Decinormal soda-lye potassium, a fluid amalgam is produced which remains | duce a distinct red colouration. consumed equal quantities of decinormal hydrochloric Following up these reactions several mixtures were obtained for reduction, by calcining sodium tartrate in admixture with a sufficiency of potassium tartrate to allow of about 5 or 6 per cent of potassium to pass into the distillate; the reaction being brought about in wrought: iron tubes, and urged to whiteness by the aid of a small blast furnace. The globules of metal thus obtained presented in appearance an exact similitude with the first-mentioned alloy, and were considerably more volatile than pure sodium. The iron tubes after cooling were cut longitudinally, and in no instance could any reduced metal be detected; the metals as fast as reduced having been volatilised, producing the aforesaid alloy. The calcining of tartrates on a large scale would natu rally be entirely out of place; but the introduction of potassium could, as far can be seen, be brought into play when employing any of the ordinary commercial methods, save that the percentage of the two would probably have to be looked into still closer in order to obtain concordant results. Everton Research Laboratory, 18, Albion Street, Everton, Liverpool. DETERMINATION OF SMALL QUANTITIES FOR the preparation of mille-normal solutions we must use pure or, at least, neutralised water, since otherwise the alkaline impurities derived from the glass vessels in which the water is preserved have a decidedly disturbing effect. Mille-normal solutions can be kept for some time in bottles of good glass. The titration is advantageously conducted in a small stoppered bottle, from 50 to 100 c.c. of the aqueous liquid under examination being shaken up with 10 to 20 c.c. of an ethereal solution of iodeosine containing 2 m.grms. of the colouring matter per litre. Such a solution, which appears almost colourless, can be obtained by diluting the ethereal solution of iodeosine to be mentioned below. If free alkali is present, the lower stratum, after shaking, appears rose-coloured, whilst on neutralisation with alkali it becomes colourless again. The titration is best completed by again adding so much solution of alkali that a distinct rose colour reappears. The small excess of about 2 c.c. thus produced must be deducted from the total quantity of solution of alkali consumed. The influence of carbonic acid upon titration with mille-normal solutions, according to the process above described, is not important. Even on titration with decinormal solutions, using an ethereal solution of iodeosine, it interferes little, as is shown by the following experiment:-For neutralising 407 c.c. decinormal sulphuric acid, which, when tested with litmus (the carbonic acid having been previously expelled by heat), required 40°7 c.c. of a decinormal solution of sodium carbonate, there were consumed, when an ethereal solution of iodeosine was used as indicator, 41°2 c.c. of the soda solution in order to pro The following instances may show the varied applicability of the method: water. The carbonates dissslved in spring- or well-water may be quickly and easily determined. The acid consumed is equivalent to the "combined "carbonic acid in the sumed for neutralisation 15.2 c.c. N/1000 sulphuric acid, 5 c.c. water from the Charlottenburg mains conwhich corresponds to 6.7 grms. of "combined" carbonic acid in 100,000 water. Titrations of drinking-water have been effected with centinormal solutions, using carmine as indicator. The solubility of calcium carbonate in pure water which has been repeatedly determined can be readily shown by this method. Finely pulverised marble was well washed, and then left for some time in contact with neutral (see below) water which had been previously deprived of carbonic acid by ebullition in platinum vessels. 25 c.c. of the solution thus obtained, after the lapse of seven days, required for neutralisation 15'5 c.c. N/1000 sulphuric acid. A litre consequently contained 31 m.grms. calcium carbonate in solution. The decom. position of neutral salts of ammonium, when their solutions are heated, can be shown by means of titration with mille normal solutions. A decinormal solution of commercial crystalline ammonium sulphate in "neutral" water, which was itself neutral, was distilled in the apparatus to be described below. Appreciable quantities of ammonia were present in the distillate; 200 c.c. were three times distilled off in succession; 25 c.c. of the first distillate required for neutralisation 32.5 c.c. N/1000 sulphuric acid; the same quantity of the distillate required 34'5 c.c., and of the third 37 c.c. of the acid. Hence it appears that the concentration of the solution in the distillatory vessel increases little. The ammonia present in the 600 c.c. of water distilled was o'01416 grm. NH3, 1.e., 0.83 per cent of the total ammonia present. The eosine reaction is particularly useful for testing the neutrality of salts. If a salt is considered neutral the solution