the part of an insulated conductor. The steam from a vessel of hot water was allowed to rise past a conductor, the apparatus being in front of a large fire, so that the air was very dry. When the conductor was charged the column of vapour was deflected from the vertical to the conductor both for a positive and negative charge. con Experiment 2 was made with the same object as Experiment 1. A gold leaf electrometer was charged so that the leaves stood open and then a cloud made to pass by the insulated leaves. As the cloud passed they were both attracted. This experiment was attended with siderable difficulty, as the moisture from the steam seemed to get on to the glass shade over the gold leaves and so form a charged conductor between the leaves and cloud. The cloud was first formed by a jet of steam from a a pipe, then by the vapour from a vessel of boiling water, and lastly, by a smoke ring or rather a steam ring. By this latter method an insulated cloud was formed, which, as it passed, was attracted by the charged leaf. Of the two latter propositions I have not been able to obtain any experimental proof. I made an attempt, but failed, through the bursting of the vessel in which the cloud was to be formed. I hope, however, shortly to be able to renew the attempt, and in the meantime I will take it for granted that these propositions are true. Faraday maintained that evaporation was not attended by electrical separation unless the vapour was driven against some solid, when the friction of the particles of water gave rise to electricity. So that unless there were some free electricity in the steam or vapour before it was condensed none could be produced by the condensation, and hence the cloud when formed would be uncharged. In the same way with regard to evaporation, unless, as is very improbable, the steam into which the water is turned retains the electricity which was previously In the condensed vapour; the electricity from that part of the cloud which evaporates must be left to increase the tension of the remainder. So that, as a charged cloud is diminished by evaporation, the tension of the charge will increase, although the charge remains the same. I will now point out what I think to be the bearing which these propositions have on the explanation of thunder storms. In doing this, I am met with a great difficulty, namely, ignorance of what actually goes on in a thunder storm. We seem to have no knowledge of any laws relating to these every-day phenomena; in fact we are where Franklin left us-we know that lightning is electricity and that is all. It is not, I think, decided whether the storm is incidental on the electrical disturbance or vice versa, i.e.. whether the electricity causes the clouds and storm or is a mere attendant on them. Nor can I ascertain that there is any certain information as to whether, when the discharge is between the earth and the clouds, the clouds are positive and the earth negative, or vice versa. Such information as I can get appears to point out the following law that in the case of a fresh-formed storm, the cloud is negative and the earth positive; whereas, in other cases, the cloud is positive and the earth negative. Again, thunder storms move without wind or independently of wind; but I am not aware whether any law connecting this motion with the time of day, &c., has ever been observed, though it seems natural that, however complicated by wind and other circumstance, some such law must exist. In this state of ignorance of what the phenomena of thunder really are, it is no good attempting to explain them. What I shall do, therefore, is to show how the inductive action of the Sun would necessarily cause certain clouds to be thunder clouds in a manner closely resembling, and for all we know identical with, actual thunder storms. In doing this I assume that the thunder is only an attendant on the storm and not the cause of it; and that many of the phenomena, such as forked and sheet lightning, are the result of different states of dampness of the air and different densities in the clouds, and really indicate nothing as to the cause of electricity. In the same way, the periodicity of the storms is referred to the periodical recurrence of certain states of dryness in the atmosphere. Thus the fact that there is no thunder in winter is assumed to be owing to the dampness of the air, which allows the electricity to pass from and to the clouds quietly. What I wish to do is to explain the cause of a cloud being at certain times in a different state of electric excitation to the earth and other clouds, and of this difference being sometimes on the positive side and sometimes on the negative, that is to say, why a cloud should sometimes appear to us on the earth to be positively charged, sometimes negatively, and at others not to be charged at all. The assumed condition of the sun and earth may be represented by two conductors S and E, acting on one another by induction, the sun being negative and the earth positive. The distance between these bodies is so great that the inductive action would not be confined to those parts which are opposed, but would in a greater or less degree extend all over their surfaces, though it would still be greater on that side of E which is opposite to S than