210 Singular Anomaly of the Sesquioxide of Iron. resist complete destruction by chromic acid, and we have endeavoured to find a method which will effect the complete oxidation and separation of these bodies so that nothing but pure anthrachinon remains. "The method is as follows: "The first part of the process, i.e., the solution of the sample in glacial acetic acid and gradual oxidation by chromic acid, as described in our former circular, remains the same. The anthrachinon thus obtained (washed with water only) is washed from the filter into a dish with 1 or 2 c.c. of caustic lye and boiled for five minutes with 1 c.c. of chameleon solution (permanganate of potash), the whole being well stirred. Should the red colour of the chameleon solution disappear, more' is to be added until an excess is visible. After cooling, the liquid is rendered slightly acid by sulphuric acid, and a few crystals of oxalic acid added, which will effect the solution of the separated manganese compounds. "The anthrachinon is then collected on a filter and washed, first with water, then with dilute alkali, again with water, and finally dried at 100° C. "The treatment of anthrachinon with permanganate solution occupies a short time only, and ensures the complete removal of all coal-tar impurities, anthrachinon alone remaining; we therefore consider the above treatment an essential perfection of our anthrachinon test. "To avoid the errors involved in collecting on weighed filters, or filters counterpoised in the ordinary way, the following plan is recommended:— "Two filters of equal size are selected and folded; the point, or cone, of one is cut off, so that, when folded together, the point of the inner projects through the cut part of the outer one to the extent of about half an inch; the two filters are then counterpoised exactly in a delicate balance, by cutting off small portions with scissors, and then folded together. The solution to be filtered is passed through both filters; after washing and drying, the inner filter, containing the anthrachinon, is placed in one pan of the balance, and the outer one in the other, adding weights till the counterpoise is perfect. "The advantages of this method are, that it obviates errors (1) from solution of the substance of the paper; (2) from scraping off some of the filter with the knife; and (3) from unequal absorption of moisture while weighing." In adding this appendix to our paper, we may state that we have not made a trial of the improvements; we hope to do so shortly upon the dark green hydrocarbon, which we have found it so difficult to oxidise with chromic acid." We give the appendix so that the readers of the CHEMICAL NEWS may be acquainted with the latest on anthracen analysis. SINGULAR ANOMALY OF THE SESQUIOXIDE By J. LAWRENCE SMITH, Louisville, Ky. IN my studies connected with the physical, chemical, and mineral characteristics of meteorites, several points of more or less importance have been observed which have as yet formed no part of my published observations on this class of bodies. One of these observations is, that the sesquioxide of iron, obtained in the result of the analysis of meteoric irons, is always decidedly affected by the magnet, although the temperature to which it may be heated prior to final weighing be but a moderate redheat, only sufficient to furnish a certain constant weight. This phenomenon was at first attributed to small particles of organic matter from the filter, that may have caused the reduction of a minute portion of the sesquioxide to a lower oxidation. Still later I was led to suspect the presence of some new magnetic metal-a metal, however, closely allied to iron (other than cobalt and nickel, as the oxides of neither of these metals are attracted by the {CHEMICAL NEWS May 1875. magnet, whether obtained from meteoric iron or from any other source). Experiments were undertaken to discover such metal, if any existed, but without success, and there the matter rested with me for several years, and until I commenced the study of the Ovifak iron, on which I am now engaged, in developing such facts connected with it as convince me that it is of terrestrial and not celestial origin, the results of which investigation I hope, before long, to submit to the Academy. The observation in connection with the sesquioxide of iron resulting from meteoric irons was applied to that from the Ovifak iron, and it was also found to be attracted by the magnet. This anomaly showing itself in two irons that I supposed to be of different origins, it was resolved to make a critical investigation of the sesquioxide of iron cbtained from both celestial and terrestrial sources; and if what follows appears too much in detail for the examination of so simple an observation, it was, nevertheless, necessary, in order to eliminate, as far as possible, sources of error, and to arrive at correct results. Materials, &c., used in the Experiments. The irons experimented