of which, on shaking with ethereal solution of eosine, is turned red by a minimum of mille-normal soda-lye, many of the salts sold as pure are not neutral. Thus, e.g., 5 grms. of Kahlbaum's sodium chloride required for the production of a red colour 11 c.c. N/1000 soda-lye, corresponding to the presence of o'0008 per cent hydrochloric acid. The same preparation, after ignition, was found to be faintly alkaline, since 5 grms. required 1-4 c.c. N/1000 sulphuric acid for neutrality. As regards the alkaline sulphates, further observations are needed. We merely remark that a commercial and apparently pure specimen of potassium sulphate had a distinctly alkaline reaction. The recognition of the neutrality of salts of the zinc group has hitherto been especially difficult, but iodeosine was here found to be the most sensitive indicator. The hydroxides of metals of the zinc group, like the alkalies and alkaline earths, form with iodeosine soluble salts of a reddish violet. In the solutions of the basic salts a red colour appears on shaking up with ethereal solution of iodeosine; but the slightest excess of free acid prevents the appearance of salts of colouring-matters. For the production of neutral zinc sulphate, the commercial salt, having an alkaline reaction, was twice recrystallised from a carefully neutralised solution; 5 grms. of the crystals dissolved in water and shaken with ethereal solution of eosine gave a colourless solution; but 2 c.c. N/1000 alkali sufficed to produce a distinct rose colour. (To be continued). NEWS SINCE De la Kive in 1830 made the observation that chemically pure zinc is almost insoluble in dilute sulphuric acid, this puzzling and interesting phenomenon has been the subject of research, but without meeting with a satisfactory explanation. The question derives increased importance from the fact that other metals, if pure, and other acids behave in a similar manner. Nitric acid alone attacks most metals considerably, even if chemically pure. Since it has been found that the same metals, which in a chemically pure state were almost insoluble in acids, are always more or less rapidly attacked by the same acids when impure, the attention of investigators was turned to the impurities and the adherents of the contact theory recognised in this different behaviour of the pure and the impure metals,-a defence against the attacks of the supporters of the chemical theory. According to the contact theory the electric current arises in a closed circuit (Zn, Pt, H2SO4) by the origination of a difference in the electrical potential due to the contact of the two metals. This difference tends to be equalised by the acid. But as an electrolyte-in this case the sulphuric acid-cannot conduct the electric current without itself being decomposed, such a decomposition of the acid is effected by the current which passes from the zinc through the sulphuric acid to the platinum. The hydrogen is set free upon the platinum and the acid radicle, in virtue of its liberated affinities combines, with the zinc to form zinc sulphate. In brief, the contact theory views the solution of the zinc as a consequence of the electric current running from the zinc to the platinum in consequence of the electric opposition of the two metals; the zinc must remain undissolved as long as no electric difference is present. Hence impure zinc, but not pure zinc, dissolves in acids. The contact theory thus seemed to solve the problem in a simple and natural manner, and in consequence it met with almost universal acceptance. But this explanation did not agree with the facts that chemically pure metals are, in general, easily dissolved by dilute sulphuric acid, if boiling, as also by cold nitric acid. These reasons led me to give up the theory above given, and to seek elsewhere for the cause of the insolubility of pure metals in acids. In this investigation I ultimately succeeded and found an explanation for the phenomena at once simple and comprehensive. I maintain that chemically pure zinc or other chemically pure metals are insoluble or sparingly soluble in acids, because at the moment of their immersion in the acid they are at once enclosed in an atmosphere of condensed hydrogen, which under normal conditions makes any further action of the acid impossible. This im. measurably thin but everywhere continuous layer of hydrogen is alone the cause of the insolubility or sparing solubility of chemically pure zinc in dilute acids. Absolute insolubility is from various reasons out of the question if we consider that both the varying density of the surface and minimal traces of foreign metals, such as we may suppose even in chemically pure zinc, must give rise to very feeble local currents, which are of course attended with a slight solution of the zinc. It is known that on the action of nitric acid upon the zinc there are formed, according to the concentration of the acid, NH3, N2O, NO, N2O3, and N2O4. There is at first formed upon the zinc hydrogen, which in its nascent tate is at once oxidised to water