on the other side. The conductor E must be surrounded by an imperfectly insulating medium to represent damp air. The formation of a cloud may then be represented by the introduction of a conductor C near to the surface of E. Such a conductor at first having no charge would attract the positive electricity in E and appear by reference to E to be negatively charged. If it was near enough to E, a spark would at once pass, which would represent a flash of forked lightning. If it were not near enough for this it would obtain a charge through the imperfect insulation of the medium. Such a charge might pass quietly or by the electric brush. When the cloud had obtained a a charge it would not exert any influence on the earth, unless it altered its position. But if the heat of the sun caused part of the cloud to evaporate the remainder would be surcharged and appear positive. Or if C approached E then C would be overcharged, and a part of its electricity would return, and on its return it might cause positive lightning. Thus, suppose that after a cloud had obtained its charge part of it came down suddenly in the form of rain. As the rain came lower its electric tension would increase until it got near enough the ground to relieve itself with a flash of lightning, almost immediately after which the first rain would reach the ground. It has often been noticed that something like this often takes place; it often begins to pour immediately after a flash of lightning, so much so that it seems that the electricity had been holding the rain up, and it was only after the discharge that it could fall. This, however, cannot be the case, for the rain often follows so quickly after the flash that there would not have been time for it to fall from the cloud unless it had started before the discharge took place. If, on the other hand, C receded from E, it would again be in a position to accept more electricity, or would again become negative. In this way, a cloud in forming, or when first formed, would appear negatively charged; soon after it would become neutral, and then if it moved to or from the earth it would appear positively or negatively charged. If the air was very dry, as it is in the summer, any exchange of electricity between the earth and the cloud would cause forked lightning, in the winter it would take place quietly, by the conduction of the moist atmosphere. In this way then there would sometimes be positive, sometimes negative lightning; sometimes the discharge would be a forked flash or spark, sometimes a brush or sheet lightning. And if clouds are formed in several layers, as would be represented by another conductor, D, outside C, then, in addition to the phenomena already mentioned, similar phenomena would take place between C and D; and if, in addition to this, we were to assume that there are other clouds in the neighbourhood, the he is acquiring the facts here given he will be laying the phenomena might be complicated to any extent. foundation of a truly scientific education. And if, further, the motion of the sun is taken into account; as the conductor S moves round E the charges in D and E would vary, accordingly as they were more or less between S and E and directly under the induction of S; i.c., the charge in a cloud would appear to change owing to the motion of the sun; thus, a cloud that appeared neutral at mid-day would, if it did not receive or give off any electricity, become charged positively in the evening. With regard to the independent motion of the clouds, there are several causes which would effect it. For instance, a cloud whether it appeared on the earth to be negatively or positively charged, would always tend to follow the sun, though it is possible this tendency might be very slight. Again, one cloud would attract or repel another, according as they were charged with the opposite or the same electricities; and in the same way a cloud would be attracted or repelled by a hill, according to the nature of their respectfve charges. Such, then, would be some of the more apparent phenomena under the assumed conditions. So far as I can see, they agree well with the general appearance of what actually takes place, but, as I have previously said, the laws relating to thunder storms are not sufficiently known to warrant me in doing more than suggesting this as a probable explanation. In these remarks I have said nothing whatever about what is called atmospheric electricity, or the apparent increase of positive tension as we proceed away from the surface of the earth. I do not think that this has much to do with thunder storms. If the law is established, it seems to me that it will require some explanation, besides merely that of the solar induction acting through the earth's atmosphere on to the surface of the earth. It would rather imply that the sun acts on some electricity in the higher regions of the earth's atmosphere, and that electricity in these regions acts again on the surface of the earth; but, however this may be, the effect of the assumptions described in this paper would be much the same. NOTICES OF BOOKS. Introduction to Inorganic Chemistry. By WILLIAM GEORGE THE student of chemistry must be sometimes at a loss to distinguish for himself the best text-book. Each of the numerous volumes