with were dissolved in a large excess of a mixture of equal parts of chlorhydric and nitric acids. The filters used were old cotton cloth (perfectly free from lint), stretched over a funnel; the quantity of iron used did not exceed 1 grm.; the filtration was rapid, and the washing very easy and thorough. Other filtrations were made in funnels, the necks of which were plugged with asbestos. The heating was conducted in thin, glazed porcelain crucibles, of about 40 c.c. capacity, and the heat applied by a small Bunsen burner of 5 cubic feet capacity per hour, the top of the flame spreading over the bottom and one-half of the sides of the crucible, thus heating to redness 2 or 3 decigrammes very readily, the amount usually employed. As for the magnet used, I would state that as sesquioxide of iron, in a strict sense, ranks among the magnetic compounds, yet this is only true when very powerful magnets are brought to bear upon it; so the magnet used in all the experiments was a feeble one,-a small horseshoe magnet, capable of sustaining about 200 grms. when the two poles were connected, and, when used, one of the angles of one of the poles was brought near to the particles of oxide experimented with. First Series of Experiments made with Pure Sesquioxide of Iron. The first sesquioxide used was prepared from the purest commercial iron, dissolved in equal parts of chlorhydric and nitric acid, diluted, filtered, and precipitated by an excess of ammonia after filtration. 1. Dried at 110° C., crushed to fine particles, but not pulverised. The magnet, being brought within 1 m.m. of these particles, would attract them feebly. 2. The same heated to 250 C. still showed signs of magnetism. 3. Heated to 425° C. the magnetism entirely disappeared, or could only be made apparent upon the very smallest particles. 4. At red-heat the magnetism had entirely disappeared, whether the heat was continued for five or ten minutes only, or during several hours. All the tests with the magnet were made when the heated oxide had cooled; and the same was done in all the experiments referred to in this paper. The second sesquioxide of iron used was prepared from pure protosulphate oxidised by nitric acid, and precipitated by ammonia. This, when dried, heated, and tested in the same manner as in the former experiments, gave similar results, viz., that hydrated sesquioxide of iron was always slightly magnetic when dried at a low temperature (even at 50°); but at high temperatures, as red-heat, all magnetic property disappeared. NEWS The third sesquioxide of iron was prepared from hemical pure iron, of which I had a small quantity resented to me by Johnson and Matthey, of London, hat had been made with great care by Mathiessen. The esults of these experiments differed in no way from those ust detailed. Second Series of Experiments, with Sesquioxide from I next took meteoric iron with which to prepare the sesquioxide, using irons from a variety of sources. Those experimented with were the Toluca, Cranbourne, Russel Gulch, Sevier Co., Robertson Co. irons, and the iron from the stony meteorite that fell at Parnallee. These irons were dissolved in the same way as the iron in the first series, viz., with a large excess of chlorhydric and nitric acids, in equal parts, and most of the excess of acid driven off over a water-bath. The solution was then diluted, filtered, and poured into a bea ker of hot distilled water, to which had been previously added a large excess of ammonia. The precipitates were washed and dried precisely as in the first series, and oxides examined by the magnet. The following were the indications given by all the oxides, with but very slight difference in degree :Dried at 110° C., sensibly magnetic when the magnet was brought nearly in contact with the fine particles. Heated to 190 C., magnetism exhibited was about the same; at 300 C., slightly increased; at 450°, decidedly magnetic and at red-heat more strongly magnetic-particles, 2 or 3 m.m. in diameter, being attracted at some little distance. It made no essential difference if the red-heat was continued only for a few minutes or several hours. It will be remarked, in this series of experiments, that the effect of heat was just the reverse of what occurred in the first series; for in those experiments what little magnetic property existed in the oxide as first prepared, disappeared entirely when the oxide was brought to a red-heat. I would here remark that the sesquioxide, as usually prepared by me in the analysis of meteoric iron, differs somewhat from the above, the iron in solution being first precipitated by boiling with acetate of soda, and the subacetate of iron subsequently converted into oxide; but the oxide thus prepared still exhibits magnetic properties analogous to that formed by using an excess of ammonia. From the above results, the question naturally arises as to what is the cause of the difference in the first and second series of