and cobalt, metals often present in iron in greater or less quantity, small quantities are thrown down along with the manganese precipitate, copper and nickel only as monoxides, but cobalt as sesquioxide; the latter, therefore, unless it be separately estimated, may bring on a slight error, so that the quantity of manganese would be found somewhat too large, viz., about half the quantity of the cobalt present.-Ber. d. Deutsch. Chem. Gesells. THE ESTIMATION OF SULPHUR IN PYRITES. Amount of Standard Sulphuric Acid taken. THE practical importance of providing a reliable method for estimating sulphur in pyrites has, I need hardly say, been fully realised by chemists who are called upon to make assays of this kind. The next method to which some attention was paid is the direct volumetric one of Wildenstein, which has lately been carefully studied by Messrs. Teschemacher and Denham Smith; I am able to confirm much that these chemists have said about it in several particulars. My own experiments satisfy me that it is desirable to conduct the titration in the presence of but little free HCI, in the entire absence of nitric acid, and to standardise the barium The experiments detailed in the present communica-chloride by iron sulphate, as nearly as possible under the tion, some of which have yielded but indifferent results, conditions which will prevail in a pyrites assay so far as were made, not so much with a view of arriving at an entirely new scheme for valuing sulphur ores, but rather to supplement our present ways and means of conducting the operations now in vogue. For most purposes, I think it may be assumed that sulphur in mineral substances is usually, if not invariably, determined as sulphate of barium, and that there are four methods in general use for this purpose-three direct, and one somewhat indirect. The direct comprise a gravimetric and two volumetric processes. The indirect is also a volumetric one, and is known as Mohr's alkalimetric method, useful in many cases, but scarcely applicable to the valuation of pyrites, since the conditions of an assay are not usually such as readily to adapt themselves to it. I tried the following alkalimetric process, which I believe has been already described by Bohlig (Fresenius's Zeitschrift) for estimating sulphuric acid, with the intention of applying it to the valuation of pyrites. It was soon evident, however, that the experimental conditions of an assay are not such as to adapt themselves readily to it. the amount of free acid and volume of fluid are concerned. Some little difficulty is experienced in deciding what shall be considered the end-point of the titration, owing to the barium chloride or sulphuric acid produces a slight so-called neutral point in the fluid when a drop of either turbidity. If hydrate of barium be added to a neutral solution of a sulphate of an alkali, and the excess be removed by CO2, the amount of caustic alkali liberated will be equivalent to the sulphuric acid previously combined in the absence of phosphates and fluorides. The solution containing the soluble sulphate in the presence of free HCl is precisely neutralised whilst boiling with sodium carbonate free from sulphuric acid. Hydrate of barium is then added in slight excess, the whole boiled for a few minutes, and the excess removed by CO2. After filtration and thorough washing of the residue on the filter by boiling-water, the filtrate is neutralised by standard acid. The following experiments were made; the volume of fluid was in each about 250 c.c.: C.C. 2 10 12 14 16 18 20 30 I may remark that the burette used to measure the standard acid at the commencement of the experiment was the one which measured the same acid when titrating the alkali, consequently the amount of acid used in the latter part of the process should have precisely equalled 3.83 3.67 that in the first stage had there not been a loss of alkali. Some of the experiments were repeated, and similar results were obtained; the error was always one of deficiency when the sulphuric acid was about half a gramme. The low results are doubtless due to the fact that boiling water fails to remove all alkali from the mixed precipitates, which is not at variance with our present knowledge of the properties of barium precipitates generally. I did not continue this enquiry further, inasmuch as it did not seem probable that closer numbers would be obtained unless the process was inconveniently prolonged. 