from which he may select has its peculiar feature: and the existence of this plurity of teaching-methods is suggestive of the fecundity of the science, But many text-books unhappily differ in little more than the authors' names; in them there is given description of the orthodox method of preparing hydrogen from water, sulphuric acid, and zinc, as well as orthodox methods of obtaining other elements; some books vary the process by recommending expensive apparatus, others go to the opposite extreme of imagining all students trained mechanicians or artizans. Few of the text-books detail fully the method of preparing, say hydrogen, from the action of sodium upon water, the decomposition of water by electrolysis, by heated oxidisable bodies, &c. ; but among the number who are content to treat chemistry as an experimental science, Mr. Valentin ranks in the first order. His book records the neatest modes of preparing an experiment, and the bearing of the experiment upon the general theory of chemical science. As a practical handbook the student need go in search no farther, for while CORRESPONDENCE. CHEMISTRY OF ACID MANUFACTURE. To the Editor of the Chemical News. SIR,-In your report of the transactions of the Manchester Is it in your power to give the whole of the paper? Dec. 28, 1872. ANALYSIS OF MANURES. To the Editor of the Chemical News. SIR,-Having just seen Mr. Reynolds's proposition for a more correct scale of valuation for the constituents in artificial manures, I would like to draw his attention, as a primary question, to the means of ascertaining the relative proportions of the ingredients he names in manures. As yet it seems beyond the means of ordinary analyses. Mr. Reynolds admits a difference in the value of mineral and bone phosphates. If, then, a farmer buys a quantity of manure under the common name of " Dissolved bones" if he sends a sample of this to a chemist, and requests to know whether all the phosphates present are bone, and if not, how much of them are from minerals, can the chemist tell this? We suspect not. We have heard chemists of very high standing confess that when dissolved they cannot tell, and even the portion not in a soluble condition, if equal ground, the proportions of the one and the other cannot be given. This is what farmers wish to know, and then the true value is of easy determination; and so long as bones are nearly double the price of mineral, and no positive means of determining when and to what extent they are mixed, the farmer will be at the mercy of the adulterator. The relative quantity of ammonia as such in a manure and ammonia latent is within ordinary analysis, and ought to be determined and relatively valued-a matter long overlooked. I would suggest that Mr. Reynolds turn his attention to the determining the relative quantities of bone and mineral phosphates when they are mixed in a manure.—I am, &c., J. NAPIER. Glasgow, 25th Dec., 1872. MISCELLANEOUS. The Royal Polytechnic.-Never was a more attractive or more varied programme issued from this institution than that provided for the Christmas holidays. It includes a lecture by Professor E. V. Gardner, F.E.S., M.S.A., on the History of a Plum Pudding, an amusing entertainment by Mr. George Buckland, and a ghost illusion entitled "The White Lady of Avenel," the incidents of which are taken from Sir Walter Scott's tale of the "Monastery." It has been skilfully arranged by one of the Directors, Dr. Croft, who takes a lively interest in, and devotes much time to, the interests of the Polytechnic. Jan. 3, 1873. Chemical Technology, or Chemistry in its Applications to the Arts and Manufactures. BY THOMAS RICHARDSON and HENRY WATTS. Second Edition, illustrated with numerous Wood Engravings. Vol. I., Parts 1 and 2, price 36s., with more than 400 Illustrations. Nature and Properties of Fuel: Secondary Products obtained from Fuel: Production of Light: Secondary Products of the Gas Manu acture. Vol. I., Part 3, price 335., with more than 300 Illustrations. Vol. I., Part 4, price 21s., 300 Illustrations. London International Exhibition, 1873.-The second meeting of the Committee on Surgical Instruments and Appliances was held on the 23rd December, 1872, at 3 o'clock, in the offices, Gore Lodge, South Kensington. Among those present were Mr. Cæsar H. Hawkins, F.R.S., in the chair, Sir William Fergusson, Bart., F.R.S., Dr. P. Allen, Mr. W. Bowman, F.R.S., Mr. B. Brudenell Carter, Mr. W. White Cooper, Dr. W. T. Domville, R.N., Dr. Arthur Farre, F.R.S., Dr. G. T. Gream, Mr. Prescott G. Hewett, Mr. J. Hilton, F.R.S., Mr. J. Hinton, Professor J. Marshall, F.R.S., Mr. T. W. Nunn, Dr. W. S. Playfair, Mr. R. Quain, F.R.S., and Mr. Edwin Saunders. Letters received from the Royal College of Surgeons, and the Royal Medico-Chirurgical dices containing the latest information and specifications relating to Society were read, and it was stated that many of the leading surgical instrument makers in London, Dublin, Paris, and other capitals had signified their intention to contribute. It was suggested that the Exhibition should be brought before the notice of the British Medical Association at its meeting in August, 1873. The committee resolved to recommend the Royal Commissioners to invite corresponding members in foreign countries, and after the transaction of general business adjourned till Monday, the 20th January, 1873. * Vol. XXVI. of the CHEMICAL NEWS, containing a copious index, A Student. The CHEMICAL NEWS is the only one. X. Y. Z.