experiments; for the conditions under which the oxides were formed are as nearly the same as it was possible to make them, and they were made side by side at the same time. In the first part of this communication it was stated that I supposed it might be due to a new metal closely related to iron, but whose oxides were all attracted by the magnet: this idea was discarded after varied and careful experiments that failed to give any evidence of the correctness of such an assumption. The only thing left for me to do was to look into the oxides of iron from meteoric iron, and see what impurities they contained, and ascertain if these impurities played any part in this magnetic anomaly. It is well known that when a solution of meteoric iron is precipitated by an excess of ammonia, the oxides of nickel and cobalt at first precipitated along with the sesquioxide will be redissolved, but only in part, leaving often from 1 to 3 per cent of these oxides mixed with the oxide of iron. It is further known that while nickel and cobalt can be very nearly separated from iron by means of the acetate of soda, still a sensible portion will be left in the precipitated oxide of iron; so much so, that when I aim at exceeding accuracy in this separation, the subacetate at first precipitated (after par ial washing) is re-dissolved and reprecipitate by the acetate of soda, and this operation repeated a second and a third time. While there is nothing in the presence of the oxides of nickel and cobalt that would lead to any explanation of the phenomenon under discussion (for neither of these oxides is affected by the magnet), still I was determined to see what would be the condition of the oxide of iron prepared from meteoric iron after four re-solutions and re-precipitations by acetate of soda, finally converting the subacetate of iron into the sesquioxide by dissolving it in aqua regia, and precipitating by ammonia. The sesquioxide thus prepared no longer possessed the properties of that indicated in the second series of expe riments, but showed the same properties as the ordinary oxide of the first series, viz., not attracted by magnet after heating to redness. It being evident that the minute quantity of the oxides of nickel and cobalt remaining in the sesquioxide, as prepared from meteoric iron, had something to do with the phenomenon under consideration, I was led to institute a third series of experiments, using the sesquioxide from ordinary iron. Third Series of Experiments with Sesquioxide of Iron prepared from Pure Iron, mixed with Oxides of Nickel, Cobalt, and other Metals. A quantity of a solution of the sesquioxide was prepared by dissolving pure iron in a mixture of equal parts of chlorhydric and nitric acids, and a portion of the solution equal to about a gramme of the metal used in each of the subsequent experiments, adding to the solution of iron a solution of the other metals in the same acids prior to precipitation by ammonia. Some of the iron solution, without any admixture, was precipitated, and the sesquioxide examined as in Series 1st, and being satisfied that the pure oxide gave no indication of magnetic attraction, the experiments were proceeded with. Exp. 1.-A solution of the mixed oxides of nickel and cobalt, as obtained from meteoric iron, was added to some of the above solution of iron (the nickel and cobalt oxides representing about 10 per cent of the iron) precipitated by large excess of ammonia, dried, and tested with magnet before and after heating to redness, and the magnetic results were just the same as if we had been experimenting with meteoric iron, detailed in Series 2nd. This experiment was frequently repeated, and always with the same result. Exp. 2.-Same experiment as last, only using oxides of nickel and cobalt from ordinary sources; the results were the same, only not quite so marked. Exp. 3.-Pure nickel oxide was used, with similar results. Exp.4.-Pure oxide of cobalt was used with similar results. While the above experiments throw some light upon the subject, they in no way explain the phenomenon, for none of the oxides separately exhibit the magnetic property. I therefore continued the experiments, using other metals besides nickel and cobalt. Exp. 5.-Ten per cent of copper, dissolved in nitric acid, was added to a portion of the solution representing a gramme of iron, and precipitated with an excess of ammonia. The oxide dried and heated as in the previous experiments. When tested with the magnet it gave indi. cations similar to the oxide from meteoric iron, but less in degree. In subsequent analyses of this oxide, thus precipitated, it was found to contain nearly 3 per cent of oxide of copper. Exps. 6, 7, 8, 9, and 10.