ turbing factor. The result of this caking is shown in the three- and four- hour results with pyrite, both being very low. Marcasite, on the other hand, invariably caked and stuck to the bottom of the bottle, but as this was a constant source of error in this case, the results show a gradual and fairly even increase. Irregular results with marcasite were now largely conditioned by the evaporation of the solution or by the fact of whether the mineral was evenly caked over the inner surface of the vessel or con¬ centrated in spots. The result of this latter way of caking will be better seen in some of the subsequent series of experiments. Table Slcowing the Relative Oxidation of Sulphur in Pyrite aud Marcasite by « Normal Solution of KMnO i at 100° C. Mineral. l-IIOUR. 2-Hour. 3-Hour. 4-Hour. 5-Hour. Pvrite with N. KMn0 4 at *100° C. 4.05 4.72 3 36 2.04 5.64 Marcasite with N. KMn0 4 at 100° C. 3.17 3.84 3.76 5.63 5.61 10 Action of 1 Per Cent. Potassium Permanganate Solution at 100°. The results of the oxidation shown in this series are chiefly remarkable as still further illustrating the action of the solution on the caked material, as shown in the four- and five-hour trials with pvrite and the four-hour trial with marcasite. This latter shows, too, the effect of having* the caked mineral massed in one spot. With this exception, the marcasite oxidations are progressive and fairly uniform (the two hour trial falls somewhat below what it doubtless should be), but the pyrite shows a sharp fall in the four- and five-hour trials. The cause of this has been indicated. That the concentration of the solution by evaporation caused an increase in the action seems to be indicated by the result of the five- hour marcasite oxidation, but this is much more strongly marked in the 5 per cent, series in the case of pyrite, which will be referred to later on. Table Shouting the Relative Oxidation of Sulphur in Pyrite and Marcasite by a 1 Per Gent. Solution of KMnO 4 at 100°. Mineral. 1-Hour. 2-Hour. 3-Hour. 4-HOUR. 5-Hour. Pyrite with 1 per cent. KMn0 4 at 100° C. 6.03 6.98 8.38 6.11 6.88 Marcasite with 1 per cent. KMn0 4 at 100° C. 6.43 | - i 5.61 6.25 8.56 7.40 9.10 Action of 3 Per Cent. Solution of Potassium Permanganate at 100° C. The average results of this series of oxidations were very good, with the exception of two members of the series—the marcasite five-hour trial and the pyrite three-hour. This latter was repeated, but with a similar low result. Leaving these two out of account, the average results show a very fairly even rate of progression, which have been brought out in the diagram (PI. i). It is evident from an inspection of the following table that the marcasite oxidations of the four- and five-hour trials were arrested by some disturbing influence. Table Showing the Relative Oxidation of Sulphur in Pyrite and Marcasite by a 3 Per Cent. Solution of KMnO± at 100° G. Mineral. 1-Hour. 2-Hour. 3-Hour. 4-HOUR. 5-Hour. Pyrite with 3 per cent. KMn0 4 at 100° C. 5.52 (7.01) 5.77 7.89 6.81 5.16 10.73 11.08 Marcasite with 3 per cent. KMn0 4 at 100° C. 5.25 5.50 (8.67) 7.97 9.42 9.80 7 55 11 Action of a 5 Per Cent. Potassium Permanganate Solution at 100° C. • This series was decidedly the most satisfactory of the experiments con¬ ducted at 100° and, with the exception of the two-liour oxidations, in case of each mineral shows a remarkably regular increase in the oxidation of the sulphur. It illustrates, too, in the case of the pyrite, the effect of the concentration of the solution due to evaporation. This causes a more rapid rise in the results after three hours’ action, and notably between four and five hours. This rise is not so well seen from the table as from the plot of the results given in PI. i. Here this rising of the curve after three hours is very marked as contrasted with the curve of the marcasite oxidations. Table Showing the Relative Oxidation of Sulphur in Pyrite and Marrcasite by a 5 Per Cent. Solution of KMn 0 4 100°. Mineral. 