by the excess of nitric acid, whilst the latter is reduced to the above mentioned nitrogen compounds. Where then are the latter formed? Certainly not directly on the surface of the zinc, from which they are separated by that stratum of water which has been formed by the oxidation of the nascent hydrogen. This stratum is certainly very thin, but it suffices to abolish the attraction between the metal and the gas which is formed. Zinc, therefore, on treatment with nitric acid, is never enclosed in a protective coating of gas, but it is exposed on its entire surface without protection to the attacks of the nitric acid. The conditions of impure zinc if treated with sulphuric acid are similar. The hydrogen formed is not liberated on the zinc, but on the accompanying impurities which are more negative than zinc. Consequently the true surface of the zinc remains always free from gas, so that the action between metal and acid goes on continuously. Similar results are obtained also with cadmium, cobalt, nickel, iron, and aluminium. Aluminium, which is otherwise scarcely or not at all attacked by dilute sulphuric or nitric acid, dissolves easily in these acids in a vacuum. It dissolves also readily in a vacuum in a perfectly neutral solution of ferric chloride which is reduced to ferrous chlo. ride with an escape of hydrogen. At the ordinary atmospheric pressure this process is much more tedious. It would be interesting to ascertain whether hydrogen gas alone possesses the property of condensing in its nascent state upon metals, and of thus rendering them unattackable by acids, or if this property is common to other gases. If this is not the case we are entitled to assume that in these phenomena of adhesion the metallic character of hydrogen plays a part of greater or less importance, and that we have here a transition to the compound of palladium and hydrogen. -Ber. Deut. Chem. Gesell., vol. xxiv., p. 1785. ELECTROLYTIC DETERMINATION OF METALS AS AMALGAMS. By G. VORTMANN. (Continued from p. 228.) Determination of Zinc. I can IN 1885 C. Luckow recommended the separation of zinc as amalgam from a weak sulphuric solution. confirm the accuracy of his statement, but a large excess of free acid may completely prevent the separation. I therefore preferred to undertake the precipitation of the zinc from the solution of the double ammonium oxalate or from an ammoniacal solution. 1. Separation of Zinc from the Solution of the Double Oxalate.-The zinc salt along with a weighed quantity of mercuric chloride was dissolved in water in a platinum capsule and mixed with an excess of ammonium oxalate (3-5 grms.). On electrolysis the zinc amalgam was deposited upon the platinum capsule as a silvery white layer and could be washed without loss. To ascertain whether the precipitation was complete a portion of the liquid was tested with ammonium sulphide. The mercury must not be in too large a proportion to the zinc (to one part zinc not more on the outside than 2-3 parts mercury), as the zinc is otherwise deposited in a spongy form. Moreover zinc is easily deposited without mercury, the addition of which has merely the purpose to facilitate the removal of the deposit from the platinum capsule. Like the other amalgams zinc amalgam does not require to be transferred rapidly to the balance, as it neither loses nor gains weight if left in the exsiccator for a considerable time. 2. Separation of Zinc from an Ammoniacal Solution.The precipitation of zinc as an amalgam from an ammoniacal solution containing tartaric acid takes place similarly to that from the double oxalate; a larger quantity of mercuric chloride must, however, be used (to I part zinc at least 3 parts of mercury), as the amalgam otherwise falls off in scales. A large excess of mercury is in this case not hurtful. * Berichte Deutsch. Chem. Gesell, The amalgam, whether deposited from solution of ammonium oxalate or from an ammoniacal liquid, is readily dissolved by dilute nitric acid; there remains a black powder consisting of platinum black which adheres to the capsule but is easily removed by friction. If the capsule is weighed after removal of the zinc amalgam, rinsing with alcohol and ether, and desiccation, it has its original weight, or it weighs only 1-2 m.grms. more than before the electrolysis. After the removal of the black powder, the weight of which is sometimes very appreciable, the loss of weight of the capsule is sometimes only a few m.grms., but occasionally may reach 5 centigrms. On account of this loss of weight the separation of zinc as amalgam cannot be recommended, and the previous method, according to which the capsule is first coated with copper or silver, is to be preferred. Determination of Cadmium. Cadmium, like zinc, can be separated as an amalgam, either from the solution of the double ammonium oxalate or from an ammoniacal solution. The latter is more uniform if the mercury is present in large excess (at least 4 parts to I part of cadmium); if but little mercury is present the amalgam separates in a crystalline