-In these experiments oxides of manganese, gold, platinum, zinc, and cadmiuni, were employed as admixtures with the iron; but in all cases the precipitated oxides differed in no way from the pure oxide of iron; when tested by the magnet evincing no attraction after being heated to redness. The above experiments embrace all that I have made up to the present time, and the results that are new may be summed up as follows: 1. That artificial hydrated sesquioxide of iron, dried at low temperatures, is attracted feebly by the magnet, but loses this property at and below red-heat. 212 Notes upon Sugar Analysis. 2. That sesquioxide, precipitated in the ordinary methods from solutions of meteoric iron, and dried at a low temperature, acts similarly to the ordinary oxide, but differs from it in that it becomes decidedly magnetic on being heated from 400° C. to a red-heat. 3. That the sesquioxide from ordinary iron, mixed with nickel or cobalt, or both, from whatever source, exhibit magnetic properties identical with that from meteoric iron. 4. That the sesquioxide of iron from meteoric iron, 5. That the sesquioxide made from iron mixed with reaction. A natural inquiry is-What is the cause of this change in the sesquioxide of iron when mixed with oxides of nickel, cobalt, or copper? Analyses of mixed oxides WCHEMICAL NEWS, NOTES UPON SUGAR ANALYSIS. DURING the last few years I have devoted much attention to the subject of sugar analysis; besides making many chemical analyses of raw sugars, I have analysed a great number of samples of crushed or refined sugar, manufactured in different sugar refineries throughout the country. When sugar is so very cheap, and every inducement thus offered to promote adulteration, when a new class of men has been constituted to watch over the food of the country, there is need for special communications upon this most interesting and important subject. Sugar, as met with in commerce, may be composed of crystallisable and noncrystallisable sugars, soluble salts, moisture, peroxide of iron, and other organic matter of undefined composition. The following analyses must not be con sidered as typical of the various kinds of sugar, for all of them vary exceedingly, but they give a fair idea of the composition of the sugars passed through the market. 100'000 100'000 100'000 100'000 100'000 100'000 100'000 100'000 100'000 100'000 100'000 100'000 100'000 100'000 100'000 Crystallisable sugar, 97 265 80°052 48734 88°475 74'074 45'916 83'877 95°377 49'956 obtainable Polarisation Degrees. 98.830 91060 71'561 94164 84°063 62761 90'863 97'158 82'513 99'795 95'548 92'348 90'341 86.000 51001 Per cent. Pernambuco. French Beet German Beet Sugar (low). Crystallised Refined Sugar. Crushed Crushed Crushed Crushed Refined Syrup. 99 999 99 999 99 999 99'998 99 998 99'994 99'999 100 111 99'999 99'998 99'999 Ico'069 100'000 99 999 99 998 have thrown very little light on the subject; for by care- It has been suggested to me, by Prof. C. F. Chandler, that the nickel, cobalt, and copper oxides present may form, with the sesquioxide of iron, a magnetic oxide, thus:-NiO+ Fe,O=(NiFe)04. The suggestion is a reasonable one, and worthy of consideration, The Crystallisable Sugar is best estimated volumetrically by Fehling's method. It must first, however, be converted into grape sugar, by protracted ebullition with dilute acid. I always conduct this experiment over a steam-bath, closing the neck of the glass flask with a perforated caoutchouc cork, carrying a glass tube bent in an obtuse angle which leads to a Liebig's condenser (turned upwards); the evaporated water thus flows back into the flask. After all the crystallisable sugar has been converted into grape sugar, the free acid is then neutralised by running in a dilute solution of pure sodium carbonate, and when perfectly cold it is diluted with water to a known volume, and then titrated in the usual way. One hundred parts of grape sugar are equivalent to 95 parts of crystallisable sugar. The results are not so satisfactory, as the boiling with dilute acid may convert some of the grape sugar into caramel. In analysing however, always found by difference, and is then quite sugars practically, the sum of the crystallisable sugar is, I reliable. NEWS Polarisation.-Frequently, recourse is had to the power which a solution of crystallisable sugar has of turning to the right the plane of polarised light. When, therefore, a liquid contains crystallisable sugar without any other substance possessing the same optical property, the quantity of sugar may be estimated by measuring the deviation which the liquid produces in the plane of polarisation. To apply this, the deviation must be known produced by a standard solution of pure crystallisable sugar; and from this we can readily calculate the deviation produced by any other standard solution of sugar the richness of which we wish to ascertain. Upon the theory of this optical method is constructed the saccharimetre-soleil of Duboscq and others. The polarising saccharometer, as considered from a practical point of view, is an exceedingly simple instrument. Sugar refiners not accustomed to manipulate it would readily mistake it at first sight for a family