3'75 3.82 3.81 3'66 I may just mention that up to a certain stage the testing for excess of barium can be done on a watch-glass. A drop of the partially clear solution is removed by means of a narrow tube pipette. A drop of the barium solution from the burette is brought near, and the two fluids allowed to run together. If the point of contact be carefully observed, the appearance of a faint opalescence will indicate sulphuric acid. The rapidity of the change and the depth of the opalescence serve to warn the operator of the approaching completion of his experiment. The absolute completion cannot be decided except on a filtered portion in a tube, and this can only be done under certain conditions, as the following observation testifies. Two measured equal volumes (4 c.c.) of the solution in a tively and at the same time one drop less than o'r of a c.c. pyrites assay were taken, and to each was added respecof the normal barium, and a like amount of normal sulphuric acid. In the tube to which the acid was added a slight turbidity appeared at the end of a minute and a half; in the other tube the fluid remained quite clear. To the latter four additional drops of barium were added, when a turbidity showed itself in a few seconds. I am not able to give an explanation of the phenomenon. Perhaps sulphate of barium is slightly soluble in a solution not containing too large an excess of BaCl2? Whatever may be the explanation, the reaction must be taken into account when verifying the strength of the barium chloride. In assaying pyrites for sulphur only by the fusion method, I have worked as follows and obtained good results. The process will no doubt be useful in laboratories which do not possess large platinum crucibles. A test-tube or piece of sealed combustion-tube, about six inches long and half an inch internal width, is fitted with a cork and delivery tube, the latter bent at a right angle and long enough to reach to the bottom of the flask in which it is intended to make the titration. The fusion mixture consists of equal parts of nitre and ignited acid carbonate of sodium, both free from sulphur, dry, and in fine powder. Nine to ten grammes is taken in an operation, together with one of pyrites, the latter must be in exceedingly fine powder; the two are mixed in a warm porcelain dish or agate mortar, and transferred to the tube without loss. The delivery tube is then inserted with its extremity dipping into the flask. A channel is made on the surface of the mixture, and the tube suitably supported is heated in small portions at a time with a Bunsen gas flame, commencing as usual with the anterior portion. When the operation is progressing favourably, the deflag ration proceeds for a few seconds after removing the flam There is no danger to be apprehended, and the tube does not crack or blow out with proper care. When the tube has been heated throughout, and the deflagration has ceased, it is then more strongly heated with a Herapath or powerful gas flame. It is a good plan at this stage to slip a coil of wire gauze over the tube, which helps to accumulate the heat. It is not, however, necessary that the contents should be fused a second time, at least this has not been done in experiments appended. The sulphur ores examined have yielded their sulphur readily. The gaseous products of the combustion which mechanically carry over with them small quantities of sulphates or sulphuric acid, being heavier than air, collect in the flask, and are washed by shaking with a little water, closing the flask with the palm of the hand. The delivery tube is also washed. That containing the fused mass is carefully broken and put in the flask, together with sufficient hydrochloric acid to dissolve nearly the whole of the iron oxide; then ammonia is added, until a precipitate of oxide reappears, and lastly as much free HCl and water as are necessary to bring the fluid to the conditions which obtained when the barium solution was standardised. I have used 2 c.c. of free acid, and the total volume of solution was 200 c.c. Experiments have been made on three samples of iron pyrites. The one containing the most sulphur has a bright crystalline fracture, and appears to contain but little siliceous matter. I have not made a complete analysis of it, as I wish to reserve a specimen. It was given to me by the landlord of a small inn at Macugnaga in the Val-Anzasca, near which place a mining company has been established to work the vein. The mineral is said to contain gold. I am uncertain of the source of the other two specimens. In the following experiments I have compared the results obtained by the method of working just described with those obtained by oxidising the sulphide with nitro-hydrochloric acid. In some laboratories both processes are used in valuing sulphur ores. The tables give the sulphur per cent on the dry sample. A single gravimetric determination was made on the Val-Anzasca specimen; the number obtained was 50'44. Nitro-hydrochloric acid was used to oxidise the sulphur, and the excess was removed by evaporation at 100° C. before adding the barium salt; the precipitation was made in the boiling solution, which was somewhat dilute. The chief cause of failure in conducting a fusion as above described is the possible incomplete oxidation of the sulphur. Such is rarely the case, provided the heat is sufficient and the mineral finely divided. To insure the latter condition it is desirable to sift the dry sample through muslin. I have not had an opportunity of extending the method to sulphur minerals generally, but there is little doubt that most, if not all, can be decomposed in this manner without loss of sulphur-that is to say, if ordinary precaution be taken. An anterior layer of fusion mixture may be dispensed with, and it is not necessary to rinse the dish; a camel-hair brush will remove any remaining particles both from it and the mouth of the