form and does not adhere sufficiently to the platinum, so that it cannot be easily washed without loss. If the amalgam contains 4-6 parts of mercury to 1 part cadmium, it is so hard that it may be rubbed with the finger without removing any portion. If the quantity of the mercury is eight times as large as that of cadmium, the amalgam is in part liquid. It does not become oxidised in the air, and can be preserved for several days without any change in its weight. Dilute nitric acid dissolves it readily, and as a rule, without a residue of platinum black; rarely is a faint blackness perceptible. As cadmium ammonium oxalate is less soluble in water than the corresponding zinc salt, and as a larger quantity of mercuric chloride must be dissolved, the oxalate method is suitable only for the separation of small quantities of cadmium (about 0'2-03 grm.). The solution containing the salt of cadmium and the mercuric chloride is mixed with about 5 grms, ammonium oxalate, stirred up, and diluted as far as possible with water. The ammonium oxalate must be dissolved in the cold, as on the application of heat a precipitate of mercurous chloride may be produced in a concentrated solution. If the quantity of cadmium exceeds o'3 grm. an ammoniacal solution must be used. In this case the solution of the cadmium and the mercuric salt is mixed with about 3 grms. tartaric acid and then with ammonia until it smells strongly, diluted with water, and exposed to the electric current until a portion of the liquid is not rendered turbid by an addition of ammonium sulphide. Determination of Lead. rapidly rinsed with water, alcohol, and ether, quickly dried by warming the capsule in the hand whilst a current of air is directed upon it, and then placed in the exsiccator. The lead amalgam, when dry, is permanent in the air, and does not vary in weight if it remains for twenty-four hours in the exsiccator. In a moist state, it is readily oxidised; hence, after rinsing with water, that liquid must be removed as quickly as possible with alcohol and ether. Oxidation, however, is not so rapid as to render the determinations inaccurate; the moist amalgam undergoes no change for about five minutes. It then gradually loses its lustre, and is coated with a yellowish white film. It is readily soluble in nitric acid, and leaves on the platinum capsule a very faint stain of platinum black. The separation of lead as amalgam can also be effected in an aqueous solution acidulated with dilute nitric acid, to which a little potassium nitrite is added in order to prevent the formation of lead peroxide. As the quantity of nitrite which a liquid containing free nitric acid can take up without a violent escape of nitrous acid is very trifling, there is formed at the positive electrode during electrolysis a thicker layer of lead peroxide than in an acetic solution. It must be re-dissolved by the frequent addition of a few drops of the solution of potassium nitrite. The analysis is in this case more tedious than with an acetic solution, and must be continued until no more lead peroxide is deposited at the positive electrode, and until no lead can be detected in the liquid by means of ammonium sulphide. The determination of lead as amalgam has the advantaee of being applicable to larger quantities of the metal than the process at present in use. (To be continued). DETERMINATION OF MANGANESE IN MANGANIFEROUS SLAGS AND ORES. THE manganese in slags from blast furnaces making spiegel, and in open hearth slags, is easily and rapidly determined in the following way: One grm. of the finely powdered slag is placed in a 4-ounce Griffin beaker, moistened slightly with water to prevent caking, and 50 c.c. of nitric acid of 1'42 sp. gr. added. Bring to a boil, and while the slag is in suspension add 3 or 4 c. c. of hydrofluoric acid. The slag is rapidly decomposed and the silica driven off. Boil the solution for a few minutes to drive off any remaining hydrofluoric acid, then transfer to a larger beaker or a precipitating flask, add more nitric acid, bring the solution to a boil, and precipitate the manganese with potassium chlorate. Filter off the manganese dioxide upon an asbestos filter. The manganese dioxide may either be dissolved in standardised oxalic acid or ferrous sulphate solution, and the excess of the solvent titrated with permanganate of potassium, and the amount of manganese calculated; or the manganese may be finally weighed as the pyrophosphate. 1. Separation of Lead from Acid Solutions.