-Dr. Morfit's work. It has been reviewed, and also advertised in our pages. In our report of the discussion of the Chemical Society, p. 306, for Mr. Prosjean read Grosjean. Vol. I., Part 5, price 36s. Prussiate of Potash: Oxalic, Tartaric, and Citric Acids, and Appenthe materials described in Parts 3 and 4. BAILLIERE AND Co., 20, King William Street, Strand. ERNERS COLLEGE of CHEMISTRY.— BER EXPERIMENTAL MILITARY and NAVAL SCIENCES, under the direction of Professor E. V. GARDNER, F.E.S., &c., of the late Royal Polytechnic Institution and the Royal Naval College. The Laboratory and Class Rooms are open from 11 to 5 a.m., and and from 7 to 10 p.m. daily. Especial facilities for persons preparing for Government and other examinations. Private Pupils will find every convenience. Analyses, Assays, and Practical Investigations connected with Patents, &c., conducted. For prospectus, &c., apply to Prof. E. V. G., 44, Berners-street, W Royal now open. Polytechnic Institution, 309, Regent Street.-Laboratory (entirely re-fitted) and Class-Rooms are ASSAYS, ANALYSES and Investigations connected with PATENTS conducted. Pupils received for Class and Private Study. Special facilities are offered to persons preparing for GOVERNMENT EXAMINATIONS. Classes are now forming for Practical Study in CHEMISTRY, STEAM, and PHYSICS. For particulars, apply to Professor E. V. GARDNER, F.A.S., M.S.A., at the Institution. North London School of Chemistry, Phar macy, &c.-Conducted by Mr. J. C. BRAITHWAITE, for thirteen years Principal Instructor in the Laboratories of the Pharmaceutical Society of Great Britain, and Demonstrator of Practical Pharmacy, Pharmaceutical Latin, &c. The Session 1872-1873 will commence on the 1st of October, when The LABORATORY will be open at ro a.m. for Instruction in Practical Chemistry as applied to Pharmacy, Medicine, Analysis, &c. Terms moderate. The CLASSES will meet as usual. The CHEMICAL and TOXICOLOGICAL CLASS on Monday and Thursday evenings at 8 p.m., commencing October 1st. The LATIN CLASS on Tuesdays and Fridays at 8 p.m., commencing October 2nd. The MATERIA MEDICA and BOTANICAL CLASS, every Wednesday and Saturday at 8 p.m., commencing October 3rd. The BOTANICAL GARDEN affords to Students desirous of Fee to either of the above Classes Half-a-Guinea per Month Letters of inquiry should be accompanied with a stamped envelope. Silicates of Soda and Potash in the state of Soluble glass, or in CONCENTRATED SOLUTION of first quality, suited for the manufacture of Soap and other purposes, supplied on best terms by W. GOSSAGE and Sons, Widnes Soapery, Warrington. London Agents, CLARKE and COSTE, 19 and 20, Water Lane, Tower Street, E.C., who hold stock ready for delivery. Source of Error in the Valuation of Pyrites. CHEMICAL 11 JAN 73 VOL. XXVII. No. 685. MUSEUM NEWS. ON SOURCE OF ERROR IN THE VALUATION OF PYRITES.* 13 a hard wedgwood mortar, and after thorough mixing a quantity was reduced to an impalpable powder in the same mortar. Two determinations of the sulphur and sand in the portion gave the following results :-Sulphur, 45'08 per cent; sand, 8.92 per cent; and sulphur, 45'04 per cent; sand, 8.88 per cent. 2. Another portion of the sifted sample was similarly reduced in a wedgwood mortar, which had been in use for some considerable time previously (this mortar was perceptibly softer than that used in the preceding experi ment). Two determinations of sulphur and sand gave By NICHOLAS GLENDINNING and ALFRED J. M. EDGER. Sulphur, 4356 per cent; sand, 1152 per cent; and THE great consumption of pyrites in industrial chemistry renders the accurate determination of the sulphur it contains a matter of great commercial importance. Judging however, from the serious discrepancies which so frequently occur in the results obtained by different chemists when operating upon the same sample, it would appear that the estimation of sulphur is either surrounded by great difficulty or performed under very defective manipulation. sulphur, 43'44 per cent; sand, 11.50 per cent. 3. Another portion of the same sifted sample, reduced in an agate mortar, gave in two determinations:-Sulphur, 45 28 per cent; sand, 6-2 per cent; and sulphur, 46.15 per cent; sand, 6'24 per cent. 4. A duplicate sample of the same ore, instead of being ground in a wedgwood was reduced in a steel mortar, and a portion brought to an impalpable powder in the agate, two determinations giving :-Sulphur, 46.25 per cent; sand, 6.11 per cent; and sulphur, 46.36 per The following is an arrangecent; sand, 6:13 per cent. ment of the results in a tabular form :Average Average Sulphur per cent. Sand per cent. Sulphur per cent. Sand per cent. 1. Wedgwood mortar alone 2. Wedgwood mortar alone It will be quite within the recollection of the members of this society that their attention was drawn by a late president (Mr. Glover) to the differences observable in the results obtained by different experimenters. This gentleman explained that in seven samples, each tested by two professional chemists, there was a maximum difference of 19, a minimum of 14, and an average of 1.6 per cent