tube. CHEMICAL NEWS DETECTION AND ESTIMATION OF PARAFFINE MAKERS of stearine candles mix paraffine with the fatty mass in quantities up to 20 per cent. Paraffine candle attribute valuable properties to such a mixture, so far as makers also mix stearic acid with their paraffine, and candle-making is concerned. The attempt to determine if paraffine be present, and if so, to get some approxivice versa, by means of the comparison of the meltingmate idea of the quantity, in a sample of stearine and point and specific gravity of such a mixture, is shown to be useless, as these vary according to the source from which the paraffine is obtained, as also in the case of the stearic acid, since the pure commercial article is by no means a chemically pure article. A good method for detecting the presence of stearic acid in paraffine has been devised by R. Wagner, viz., by treating a boiling solution of the paraffine in alcohol with an alcoholic solution of neutral acetate of lead, when, if stearic acid be present, a dense floccular precipitate appears, but none if it be absent. The best method, and one which can be used quantitatively as weli as qualitatively, is described as follows: Not less than 5 grms. of the candle are taken and treated with warm solution of hydrate of potash, which must not be too concentrated. A soap is formed with the stearic acid, whilst the paraffine is left unaltered. Salt is thrown into the solution, whereby the soap is separated out as a soda soap, and in precipitating takes down the paraffine with it. The soap obtained is thrown on the filter and washed with cold water or very dilute spirits of wine. Thus, firstly, the salt is washed out, and finally, the soap is brought into solution and likewise washed through the filter, leaving the paraffine, which is then dried at a temperature below 35° C., so as not to fuse and after repeated washing with this solvent, the ethereal it. The paraffine is then treated on the filter with ether, solution is carefully evaporated in a weighed porcelain crucible, in the water-bath, at a low temperature. The residue, consisting of the paraffine, is then weighed, and the stearic acid is estimated by difference. ON THE MEANS OF REGULATING GAS-FLAMES JEANNEL and Martensont have recently published descriptions of regulators of temperature, by means of which a constant temperature higher than that of the boiling. point of mercury may be kept up. It is unnecessary to enter here into details of the construction of these apparatus; suffice it to say that air is applied in them as the expanding medium. While engaged in a series of experiments on the process of dissociation of oxide of mercury, I required a high, but constant and only slightly varying, temperature for a considerable length of time, and for that purpose constructed a modification of Schlösing's apparatus, in which, in lieu of a mercury-reservoir, an air-reservoir is used, consisting of four glass tubes placed side by side and tied to each other, each tube 15 centimetres long by 2 centimetres diameter, This apparatus was placed into an airbath, through a slit cut in the top of it, at the side where the door is placed, care being taken that the bath is air *Polyt. Journ., No. 204, p. 460; Chem. Centralbl., 1872, p. 497. + Pharm. Zeitschr. f. Russland, 1872, No. 11, p. 136; Chem. Centralbl., 1872, p. 513. | tight, while it is heated by means of gas, which can be supplied at any desired pressure. When it was desirable to heat the bath to a high temperature, say 250°, the quantity of gas required for that purpose was found too large to be regulated by the apparatus, since the distance between the supply-pipe of the gas and the caoutchouc caps (probably those connecting the above-mentioned glass tubes) had to be made too great, this being due to the great loss of heat (by radiation) from the non-heated sides of the air-bath, which, in order to be heated to the boiling-point of mercury, requires a very large bulk of gas. It is through the kindness of Professor Gunning that I have been enabled to make these experiments, because, as the pressure usually kept up in street-gas-mains is not strong enough for this purpose, I was compelled to supply gas to the burners by means of a separate gas-holder. When the supply of gas reached a given pressure I was enabled to bring the temperature of the air-bath up to 350° by the use of four Bunsen burners, which temperature could be kept up constantly with very slight variations; with five such burners I could bring the temperature up to 362°. It is of course evident that neither Jeannèl's or Martenson's instruments, nor my own, can be used for regulating temperatures above that of the boiling point of mercury, an observation more particularly applicable to the apparatus of the first-named gentleman, in which the outlet opening is very small; perhaps by placing the instruments in a bath of molten metal (lead or zinc, for instance) regulation of the temperature might still be possible, but the instruments are not well suited for such use. If an air-bath were so constructed that the loss of heat from the metal by radiation were either entirely prevented or greatly reduced, it might be