-In the determination of lead as an amalgam, I encountered difficulties, as the lead was partly separated out at the positive electrode as a peroxide, and cannot be removed readily by reducing agents, or, in some cases, not at all. Tartaric and oxalic acids, ethylic alcohol, potassium iodide had little effect. The removal of the lead peroxide was best effected by means of nitrous acid. I dissolved the lead salt and the mercuric chloride in water; adding 3-5 grms. sodium acetate and a few c.c. of a concentrated solution of potassium nitrite. The white precipitate formed was dissolved in acetic acid. The clear yellow solution thus obtained was submitted to electrolysis. As long as the liquid contained nitrous acid, no lead peroxide was formed, but towards the end of the process there appeared on the positive electrode a brown stain, which quickly disappeared on the addition of a few Drive off any remaining hydrofluoric acid, and be sure drops of solution of potassium nitrite. When a portion that the tartaric acid is all decomposed. Add more nitric of the liquid gave no colouration with ammonium sul-acid, boil, and precipitate the manganese with potassium phide, the liquid was decanted off, the lead amalgam was For manganese in ores:-1 grm. or o'5 grm., according to richness of the ore, is placed in a 4-ounce Griffin beaker, moistened with water, and 50 c.c. nitric acid of 1'42 sp. gr. added. Bring to a boil and add a few small pieces of tartaric acid. Ores made up largely of the mixed oxides of manganese go into solution readily. When the solution clears, add a little hydrofluoric acid to decompose the residue. Where there is much oxide of iron in the ores it is better to add the hydrofluoric acid along with the tartaric acid. chlorate. NEWS 13, The nitric acid breaks up all the other oxides of manganese into the oxide and dioxide, and dissolves the monoxide, while the tartaric acid causes the solution of the dioxide. Considerable tartaric acid may be used, as it is readily decomposed by boiling in the nitric acid solution. By this method the manganese can be determined in about forty-five minutes, while a much longer time is required by the ordinary way of solution in hydrochloric acid, with separation and fusion of the residue and consequent accumulation of salts.-Journal of Analytical and Applied Chemistry, August, 1891. as is used in the humid assay of silver, dilute nitric acid is added and heat applied. When action ceases, the silver present in solution is determined volumetrically as usual. The silver found, less the amount added, gives the silver in 500 m.grms. of bullion. There are two objections to this method: (1) The insolubility of lead nitrate in nitric acid necessitates the use of dilute acid, which leaves much alloy with the geld; (2) the alloy to be removed is one-third silver. This is very important when small quantities of silver are to be determined, as in the case of gold coin, which rarely contains more than o'003. Cornets, after three boilings with acid (ten minutes each), retain from o'oor to o'002 of silver; so it will be seen at a glance that the amount retained after one boiling with dilute acid will be A METHOD OF OBTAINING AMMONIA-FREE much greater, and the gold residue will contain more WATER. By D. B. BISBEE. A VERY easy way of obtaining ammonia-free water, which I have used for some time and have never seen mentioned, is to acidulate the water, before distilling, with sulphuric acid. The acid holds all ammonia in the retort, the first portions even of the distillate being ammonia-free. But this acidulation naturally causes the nitric and nitrous acids in the water to distil over. For some purposes, however, nitrates are not objectionable. At the laboratory of the Iowa Agricultural Experiment Station, Ames, Iowa, when we wish to obtain chemically pure water for any use, we take distilled water, which is nitrate-free, acidulate with sulphuric acid, and distil, at once getting pure water. Journal of Analytical and Applied Chemistry, August, 1891. THE USE OF CADMIUM IN ASSAYING By CABELL WHITEHEAD, CADMIUM, as a substitute for silver in assaying gold bullion, was first used by Balling (Crookes's "Select Methods in Chemical Analysis"). He states that the gold is entirely parted from all metals, except the platinum group, when its alloy with cadmium is boiled with strong nitric acid for one hour, followed by a second boiling for ten minutes with fresh portion of same acid. For general assays of gold bullion I do not think that Balling's method with cadmium will bear comparison in point of accuracy with the old one of quartation with silver and cupellation. But, with certain modifications which will be suggested later, it will be found rapid and satisfactory for a preliminary assay. It is, however, in the estimation of small quantities of silver in gold bullion containing considerable amounts of copper or platinum, that I have found cadmium to be a most efficient aid. In The difficulty in determining silver in the presence of platinum by cupellation is well known to assayers. such cases the following has been the method in the United States Mint at Philadelphia: An approximate assay is made by cupellation, after which, and based on this approximation, enough pure silver is added to 500 m.grms. of the bullion to make in all at least 1004 m.grms. of silver present in the assay. This weighed silver and bullion is wrapped in a sheet of lead weighing 25 grms., and the whole placed on a hot cupel in the muffle furnace. As soon as the lead "clears" the cupel is withdrawn ; when cold the