telescope; but, in reality, it is constructed of three parts, two of which are fixed, the other movable. The latter tube, which is of standard dimensions, is destined to hold the sugar solution the richness of which we wish to ascertain. There is a little movable tube, or eye-piece, to which the eye is applied, and which may be worked out and in until the polariscope has been properly focussed. The latter should be set in such a position in the laboratory that the flame of a gas lamp will shine through the whole arrangement, including the sugar solution, and care should be exercised to have the apartment as dark as possible. A mechanical arrangement, or button," serves to regulate the saccharometer-i.e., to cause the zero of the scale to coincide with the indicator. Another large horizontal"button" renders the observed tint, or shade, uniform, while a third arrangement assists in changing the colour of the disc best suited to the eye of the operator. There is also a graduated scale which supplies information about the richness in crystallisable sugar of the solution which may be under examination. Before beginning the experiment, obtain an argand lamp, and elevate it about 2 feet above the level of the laboratory bench. Secondly, cover it completely with a copper wrapper, in the side of which must be drilled a circular hole, corresponding exactly in size with the glass at the end of the polariscope. Now light the gas lamp, and trim the flame so that a circle of the yellow portion will be obtained in the brightest degree possible. Place the saccharometer close up to this, and in such a manner that the reflected light will traverse through it in an horizontal direction. Secondly, let a tube be filled with distilled water, and set it in the instrument in the place reserved for it, between the eye-piece and the object-glass; then let the polariscope be carefully worked into the focus which suits best the sight of the operator, and until a circular surface or disc is obtained, divided into two equal parts equally tinted or coloured, or two separate tints or colours parted from one another by a black line, which must appear well cut and quite perfect. If, as generally happens, the two half circles have not the same tint or shade, ie., they do not blend, then turn the large horizontal button from left to right, or right to left, until the tint of the two circles is perfectly identical, and until the eye can no longer discern any difference between them. Again, it is not enough that the two half circles should be of the same tint; it is, further, necessary that the experiment should be conducted with the greatest possible amount of accuracy, and that this tint of uniformity should be the most sensible one possible. Every analyst, in using the saccharometer, should find out the shade of colour best suited for his eyesight; he ought to adopt it in all his experiments, as it is not advisable to use a number of colours during polarisation. If, while looking through the saccharometer, we make the milled head turn, we shall perceive that the colour of the two half discs changes incessantly, and that they repeat after each half Supposing that the operator stops the rotation when the two half discs are of a yellow tint, both of which turn. should be identical, beginning from the instant the milled head is rotated very gently in the same direction, green follows yellow, blue, green, indigo-blue, and lastly violetindigo. If we look attentively, we shall come upon a certain shade, for which the uniformity of tint at first established for the yellow no longer exists; we shall see a difference which we might not at first notice. The same experiment repeated many times, and on different days, allows us to verify the shade which presents a different tint, when with another colour we see the quality of tint is always the same. Now, such a shade for polarisation is the most suitable, and the operator ought always to remember it and adopt it as his starting-point. For the greater number of experimenters, the delicate shade is the violet-blue tint, which recalls the colour of the flower of the foxglove, and for others yellow, or some brilliant colour. The violet-blue possesses this advantage, that if, while looking at it, we turn the milled head of the saccharometer infinitely little, one of the semicircles passes suddenly to red, the other to green. Consequently, upon rotation, the operator instantly stops whenever uniformity of tint is established. Observe if the zero of the scale corresponds exactly with the indicator; if this coincidence be not perfect, then make it so, and the polariscope is adjusted. Take out the tube of water, and substitute for it the tube containing the sugar solution. If, working by inversion, uniformity of tint no longer appears, re-establish by rotation and proceed as before. This done, note the mark on the scale to which the mark upon the indicator corresponds. If the saccharometer has been