possible for my modification of Schlösing's apparatus, if of larger size, to be found to answer for regulation and constant maintenance of higher temperatures; as long as this is not effectually done we need not hope to be able to regulate high temperatures. I say this because the assertion to the contrary made by Martenson and Jeannèl, based simply upon the fact that air is the expanding medium in their apparatus, is not proved by facts. The instrument used by me enables me to regulate with great precision the temperature of either an air- or oil-bath, since the limit of variation of temperature is only about. The volume of air heated in my instrument is greater than in those alluded to, and my instrument is also more air-tight than theirs; I can therefore, upon experimental grounds, recommend the use of my modification of Schlösing's apparatus whenever a very constant and only slightly varying temperature is required.-Ber. d. Deutsch. Chem. Gesells. ON SOIL ANALYSES AND THEIR UTILITY.* By EUG. W. HILGARD, State Geologist of Mississippi. (Continued from p. 8). A MUCH graver defect is the failure to determine separately the organic matter (" humus ") and the chemically combimed water; and to this is owing, in a measure, the unsatisfactoriness of the analyses as regards information on the physical character of the soils. A large amount of water of hydration indicates in ordinary cases a correspondingly clayey soil, where heaviness in working may, or may not, be relieved by a large amount of "humus." The "volatile matter "" item, however, gives us no information whatsoever on these vitally important points; and there is, unfortunately, no simple method by which the determinations in question can be effected even approximately. That they should form part of every soil analysis is obvious, if only on account of the importance of "humus." I have attempted to obtain a reliable scale of the different degrees of "heaviness" of soils, from the de termination of their maximum absorption of hygroscopic moisture at ordinary temperatures. I find that at temperatures from about +7° to +21°, the amount of aqueous vapour absorbed by a thin layer of soil exposed to a saturated atmosphere remains very nearly constant, being for Read at the Dubuque Meeting of the Am. Assoc. Adv. Sci., August, 1872. ,, scopic power, as well as of “ there being of course, all intermediate grades of hygroheaviness." It appears that for this interval of temperature the decrease of absolute absorbing power in the soil, resulting from the rise of temperature, is just balanced by the increased amount of vapour diffused in the air-not an unimportant circumstance with regard to vegetable life. seriously with the correctness of the estimate as There are, however, two soil ingredients which interfere 46 heaviness," derived from the coefficient of absorption, viz., humus and ferric oxide. Both of these are highly hygroscopic, yet both counteract the "heaviness" caused by excess of clay. Moreover, there is a class of soils (viz., fine siliceous silts) whose exceeding "heaviness" in cultivation is much complained of, yet whose absorbent power is very small. When, as in the majority of cases, the surface soil has effect of the "humus" may be sensibly eliminated by been directly derived from the subsoil, the disturbing comparing, not the soils, but the subsoils, in this respect. sissippi soils analysed but three or four whose agricultural As to the ferric oxide, there are among about 200 Misqualities would have been seriously under-estimated by a reliance upon the coefficient of absorption alone. But I do not for a moment admit that in a material so complex, both in its composition and mode of action, any tural, may be relied upon to characterise the soil: or, as one or few data, whether chemical, physical, or agriculculture." So far from this, I consider that a proper inProfessor Johnson expresses it, "to do violence to agriterpetration of the analytical results must take into consideration, not only all the chemical and physical facts observed on the specimen, but all that has been relations to drainage, &c.; as well as all that is known or can be observed in loco-the location, depth, derivation, concerning the qualities or peculiarities of the soil, both in its natural state and in cultivation. As Professor Johnson says, it should "form part of a system of observations and trials; must be a step in some research; must stand, not as an index to a barren fact, but as the revelator of fruitful ideas." to Such, precisely, has been my object from the beginning sixteen years past. of my researches on the soils of the Mississippi, for Clearly, the difference between Professor Johnson's position and mine is one of degree only; yet this difference is not a slight one, since while, as before remarked, I have made, or caused to be made, some 200 analyses of soils and subsoils, his classic works on the growth and nutrition of plants do not contain so much as a tabular exemplification of the composition of various soils, as resulting from chemical analysis. If, then, "the probabilities of its uselessness in direct application to practice are so great," as Professor Johnson