of sulphur, but differences still greater have come under our own observation. This subject was afterwards noticed by Messrs. Teschemacher and Smith in their article "On the Estimation of Sulphur by Barium," published in the CHEMICAL NEWS, vol. xxiv., 3. p. 61. They were of opinion that the observed differences might be traceable to different methods of testing, and adduced a number of results by way of proof that 4. the method then in general use was not reliable. Shortly after the appearance of their article we published results in support of that method (see CHEMICAL NEWS, vol. xxiv., p. 140), and we now beg to draw your attention to a source of error which precedes the analytical process, and precludes the possibility of correct results being obtained,. We allude to the use of the wedgwood mortar in the preparation of the portion for analysis, and are persuaded that its use obtains very generally in both professional and private laboratories. a In the midst of all the contention respecting the differences between results, it does not seem to have occurred to the experimenters that the cause of difference might be introduced in the preparation of the sample; and yet it must have been observed that, to whatever use wedgwood mortar has been applied, a very perceptible wearing is effected in course of time. A few years ago we devoted some attention to the differences which were causing so much annoyance; and from the circumstance of mortars becoming somewhat worn, we inferred that this might be the solution of the problem, and subsequent experiments proved the accuracy of this conclusion. We prepared a sample by passing it through a sieve of 529 meshes to the inch to ensure uniformity of composition, and then reduced portions to an impalpable powder for analysis; and then estimated in the several portions the sulphur and sand. The sulphur results were found to differ to an extent in some cases of upwards of 2 per cent, and this was accounted for by the amount of sand found in their respective portions. We at once saw that the wedgwood mortar was quite unsuited to the purpose, and have since prepared samples in a different way. We will now proceed to lay before you some results obtained quite recently from samples prepared in various ways. 1. A sample of pyrites was passed through a sieve of the above description, the coarser particles being reduced in *Read before the Newcastle-on-Tyne Chemical Society, Dec. 19, 1872. 43'56 It will be observed that the difference between the results of Nos. 4 and 2 is, sulphur 2.80, and sand 5'39 per cent; and between Nos. 4 and I, sulphur 1'24, and sand 4 and there is only 278 per cent; whilst between Nos. the slight difference of sulphur o'09, and sand o'10 per We may mention that A represents the results of the metropolitan chemist, and B those of the local chemist. Our sulphur results were obtained by the method described in our article in the CHEMICAL NEWS before alluded to, which we can, with great confidence, recommend as capable of giving concordant and reliable results. We may remark that the sulphur result of No. 2 is not sufficiently high by 0.3 per cent if the difference between its sand and that of No. 4 be taken as the full extent of We have, however, so repeatedly the contamination. found this to occur that we assume that some portion of the detached mortar is soluble in the acids employedprobably not the wedgwood itself, but matter which the mortar may have absorbed during its use, It is obvious from these results that by using the wedgwood mortar very serious differences may arise, their extent varying of course with circumstances, e.g., the hardness of the sample, the hardness of the mortar and its condition, and the industry of the operator. In conclusion, we hope we are justified in assuming that the results we have had the pleasure of laying before you afford sufficient proof that the use of the wedgwood mortar involves a more or less serious depreciation of the sample, and that under these circumstances the_proba- | remains in solution, and this quantity does not vitiate the bility of the attainment of a result representing the true results. In order to obviate the washing of the precipitate, value of the sample as it reached the assayer's laboratory I dilute the previously-cooled fluid to 500 c.c. I next is of a most remote character, and an occurrence which filter it through a dry filter, and take 250 c.c. of the filtrate, can only be due to an error in analysis. equal to 0.55 grm. of the quantity of the material originally taken, and in these 250 c.c. I estimate the manganese. When the manganese is intended to be precipitated by a further addition of acetate and bromine, and the quantity of dioxide of manganese then acidimetrically determined, we again meet with the difficulties occasioned by the ready decomposition of the acetate of manganese, and the capability of the dioxide of entering into combination with the lower oxides of manganese, and consequently the protoxide. These difficulties are best met by the following method of operation :-10 grms. of sodic acetate are dissolved in 150 c.c. of water, 50 c.c. of bromine-water are added, and, at intervals of half an hour, 50 c.c. of the manganese solution, care being taken that at the third half-hour 50 