adjusted very accurately, the indicator should exactly register 100o, distilled water being o°. water. With regard to the preparation of the sugar solution, it is usually of such a strength that 16:35 grms. of pure crystallisable sugar are dissolved in 100 c.c. of distilled In polarising raw sugar solutions, it is best to precipitate the extractive organic matter from them, by the addition of a few drops of tri-plumbic acetate solution, and then pass the mixture through a filter of chemically pure animal charcoal. In this way, all traces of the lead salt are removed, which is absorbed by the charcoal, and the sugar solution is obtained as clear as water. In the case of beet-sugar solutions, the addition of the head is unnecessary, the charcoal filter being sufficient. For beet sugars, molasses, syrups, or "sweet waters," the saccharometer answers admirably, the rapidity with which the results are obtained from it, together with its extreme accuracy, places it at the top of all other methods for the estimation of crystallisable sugar. (To be continued.) PROCEEDINGS OF SOCIETIES. CHEMICAL SOCIETY, Thursday, May 6, 1875. Professor ODLING, F.R.S., Vice-President, in the Chair. AFTER the names of the visitors had been announced an i the minutes of the previous meeting read and confirmed, the names of Messrs. A. A. Nesbit, W. H. Martin, C. T. Blanchard, B.A., G. Crampton, J. C. Butterfield, J. W. Swan, A. W. Blyth, R. Stellon, G. Purdie, Jun., and the Rev. W. J. J. Welch, M.A., were read for the first time. Messrs. William Mogford Hamlet, Henry Mitchell Hastings, H. S. Carpenter, Alfred Southall, and William Alexander Little were duly elected Fellows of the Society after their names had been read for the third time. The first paper, "On Andrewsite and Chalkosiderite," was read by the author, Profe: sor N. Story Maskelyne. This mineral, which is named after Professor Andrews, is found in Cornwall in globular forms of a dark green 214 Supporting Crucibles in Gas Furnaces. CHEMICAL NEWS, May 14, 1875. colour, often associated with cuprite, and in habit singu- | phosphoric acid, alumina, ferric and ferrous oxide, and the larly resembling wavellite. It occurs on gangue of ferru- use of nitric acid, required for the molybdic acid process, ginous quartz, and in many respects resembles the well- necessarily complicated the matter of the determination known mineral dufrenite. It may be represented by the of the ferrous oxide. formula 2(2Fe2P2O8+Fe2H2O4)+CuH2O2. The other mineral which accompanies Andrewsite in bright green crystals was found by comparison with specimens of old date in the British Museum to be identical with the chalkosiderite of Ullmann. This mineral, which was examined before the blowpipe by Ullmann (who found the copper and iron, but missed the phosphoric acid) has hitherto been erroneously placed by mineralogists with dufrenite. It occurs in clustered groups or sheaves of blunt striated prisms, the minute size and the rounded striated faces of which render accurate measurements exceedingly difficult. Its composition may be represented as― 2Fe2P208+FezH606 + CuH2O2+4H2O. Professor ODLING having thanked the author, The SECRETARY read a communication from Dr. W. Flight entitled "An Examination of Methods for Effecting the Quantitative Separation of Iron Sesquioxide, Alumina, and Phosphoric Acid." After reviewing the various processes for the separation of phosphoric acid from alumina and also from iron, and giving the details of various modifications of some of these processes which he had tried, the author described the process ultimately adopted, which consisted in boiling the solution containing the alumina, iron, and phosphoric acid, and which should not be very acid, for two or three hours with excess of sodium hyposulphite. All the alumina with a part of the phosphoric acid was precipitated, whilst the iron with the rest of the phosphoric acid remained in solution. From this solution the iron could at once be precipitated by ammonic hydric sulphide, and converted into ferric oxide, whilst the alumina and phosphoric acid in the precipitate were separated by treatment with excess of caustic soda and barium chloride, which precipitates the phosphoric acid in combination with the barium, whilst the alumina remains in solution. In washing the precipitate a few drops of soda solution must be added to the wash-water, as the phosphate is decomposed by pure water. This phosphate was then decomposed by sulphuric acid, and the phosphoric acid determined in the usual way. In conclusion, the author noticed the decomposition which solution of sodium phosphate undergoes in contact with glass or porcelain from its removing the silica from it, and also a notice of some experiments made with a view to convert orthointo meta-phosphoric acid. The CHAIRMAN said they were much indebted to Dr. Flight for this process, which, in his hands, at least, gave such satisfactory results. Mr. WARINGTON said