bly, three phenol H's replaced by Sb as well as three acid H's. This formula is deduced from the following analyses of antimony tannate. All the tannates described in this paper were prepared in the same manner. The solution, containing five to ten grams of tannic acid per liter, was heated to 60° C. in a water bath. Tartar emetic in strong solution was added, then 30 c.c. of ammonium acetate per liter. After the whole was shaken and allowed to settle, it was filtered and dried at 100° C. to 105° C. for three or four days in an ordinary air-bath. The bulky, yellowish-white, gelatinous precipitate at first formed became, when dried, yellowish to reddish-brown, transparent, amorphous, and broken into small angular fragments. The antimony was determined as a sulphide, and the calculated results are probably a little too high. As to the process of titration, our first experience coincided with the statement of Gauhe (Zeitschrift für Analyse, 1863, iii, page 122), that the end of the reaction was difficult to seize, and that the dilute solutions remained turbid. Even after we found an indicator, the process in our hands gave varying results with varying quantities of ammonium chloride and with different proportions of water. We then made a series of tests with other substances, viz: alum, salts of sodium, etc., as precipitating agents. The one chosen as a result of these tests was ammonium acetate, prepared by mixing in the right proportion glacial acetic acid and strong ammonia water of known strengths.* 1 c.c. of this preparation added for every 25 or 30 c.c. of the total bulk, will give a clear, supernatant liquid after standing a few minutes. Without stating in detail the steps of the investigation, we give our process as we now use it. * We used acetic acid containing 95 per cent C2H.O2 and ammonia containing 27 per cent NH3, making about 600 grams NH.C2H,O2 for one liter of the solution. AM. JOUR. SCI.-THIRD SERIES, VOL. XVI, No. 93.–SEPT., 1878. The weighed quantity of the substance in which the tannic acid is to be estimated is taken in sufficient quantity to allow of, at least, three aliquot parts, each portion of 50 to 100 c.c. containing 100 to 300 grams of tannic acid. After the solution, made by digestion with water, is made up to a known bulk, three or four portions are measured out and set in a water bath to be heated to 50° or 60° C. The standard solution of tartar emetic contains 6.730 grams per liter of the CH,KSO, salt dried at 100° C. 1 c.c. is considered to correspond to 010 gram of tannic acid. Gerland's formula would require 5.222 grams per liter for the same value. The estimation is facilitated by obtaining a maximum and a minimum point at the first reading, as one portion is settling while the other is being treated; therefore tartar emetic is added from a burette to one portion in excess of the probable quantity required, and to another in less amount. The antimony tannate is then precipitated by the requisite number of cubic centimeters of ammonium acetate, and allowed to settle. A drop of the clear liquid is added to a drop of sodium hyposulphite on a hot porcelain plate, and if the tartar emetic has been added in excess, the deep orange color of the antimony sulphide will at once appear. When this point is reached by successive additions of the standard solution to the minimum portion, we add to a third portion the estimated quantity, and test the clear liquid as a check on the loss occasioned by taking out several drops. We have found it easier to carry the titration to a decided orange tint, and to subtract 5 c.c. of tartar emetic solution for 100 c.c. of liquid, rather than to try to seize the first faint tinge, as most of the substances to be titrated contain coloring matter which give a yellowish or reddish tint, but not an orange color. Gerland states that neither gallic acid nor the coloring matter contained in certain substances affects the results. This seems to be true so far as gallic acid is concerned, but the discussion of the relation of the coloring matter to the precipitate, together with the results of our titrations and combustions of antimony tannate from hemlock bark, oak bark, sweet-fern leaves, etc., must be reserved for a future paper. Massachusetts Institute of Technology, Woman's Laboratory, July, 1878. ART. XIX.-On some Seleniocyanates; on the Electrolytic Estimation of Mercury; some Specific Gravity Determinations. Being Parts VII, VIII and IX of Laboratory Notes from the University of Cincinnati; by F. W. CLARKE, S.B., Professor of Chemistry. VII. On some Seleniocyanates. IN 1855 Buckton discovered and described the double sulphocyanates of platinum.* Of these, the potassium salt is perhaps the one best known, partly because of its beauty, and partly because of the ease with which it may be prepared. Recently, my attention having been called to this compound, it occurred to me that it might be interesting to prepare the corresponding seleniocyanate. Accordingly I assigned the task to Mr. W. L. Dudley, a student in the University of Cincinnati, who had little difficulty in attaining to success. When an alcoholic solution of potassium seleniocyanate is added to a similar solution of platinic chloride, a heavy reddish brown precipitate is immediately formed. This, upon boiling, becomes darker in color, and apparently in part dissolves. The filtered liquid deposits crystals of the new salt, mixed with a reddish sediment of selenium; and these, although they are slightly unstable, may be purified by recrystallization from alcohol. The crystals are usually very small; mere scales in fact; although on one occasion they separated out as regular six-sided tables, several millimeters in diameter. By reflected light they are nearly black; but by transmitted light, deep garnet red. Specific gravity, 3377 at 10°2, 3.378 at 12°5. The weighings were made in benzol. Determinations