seems to hold, I have committed a grievous error, and squandered the substance of the State. I think that the considerations already adduced should plead measurably in extenuation of my course. But I will now state succinctly what services, in my view, soil analyses may fairly claim to be capable of performing, when conducted substantially in the manner, to the extent, and under the conditions defined above. I take it for granted that, if in the determination of the mineral ingredients we were able to distinguish clearly from one another the portion immediately available to In such cases, the surface soil is always more sandy than the subsoil. plants from that which is in an unavailable form, we would go far toward accomplishing what was originally claimed for soil analysis; and this Dr. Peter attempted to do by treatment of the soils with carbonated water. It cannot be coubted, however, that plants, as well as agriculturists, have at their disposal much more powerful, or at lest more energetic, solvents; and that, therefore, a determination of those ingredients which may fairly be considered practically within the reach of agriculture, must go deeper than does that with carbonated water. The proviso is important, but that with a proper local knowledge these allowances can be made, and that in most cases the information thus gained regarding the nature and treatment of the soil will be vastly more complete and reliable than the judgment of any number of "old intelligent farmers," my experience has fully convinced me; witness the egregious mistakes daily made by such in the selection of new lands. Moreover, a small minority only of farmers is likely to possess the requisite "age and intelligence;" and it is quite important that the multitude of those less fortunate should have the benefit of all the help science can give them. "" I will adduce but one" odious example " of a widely prevalent error in reference to the character of a class of soils that I have as yet been unable to eradicate, even from among the "old and intelligent;" who are unfortunately very much given to theorising on inadequate premises. Our prairie soils are notoriously limy; they are also very sticky;" and the mud takes the hair off the feet of cattle, Ergo, every "sticky" clay soil in the State is called, considered, and treated as a "prairie" soil, especially if the hardened clods adhering above the hoofs of cattle should carry the hair with them. If such soil is unthrifty, and rusts cotton, it is because" there is too much lime in it," which "scalds " the seedlings. In matter of fact, most of these soils are notably deficient in lime, so as to be most directly and immediately benefitted by its application wherever it has been tried, in accordance with my suggestion. The lime here acts probably as much chemically as physically; the clay being rich in potash, as per analysis. While the physical defects of these soils are doubtless the main cause of the crop failures, yet analysis has suggested a remedy which relieves, for the time being, from the necessity of the more costly improvements; lime being comparatively easy of access. Analogous cases are far from infrequent, both in this and in the adjoining States; and I have been led to attach special importance to the determination of lime in soils, from the (not unexpected) rule which seems to hold good very generally, viz., that, cæteris paribus, the thriftiness *See, for example, the article "Heavy Flatwood Soil," in my Miss. Rep., 1860, pp. 276, 279. of a soil is sensibly dependent upon the amount of lime it contains; while, at the same time, in the usual mode of culture without return to the soil, the duration of fertility is correspondingly diminished, and its cessation is very abrupt wherever much lime is present. It may be said that, after all, this is but what, from data already known, might have been expected. Granted; then, a fortiori, soil analysis, involving the determination of lime, is of considerable use in determining the present and future value of soils. In speaking of the "amount" of lime, I must be understood to refer, not so much to its absolute percentage, as to its quantity in comparison with that of potash, which, with phosphoric acid, is what all our fertilisers chiefly aim to supply. Their determination must, of course, be considered of prime importance, since their absence or extreme scarcity is fatal to profitable fertility; while, when they are present, even though immediately available for absorption to a slight extent only, we possess in lime, ammonia, &c., and the fallow, ready and powerful means for correcting their chemical condition. Here again, the practical value of soil analysis is direct and indisputable. It is of no small interest to know whether the soil we intend to cultivate contains o 75 per cent of potash and 0.25 of phosphoric acid, soluble in HCl, or only the fifth or tenth part of these amounts. One will bear improvement of all kinds-will pay for underdraining, terracing, &c. ; while the other, quite similar in aspect perhaps, would not, according to Liebig's testimony, ordinarily be capable of profitable culture. Again, it is well known that