he was able to confirm many of the statements of the author, especially that phosphoric acid was not completely precipitated by magnesia when alumina was present. On the other hand, with iron the separation was perfect. One point of especial interest to him, as he had paid some attention to the decomposition of calcium phosphate by water, was the decomposition of barium phosphate under similar circumstances. There was, however, this difference, that the latter appeared to be more readily attaked, for cold water had a perceptible action on it, whilst the former was only decomposed by hot water. Dr. DEBUS understood Dr. Flight to state that Sonnenschein's molybdic acid method in the presence of alumina and iron, although accurate enough for ordinary purposes, was not so for scientific purposes. He, however, gave no comparison of results obtained by the two processes. As far as he was himself concerned, he had always obtained excellent results by this method. Professor MASKELYNE, in reply, said his own impression was that the barium phosphate was slightly soluble in pure water, and was not decomposed by it. With regard to Sonnenschein's method, the objection to it was that in this particular instance it was necessary to determine A paper "On Sodium Ethyl-Thio-Sulphate," by Mr. W. RAMSAY, was then read. The author finds that the precipitates obtained on adding silver nitrate or a mercuric salt to sodium ethyl-thio-sulphate are very unstable, being readily decomposed, with formation of metallic sulphides; also, that when the sodium salt is heated with phosphoric chloride, phosphoric oxychloride and ethyl disulphide are produced, but no sulphuryl chloride, as stated by Spring. At the same time, a small quantity of a brown oil is formed, which is freely soluble in alcohol and ether with a beautiful crimson colour, but only sparingly soluble in water. Mr. JOHN WILLIAMS then read a paper "On a Milligrade Thermometric Scale." After pointing out the disadvantages attending the use of the centigrade scale, arising from the size of the divisions, and the comparatively high temperature of the zero-point, necessitating the use of the minus sign, the author suggested the substitution of the freezing- and boiling-points of mercury for those of water, fixing the one at -40° C. and the other at 360° C., and dividing the scale into 1000 parts. The freezing-point of water would then be 100°, and its boilingpoint 350°, whilst 5 milligrade degrees would be equal to 2° C. Dr. ODLING remarked that when the centigrade thermometer first began to come into use in this country, the objections made to it were the largeness of the degrees and the high zero-point. It possessed the great advantage, however, that the determination of the melting-point and boiling-point of water could be made easily and accurately. The enormous amount of literature and of work in which the centigrade was used would be a great obstacle to the introduction of Mr. Williams's milligrade scale. The last paper was "On a New Method of Supporting Crucibles in Gas-Furnaces," by Mr. C. GRIFFIN. In the author's gas-furnace, a description of which was communicated to the Society in 1870, the perforated plumbago cylinder, and the trivet-grate on which the crucible is supported, are liable to break when white-hot, occasionally giving much trouble; moreover, the latter has the disadvantage of interfering with the direct action of the flame on the crucible. This, however, is entirely obviated by the new burner, in which a space is left round the central jet, which has fitted over it an atmopyre similar to those used in Hofmann's combustion-furnace. bottom of the crucible rests on this, and the plumbago cylinder is thus relieved of all pressure. These new burners are very economical and of great power, a small one, burning 20 feet of gas per hour, being capable of melting half-a-pound of cast-iron in thirty-five minutes; or of heating a muffle, 5 inches long and 3 wide, to a temperature sufficiently high for assaying. Several varieties of the furnace were exhibited, one of which was in action. The Dr. ODLING having thanked the author for this addition to the many pieces of apparatus which he and his father had introduced to the notice of chemists, the meeting was adjourned until Thursday, May 20, when the following papers will be read : (1.) "Note on Milk in Health and Disease," by A. Smee, jun. (2.) "The Effects of Pressure and Cold upon the Gaseous Products of the Distillation of Carbonaceous Shales," by J. J. Coleman; (3.) "On some Nova Scotian Triassic Trap Minerals," by Professor How; (4.) “On some Points in the Examination of Waters by the Ammonia Method," by W. H. Deering; (5.) "On the Agricultural Chemistry of Tea Plantations of India," by D. Campbell Brown; (6.) "On the Structure and Composition of Certain Pseudomorphic Crystals having the Form of Orthoclase," by J. A. Phillips; (7.) "On Nitrosyl Bromide and on Sulphur Bromide," by M. M. P. Muir. |