of platinum and potassium came out as follows: There is, therefore, no reasonable doubt that the new salt is represented by the formula K,Pt(CSeN),, and that it is strictly analogous to Buckton's sulphocyanate. An attempt to prepare gold salts resembling the sulphocyanates described by Clevet was only partially successful. When alcoholic solutions of potassium seleniocyanate and neutral gold chloride are mixed, a red precipitate falls, which consists in large part of free selenium. The pale orange-yellow filtrate from this precipitate yields by spontaneous evaporation a crystalline crust, which under the microscope is seen to be made up chiefly of minute, deep red prisms. These crystals are so * Chem. Soc. Quart. Journ., vii, 22. + Jahresbericht, 1865, p. 295. very unstable that we could obtain but a very small quantity of them, and in a somewhat impure condition. They yielded 48.31 per cent of gold, whereas the salt KAu(CSeN),, analogous to the potassio-aurous sulphocyanate of Cleve, should contain but 43-94. As the new salt was prepared by a method precisely similar to that which gave Cleve his sulphocyanate, there can be little doubt that we had to deal with the corresponding seleniocyanate, mixed with free gold. If we had been able to command larger quantities of material, we might have been able to prepare the compound in a state more nearly approaching purity. No seleniocyanate resembling Roesler's potassium chromosulphocyanate, K,Cr(CSN), 2, 8H,O,* could be obtained. When aqueous solutions of chrome alum and potassium seleniocyanate are mixed, selenium is precipitated, and no trace of any double salt seems to be formed. VIII. On the Electrolytic estimation of Mercury. In 1865, Wolcott Gibbs published his well known method for the electrolytic estimation of copper. More recently, Merrick has shown that a modification of the same process is applicable to nickel and to zinc.‡ Having occasion recently to make a number of copper determinations by this method, it naturally occurred to me that it might be extended still farther, especially to the cases of cadmium and mercury. With cadmium I was disappointed; but with mercury, successful. Cadmium may indeed be completely precipitated by electrolysis from an ammoniacal solution, but it comes down in a spongy, porous form, enclosing various impurities which cannot be readily washed out. Accordingly the results came out several per cent too high. The mercury, however, gave results in every respect satisfactory. A solution of mercuric chloride, slightly acidulated with sulphuric acid, was placed in a platinum dish connected with the zinc pole of a six-cell Bunsen's bichromate battery. The wire from the carbon pole terminated in a thin slip of platinum foil, which dipped into the solution. At first, mercurous chloride was precipitated, but this by degrees was reduced to the metallic state, so that after an hour or so there remained in the dish a clean mass of mercury, covered by a solution in which ammonia failed to produce the slightest turbidity. When I poured off this clear acid solution the mercury became covered with a thin tarnished film, which at first annoyed me considerably. I soon found, however, that this annoyance could be avoided very easily. I simply drew off the solution from This Journ. xxxix, 64. American Chemist, October, 1871; Chem. News, xxiv, 100, 172. * Journ. für Prakt. Chem., cii, 316. above the mercury by means of a pipette, and replaced it with clean water; doing this several times before disconnecting the platinum dish from the battery. Then, upon decanting the very feebly acid supernatant liquid, the metal remained perfectly bright and clean. It was only necessary after this to rinse thoroughly with pure water, then with alcohol, and lastly with ether, and to dry under the receiver of an air-pump. Two determinations made with mercuric chloride gave respectively 73.76 and 73-85 per cent of mercury. Theory 73.80. There are no difficulties in the process, and no appreciable sources of error. Although I have made actual determinations of mercury only with the chloride, I have tested other salts of the metal and have found that the precipitation is similarly perfect. In one instance I employed a solution of mercury containing a heavy precipitate of basic sulphate. This precipitate was readily and completely decomposed by the electric current, so that ultimately nothing but metallic mercury remained visible in the solution. In every case, mercurous compounds appear to be thrown down first, so that their final disappearance furnishes a sharp end reaction to indicate when the operation is complete. IX. Some Specific Gravity Determinations. The following specific gravity determinations represent work done by my students and myself during the school year 18771878. Those portions of the work which were entrusted to students were carried out under my immediate supervision, and every precaution was taken to ensure a fair degree of accuracy. The salts were all weighed in benzol, and the figures refer to water at its temperature of maximum density as unity. To Mr. W. H. Creighton and Mr. E. F. Wittmann I assigned mercuric cyanide and some of its double compounds. For the cyanide itself, HgCy,, we found a sp. gr. of 40262 at 12°, Creighton; 4-0026 at 22-2, Wittmann; and 4.0036, 14°2, F. W. Clarke.* For the oxy-cyanide, HgCy, HgO, Mr. Creighton found 4:437 at 19°2, and I myself, in two determinations, 4-428 and 4·419 at 23°.2. For the double salt HgCy, HgCl,, Mr. Wittmann obtained the values 4:531, 21°-7, and 4.514, 26°. For the double cyanide of mercury and potassium we have, from experiments made by Mr. Creighton, 2-4470, 21°-2; 2.4620, 21°5; and 2·4551, 24°. This salt is the well known. 2KCy. HgCy.. Mercuric bromide, prepared by Mr. Miles Beamer, gave 5.7461, 18°, and 5·7298, 16°.† * Bödeker, Jahresbericht 1860, gives for HgCy, the value 3.77, 13°. |