the same species of plants may occupy soils of widely different quality and value. True, an attentive observer will in such cases see differences in the mode of development;* yet these are often such as to escape ordinary remark, and grievous disappointments frequently arise from this source, with new settlers especially. It is of no small importance to be able to identify, as well as to distinguish, soils resembling each other; and this soil analysis can undoubtedly do, if there is any virtue in the law of probabilities even-admitting all that may otherwise be said against their reliability. Even if no other direct benefits than those already mentioned could be obtained by the chemical and mechanical analysis of soils (which I do not admit, and expect to prove otherwise hereafter) ; even if we leave out of consideration the addition to our general knowledge which may fairly be expected to result from extensive series of such investigations, carried out upon a uniform plan, whereby accidental errors (whether caused by "birds or squirrels," or analytical and other mistakes) will be eliminated; even thus, I contend that the practical and theoretical value of soil analyses is sufficiently great to justify whatever labour and expenditure may be bestowed upon them by state and national surveys; and that the neglect with which this branch of research has of late been customarily treated, is the more to be regretted as no probable amount of private effort can accomplish what must, of necessity, be done on an extended scale and with the prestige, voluntary assistance, and interest not usually accorded to any but public enterprises. And with due deference to the author of the two volumes whose extraordinary merits no one appreciates more than myself, I call upon my colleagues in State surveys, especially in the West and South, to re-consider this subject before it is too late, and a legislative fiat declares their work to be "finished." It is true that the agricultural colleges must and will take up and continue, as far as possible, the investigation of the agricultural peculiarities of each State; but the special and local experience acquired by those conducting a field survey, as well as their opportunities for extensive and comparative observations, are unfortunately "not transferable," even to the finest quarto report. In order to attain their highest degree of usefulness, our agricultural *Miss. Rep. 1860, p, 203. NEWS colleges should teach, not merely general principles, together with a sufficiency of the handicraft of agriculture; but they should be enabled to point out to each student, with reference to his particular neighbourhood, How Crops Grow, and How Crops Feed.-Am. Journ. Sci. THE CHANGES WHICH COAL UNDERGOES By H. ENGELMANN, E.M. THE subject of loss of carbon, or rather of deterioration, which stone coal suffers by exposure has of late attracted much attention amongst American mining engineers and metallurgists. The different coals are not equally affected by exposure. Their texture, their chemical composition, and the impurities which they contain, exercise consider able influence. Under otherwise equal conditions those stone coals suffer most which contain a large proportion of easily decomposed hydrocarbons, or which have little cohesive strength. Gas coals, after having been stored long, make less and poorer gas than when they are fresh from the mine, and coking coals lose their coking quality, more or less; some kinds are said to deteriorate very markedly within a few days after being mined. The nature of the changes which take place with the coal, and the conditions which influence them, still form a fruitful field for investigation. A large number of interesting experiments on this subject were made by Dr. Richter, Professor at the Mining School at Waldenburg, in Prussia, which deserve to be far more widely known than they appear to be in this country at least. A detailed account of them may be found in Dingler's Journal, 1870. I will confine myself to stating some of his principal results. As soon as the coal is mined it begins to absorb oxygen, rapidly at first, then more slowly. At first this action appears to be physical, but it soon becomes chemical, when the absorbed oxygen combines with the hydrogen of the coal to form water, and, with the carbon, to form carbonic acid. Heat intensifies the chemical action. Powdered stone coal fresh from the mine, heated to a temperature between 350 and 400 deg. F., increased in weight; although carbonic acid and aqueous vapour are disengaged, more weight of oxygen is absorbed. After a while a rather constant weight is obtained, and, by chemical analysis, the coal is then found to contain oxygen and hydrogen very nearly in the relative proportion in which they combine to water, which has not been the case in the fresh coal. The property of the coal thus rapidly to absorb oxygen depends mainly upon its proportion of free hydrogen. Of the carbon of the coal only a few per cent (5 or 6) combine thus rapidly with the oxygen at the stated temperature, while the rest of the carbon is far more stable. Different stone coals heated to the boiling-point of water until