button is flattened and put in a bottle such A communication to the Chemical Section of the Franklin Institute, Sept. 15, 1891. silver than was present in the original bullion. The presence of 10 per cent or more of copper in high grade gold bullion makes the accurate determination of silver by cupellation impossible. Not only is gold taken into the cupel in large quantities by the copper, but it is also left in specks over the entire surface covered by the assay, thus making a "proof" practically worthless. The method which I am about to describe was devised for the estimation of silver in gold coin, and has been in use in the laboratory of the Bureau of the Mint for the past year, where it has given such satisfactory results that it is thought a brief description may prove of interest to others engaged in the same line of work. Five hundred m grms. are weighed into a porcelain crucible and covered with 10 grms. of potassium cyanide. The potassium cyanide is melted over a Bunsen burner or preferably a blast lamp. When the cyanide is in quiet fusion I grm. of cadmium is dropped into the crucible, where it quickly melts and forms a bright, homogeneous alloy with the gold. After gently shaking, so as to bring the cadmium in contact with every particle of bullion, the crucible is removed and the contents poured on a clean porcelain slab, where it soon solidifies. The alloy will be found in one piece and is easily detached from the potassium cyanide. It is now washed in warm water, dried, and placed in a diamond mortar, when several sharp blows with a hammer quickly reduce it to powder. This powder is carefully transferred to an assay bottle, 1004 m.grms. of pure silver added and 10 c.c. of nitric acid-32° Baumé-poured on. In from five to ten minutes (depending upon the heat used) the solution is complete and all action has ceased. The bottle is now cooled and 100 c.c. of normal salt solution is charged and the bottle shaken. The precipitation is finished with the decinormal solution. This assay is accompanied by another called a "proof," made of 1004 m.grms. of pure silver dissolved in the same amount of acid. Now the excess of silver found in the assay, over that shown in the "proof," is the amount contained in 500 m.grms. of coin. This doubled gives parts of silver per thousand. Example :-An alloy composed of 499 m.grms. pure gold, I m.grm. of silver, and 1 grm. of cadmium, treated as above described, after being charged with 100 c.c. normal salt solution, required 5 c.c. decinormal solution for complete precipitation of the silver present. A proof assay, carried along as check, upon 1004 m. grms. of pure silver, required in addition to 100 c.c. of normal salt solution, 4 c.c. of decinormal solution for complete precipitation to 1004 c.c. decinormal solution. Hence, each c.c. decinormal equals I m.grm. silver, and I c.c. decinormal solution required by the bullion, in excess of that called for by the silver added, shows the bullion to contain I m.grm. silver in 500, or 2 parts per 1000. = It may be asked by those not familiar with mint appliances and usages, "Why not titrate directly the silver brought into solution with the cadmium instead of adding a known weight of pure silver and finding the desired result by difference?" on URANINITE, AND ON THE COMPOSITION The reply is that the small amount of silver present in | ON THE OCCURRENCE OF NITROGEN IN this class of bullion would, as chloride, not "clear shaking, and much time would be consumed in finding the end reaction. By the method described the usual apparatus and solutions may be availed of and results rapidly obtained. When no such reasons exist the sulpho cyanide method alluded to at the end of this article is recommended. Example in Gold Bullion. An approximate assay gives by cupellation 0.035 silver, hence 500 m.grms. will contain about 175 m.grms. of that metal, and 986.5 m.grms. must be added to bring the total silver up to 1004 m.grms. If copper is not present about 50 m.grms. is added, it being found that the alloy of copper with cadmium is very brittle, and the resulting button is easily crushed to powder. The sample having been fused with cadmium in presence of potassium cyanide is powdered and subjected to treatment with nitric acid as above described. After charging with 100 c.c. normal salt solution, and shaking, the assay required 4'5 c.c. decinormal solution for end reaction. A "proof" consisting of 1004 m.grms. pure silver in solution, treated in same manner, required but 3 c.c. decinormal solution. Hence the assay contained 4'5-3=1.5 m.grms. more silver than the proof, or 1004 +15=1005'5 m.grms. in silver in all. This, less the 9865 m.grms. silver pur. posely added, gives 19 m.grms., as the silver present in 500 m.grms. of bullion taken, or thirty-eight parts in each thousand instead of thirty-five parts found by cupel. lation. In favour of the new method it may be said (1) That the ready solubility of cadmium in nitric acid of any strength makes it possible to dilute (if the term may be used), the