c.c. of bromine-water be again added to the fluid, to which heat should not be applied. NOTE. Since reading the above paper, we have made experiments substituting the porcelain in lieu of the weedgwood mortar in the preparation of the test sample, and the results obtained show that it also contaminates the sample to a serious extent, as will be seen from the following results :-A sample of the same pyrites as was used in the preceding experiments (and which gave by the steel and agate mortars in two determinations-Sulphur, 46 25 per cent; sand, 6'11 per cent; and sulphur, 46 36 per cent; sand, 613 per cent) gave when the porcelain mortar was used-Sulphur, 45'3 per cent : sand, 8.21 per cent; and sulphur, 45.36 per cent; sand, 8:23 per cent. Or, comparing their averages, the results of the porcelain mortar show an excess of 21 per cent of siliceous matter, and a deficiency of o'97 per cent of sulphur. | As, in these days, iron which contains more or less manganese is considered to be more or less valuable, it occurred to me to try to devise a method whereby the estimation of manganese in the above-named materials might be rendered more expeditious and less liable to the errors ordinarily inherent in that estimation. At a future period I intend to publish an exhaustive account of my researches; I therefore now communicate only what is of direct practical utility. When a hydrochloric acid solution of perchloride of iron is neutralised by means of carbonate of soda, so as to cause a permanent precipitate, and the latter is cautiously re-dissolved by the addition of some hydrochloric acid, a liquid is obtained which contains fourteen times its equivalent of ferric hydrate in solution, yet it is not precipitated by boiling. In order, therefore, to precipitate the iron by means of acetate of soda, aided by boiling heat, there is, theoretically, only as much acetate required as is sufficient to convert the chloride into acetate, viz., 3 molecules of acetate to 15 atoms of iron, or I part, by weight, of crystallised acetate of soda to 2 parts, by weight, of iron. It is, indeed, an easy matter to precipitate completely, by boiling, from previously carefully neutralised iron solutions, I'I grm. of iron upon 500 c.c. of fluid by means of 1 grm. of acetate of soda, even when, in order to prevent the possible decomposition of other acetates, i grm. of acetic acid had been added. Under these conditions, only very small portions of manganese are thrown down along with the iron-for instance, from 13 per cent only about o'02 to o'05 per cent, and less when the quantity of manganese is smaller. From what has just been stated, it is quite evident that the loss often experienced in the estimation of manganese by the method just alluded to is mainly due to the fact that too large a quantity of acetate of soda is employed, in consequence of which a portion of the chloride of manganese is also converted into acetate, which salt is readily decomposed into protoxide and acid, a fact almost entirely overlooked. It is probable that my improved method may also be applied to the separation of iron from zinc, copper, nickel, and cobalt, Before I found out this method I used sulphate of soda, for decomposing the previously neutralised chloride of iron. Although this salt may even be used in excess without any manganese being precipitated, I found that I grm. of it is sufficient to separate 11 grm. of iron; only a very small quantity of it In this manner, the manganese dioxide is thrown down from so dilute a solution that only from 002 to 0.03 per cent of manganese out of 13 per cent are lost in the shape of protoxide. On the other hand, small, but, with the whole quantity, proportional, quantities of manganese remain, either in solution as permanganate, or are so fixed to the sides of the glass vessel in which the operation takes place, that it is necessary to re-dissolve it. After the free bromine has been driven off by heat, the precipitate is filtered off, washed with dilute solution of acetate of soda, and then treated, together with the filter, with from 5 to 15 c.c. of a solution of chloride of antimony (1 to 5) and 15 c.c. of concentrated hydrochloric acid, after which the fluid is diluted with 100 c.c. and titrated with decimal permanganate solution, I c.c. of which then equals o'5 per cent of manganese. When the quantity of manganese is less (below I per cent), all quantities are tripled, and 5-6ths of the filtrate from the iron precipitate, previously concentrated by evaporation, are treated for manganese, I c.c. of permanganate becoming in this case equal to o'1 per cent of Mn. The titre of the solution of permanganate is found, either by comparing it with a solution of potassic bichromate of known strength by means of chloride of antimony, or by investigating a known quantity of pure protoxide of manganese according to the method described above. By mixing pure solutions of manganese (I often used for this purpose solutions of permanganate of known strength, with solutions of iron, the manganese in which had been previously determined, in variable proportions), I made mixtures for testing, in which the quantity of manganese varied from o'1 to 13 per cent. As instances of at least four experiments, I quote the following: - Per cent Mn. |