their weight remained constant, would absorb in a humid atmosphere at 60° F., from 2 to 7 per cent of water, and it was remarkable to observe that some solid pitch coal would absorb three times as much water as a soft laminated coal. The faculty of absorption could not be judged from the appearance of the coal, but coals from the same stratum exhibited considerable uniformity of behaviour. The coals which absorbed most water were also those which absorbed most oxygen. Twenty grm. coal absorbed in the first twenty-four hours after mining from 2 to 9 cubic centimetres oxygen. Stone coal absorbs carbonic acid even more eagerly than oxygen, taking up three times as much of it. At higher temperatures, the chemical action of the oxygen is increased, and a slow combustion takes place. The influence of humidity on the deterioration of coal is complicated, and Dr. Richter's experiments did not lead to very definite results in that respect. Air-dry coal absorbs the oxygen far more rapidly than moist coal, and coal which has been artificially dried absorbs it still more eagerly, taking up at the same time some nitrogen from the air. On the other hand, humidity induces decomposition of the iron sulphuret contained in most coals, which in turn accelerates the chemical changes of the coal by creating heat, by causing it to split and slack, and probably, also, by inducing chemical action between the oxide of iron formed and the coal, if not between the oxygen and coal directly. Light appears to exercise little influence. When coal has been exposed some time, and absorbs oxygen with little avidity, this absorption is a little greater in the dark. These were the principal results of Dr. Richter. An interesting experiment was made some years ago in Germany to test the deterioration by exposure of Silesian gas coal. A quantity of coal slack was divided in three parts. One part was directly used in the gas factory, another after having been housed one month, and the third after one month's exposure in the yard. The relative proportions of gas obtained were 135, 111, 95. The losses by exposure were, therefore, 17.2 per cent and 29'5 per cent. The gas coke from the first lot was serviceable; from the second and third unserviceable. It can hardly be doubted that this affinity of the oxygen for the coal has contributed much towards determining the quality of the coal which the different strata now present, but acting slowly in the course of ages, the effect has not been an apparent decomposition, but merely a difference of quality. The bituminous coking and gas producing coals have been least affected. They retain the largest proportion of hydrogen uncombined with oxygen and the least combined. The sinter or sand coals, which coke little or not at all, and furnish a poor gas, contain less hydrogen uncombined with oxygen, and more oxygen and hydrogen combined, than if they had been partly deteriorated by exposure. How little chemical equilibrium exists in a coal stratum is evident from the immense amount of carburetted hydrogen gas which is evolved in the coal mines, even in those which are not subject to dangerous accumulations of it in the form of fire-damp, and which in many, especially in deeper mines, can be heard escaping in minute bubbles from the sides of the rooms, making a peculiar noise. The quality of the roof and the quality and thickness of the superincumbent rock formations have exercised an important influence in determining the quality of the coal strata. Not seldom the deeper strata of a coal basin are the most bituminous ones. The more recent brown coals, the coals which are so extensively developed throughout the region of the Rocky Mountains, which are of cretaceous age, and present the appearance of stone coals, contain generally little hydrogen compared to their large proportion of oxygen. They should on that account not be apt to decompose readily, but the large amount of hygroscopic water which they contain, and their lack of cohesive strength, render many of them an easy prey to deterioration by exposure. I have seen many car-loads brought into this city (Salt Lake), of which the uppermost pieces, from an exposure of several days to the scorching sun and drying winds, were cracking and exfoliating very much like burnt lime in a moist atmosphere. These coals would certainly be of superior quality if they had been buried deeper in the bowels of the earth, and protected by heavy deposits of dense rock strata, which would have prevented the loss of so much of their bitumen, and rendered the access of oxygen more difficult. These coals are, however, not equally devoid of bitumen, and a locality in Sanpete Valley, U. T., presents a curious example of the influence of a solid casing. Near the village of Wales there is an outcrop of this brown coal, which is far more bituminous than the average. It is encased between solid beds of a slaty limestone, which forms a foot wall and roof of great stability. There is no clay seam or shale intervening. The whole thickness of the bed is about forty inches at this point, which appears to diminish to both sides, but it is not all coal. In its upper part it encloses several inches |