silver present in gold bullion, to any desired extent, while, on the other hand, the difficult solubility of lead necessarily limits the dilution possible by the time required for its even imperfect extraction. The inevitable small portion of other metals retained by the gold after treatment with nitric acid may therefore, by the use of cadmium, be made to contain but an infinitesimal quantity of silver. (2) The brittleness of the button obtained permits its being crushed to a powder, in which condition the alloy rapidly yields its soluble portion to nitric acid, and the time required for an assay is materially shortened. (3) The low temperature required enables the chemist to dispense with all special appliances. A Bunsen burner will well answer the requirements for heating purposes, and little more is needed beyond a standard salt solution. No muffle or rolls are wanted. In a laboratory where few assays are made the following method might be followed and very satisfactory results obtained. After alloying and crushing, treat in a parting flask with 15 c.c. of nitric acid, 32° Baumé, for ten minutes; pour off this acid into a beaker, add 15 c.c. acid same strength and boil for ten minutes longer, pour off again in same beaker, wash with hot water and take out in an annealing cup, dry and heat over blast lamp, weigh and deduct o25 m.grm. for cadmium retained. Twice this weight gives the goid fineness. The acid and washings are evaporated to drive off free nitric acid, and silver determined either as chloride, or volumetrically with sulphocyanide with ferric indicator. Cadmium nitrate does not interfere in the least with determination of silver, either as chloride or as sulphocyanide. Detection of Paraphenetidine in Phenacetine.— L. Reuter. The author introduces o'5 grm. of the speci men into 25 grms. of melted chloral hydrate and mixes. If the phenacetine is pure the solution is clear and colourless. But mere traces of paraphenetidine colour the solution at once of a more or less intense violet.Zeitsch. f. Anal. Chem., xxx., Part 3. OF URANINITE IN GENERAL.* By W. F. HILLEBRAND. Detection and Examination of Nitrogen. Ar an early stage in the work it was noticed in a few instances, where the sealed tubes in which UO2 was to be estimated were left in an upright position for a short time, that an extremely slow disengagement of gas took place, which was especially noticeable if the tubes were gently tapped on the table occasionally, whereby the bubbles enmeshed by the powder were set free and rose through the liquid in the tube. This was supposed to be due to the presence of traces of carbonates in the mineral, and no particular attention was paid to it, although it seemed strange that they should not be entirely decomposed in a very few moments after contact with the acid. Later, when making the first estimation of UO2 in Glastonbury uraninite above mentioned with hydrofluoric acid, solution in this case being made in the cold and with free access of air, it was observed that this slow escape of gas was continuous throughout the whole ten days required for decomposing the mineral. This seemed to exclude the possibility of its being carbonic acid, and it was determined to collect and examine it if possible. This was done in a manner to be hereafter described. It may be here mentioned that U3O8 and UO2 prepared from the residues obtained during this work, when examined under conditions precisely similar to those obtaining with uraninite, failed to give off the least trace of gas. The gas was colourless, odourless, a non-supporter of combustion, unchanged by mixture with air, neutral to moistened litmus papers, not absorbed by caustic alkalies, and insoluble in water; at least its coefficient of absorption was so small as to be inappreciable without elaborate experimentation. In a Bunsen's absorption tube, containing gas from the original bröggerite, a potash ball caused no diminution in volume, neither did potassium pyrogallate, thus showing the absence of carbonic acid and of oxygen. Transferred from the absorption tube to a eudiometer, the gas was then subjected to the tests prescribed by Bunsen ("Gasometrische Methoden," 2nd ed., pp. 73, 74. Two-thirds of which, or 32'67, is almost exactly the volume of the hydrogen introduced. Were it not for the relatively large contraction after the first explosion, o'80 the above tests would indicate almost incontestably that units, representing 5'71 per cent of the initial volume, the gas could be nothing but nitrogen. Considering the small volume operated upon, and the great disadvantages under which the eudiometric experiments were made, the above is not a surprising error, and, since in other experi ments it did not recur, I have no hesitation in regarding it as without significance. Gas liberated from bröggerite by hydrochloric acid had, after treatment in the absorption tube, in the eudiometer, From "Bulletin No. 78, U.S. Geological Survey, 1889-90." An abstract of this paper was published in the Am. Jour. Sci., vol. xl., p. 384. |