1-Hour. 2-HOUR. 3-Hour. 4-Hour. | 5-Hour. Pyrite with 5 per cent. KMn0 4 at 100° C. 7.95 7.52 9.85 11.86 14.98 Marcasite with 5 per cent. KMn0 4 at 100° C. 8.38 8.38 13.27 14.85 16.36 The graphic representation of the rate of oxidation of sulphur in these two minerals is plotted from the following tables, the first of which shows the average amounts of oxidation of sulphur compiled from the several small tables already given. In this compilation I have omitted a few of the results given in parentheses in the small tables as they are evidently incorrect, and would vitiate the averages. On the other hand, I have sim¬ ply retained the parentheses of the small tables in certain cases where but one result was obtained. The second table, also compiled from the sev¬ eral small tables, shows for the experiments at ordinary temperature the result of selecting, wherever possible, such individual determinations as show a progressive increase in the oxidation. In this way more regular series of results are obtained. Inasmuch as the variations are so marked, this method of selecting results seems justified, at least for obtaining a com¬ parison of the relative rates of oxidation of the sulphur in the two min¬ erals. The results of Table i are plotted in PI. i, and those of Table ii in PI. ii. 12 I. Table Showing Average Relative Oxidation of Sulphur in Pyrite and Marcasite by Solutions of KMn.O 4 at 22° and at 100°. See PI. i. Mineral. 1-Hour 2-Hour 3-Hour 1 4-Hour 5-Hour On Pl. Pyrite T -J-„ N. cold. .78 1.17 1.38 1.74 1.72 220 Pi Marcasite N. cold. 1.07 1.86 2.04 (1.25) 2.38 220 M 1 Pyrite 1 per cent, cold. 1.71 (1.43) 1.86 1.85 1.80 22° P 2 Marcasite 1 percent, cold.... 1.22 1.21 2.06* 2.32* 2.24 22° m 2 Pyrite 3 per cent, cold. 2.55 2.27 2.80* 2.55 2.55 22° P 3 Marcasite 3 per cent, cold_ 2.80 2.25 2 87* 2.88* 2.80 220 m 3 Pyrite 5 percent, cold. 2.77 3.09 2.77 2.89 3.02 22° P 4 Marcasite 5 per cent, cold_ 2.31 3.41 3.30* 3 16 3.78 22° m 4 Pyrite N. 100°. 4.05 4.72 3.36 2.04 5.64 100° Pi Marcasite N. 100°. 3.17 3.84 3 76 5.63 5.61 100° Mi Pyrite 1 per cent. 100°. 6.03 6 98 8.38 6.11 6.88 100° P 2 Marcasite 1 per cent. 100°. 6.43 5.93 8.56 7.40 9.10 100° m 2 Pyrite .3 per cent. 100°. 6.26 6.83 6.81* 10.73 11.08 100° P 3 Marcasite 3 per cent. 100°- 6.47 7.97 9.42 9.80 (7.55) 100° m 3 Pyrite 5 per cent. 100°. 7.95 7.52 9.85 11.86 14.98 100° P 4 Marcasite 5 per cent. 100 c .... 8.38 8.38 13.27 ' 14.85 1 16.36 100° II. Table Showing Selected Results of the Oxidation of Sulphur in Pyrite and Marcasite by Solutions of KMnO 4 at 22°. See PI. ii. Mineral. 1-HOUR 2-Hour 3-Hour 4-Hour 5-Hour On Pl. Pyrite N. cold. .78 1.17 1.38 1.74 (1.72) 22° Pi Marcasite N. cold. 1.07 1.86 2.04 (1.25) 2.38 220 Mi Pyrite 1 per cent, cold. (1.71) 1.47 1.85 1.90 1 89 22° P 2 Marcasite 1 per cent cold. 1.16 1.29 1.93 1.95 2.15 220 Mo Pyrite 3 per cent, cold. (1.65) 2.31 2.80 (2.62) 2.81 22° P 3 Marcasite 3 percent, cold. (2.72) 2.33 2.87 2.88 2.83 22° m 3 Pyrite 5 per cent, cold. 2.39 3.03 3.22 (2.89) 3.24 22° ^4 Marcasite 5 percent, cold. 2.52 3.06 3.82 (3.16) 4.17 22° m 4 In plotting from these tables a vertical scale of one-lialf inch per one per cent, of sulphur oxidized and a horizontal scale of three-quarters inch per hour of oxidation have been employed. Wherever there was a drop in the results the gap has been bridged over by the main curve, thus mak¬ ing the curves as nearly as possible progressive, but the drop has been plotted in fine dotted lines. On examining these curves the points as to the relative rates of oxidation that have been mentioned are shown very graphically, the curve of the marcasite oxidation rising in each case above the corresponding one of pyrite, but both showing a similar form. Of these curves the normal solution cold and the 5 per cent, solution hot * Obtained by omitting discordant results. Plate II. ,il aisil •- QrravAe' cpr/T'/rvo