it as a means of determining the temperatures by observation of the electromotive forces produced. The study, as far as it concerned the separation of arsenic and antimony, is not completed, hence the results are reserved for another publication. Sufficient data were obtained to show that there was a possibility of separating arsenic and antimony by the difference in decomposition points of their diluted hydrides, and the investigation will be continued in this laboratory. Concerning the arsenic hydride, it will be sufficient to say that an amount of diluted hydride corresponding to 0.04 mg. of arsenious oxide is completely decomposed at 340° if the length of capillary heated is 3.5 cm. ; at 410° if the length is 2 cm. ; and at about 450° if the length is 1 cm. Conversely, no arsenic is deposited from this amount at 330° with a heating length of 1 cm. ; none at about 300° if the length is 2 cm. ; and none at about 250° if the length is 3.5 cm. We were unable to get as definite results on the decomposition of antimony hydride, since the difficulty in deposition of the antimony mirror on the glass tubing used, which was afterwards solved as explained later, prevented the results from being uniform. We think, however, that we are safe in saying that amounts of hydride from 0.1 mg. antimonious oxide and under are entirely decomposed at a temperature of 300° and a heating length of 2 cm. Having shown that practically all the antimony may be evolved as hydride from the small amounts to be used in our work, and that the hydride, diluted with hydrogen, can be entirely decomposed by heating the capillary through which the gases pass at a temperature which is easily controlled, we then proceeded to develop a method for estimating minimal amounts of antimony. The Apparatus. The apparatus 10 is essentially the same as that proposed by Sanger for the estimation of small amounts of arsenic. Slight modifications of this have been introduced by subsequent workers, many of whom have overlooked the original article. As shown in the figure, the parts comprise a constant hydrogen generator, two reduction flasks, and two heating tubes, so that duplicate determinations may be carried out at 10 This apparatus, which is shown in Figure 2, is essentially the same as that described, but not illustrated, in the paper of Sanger (1891) above referred to. An apparatus of practically the same principle is shown on page 226 of Volume II of the Final Report of the Royal Commission on Arsenical Poisoning (Appendix 22, Report to the Commission by McGowan and Finlow on the methods employed in testing for arsenic); London, Eyre and Spottiswoode, 1903. the same time. The generator which we have found most convenient is one proposed by Richards, and is similar to the forms described by him some years ago.11 Any form of constant generator will answer which provides for withdrawal of the spent acid without disconnection. The zinc in the generator is conveniently sensitized, according to the suggestion of Gooch,12 by brief treatment with a solution of cupric sulphate and subsequent washing. The acid is sulphuric, at a dilution of one to eight. The hydrogen from our generator shows no arsenic or antimony when run for hours at a time. As the hydrogen with which the antimony hydride is heated must contain no hydrogen sulphide, as hereafter shown, the hydrogen from the generator is passed through a ten per cent solution of cupric sulphate contained in an Allihn or other suitable washing bottle. From this, the hydrogen, which needs no further purification, passes to a Y-tube with two glass stopcocks. To these stopcocks are attached the reduction flasks, which are wide-mouth bottles of 60 to 75 c.c. capacity. These are fitted with a pure rubber stopper with three holes. Through one hole passes, to the bottom of the bottle, a right-angle tube connected with the stopcock; through the second a tube passing nearly to the bottom of the bottle and extending 3 to 5 cm. above the stopper. The upper end of this tube is open, the lower somewhat constricted. In the tube is placed a small funnel, blown from narrow tubing, through which the solution to be reduced is added. The third hole in the stopper carries the right-angle delivery tube for the hydrogen which passes just below the stopper. On its upper end is a rubber stopper over which is placed a 15 cm. (total length) tube filled with calcic chloride in fused sticks. We have found it convenient to half fill the tube with calcic chloride, and then to introduce a second, smaller tube filled with the chloride. By this arrangement the rear portion of the chloride, which soon becomes moist, may be frequently and conveniently renewed without disturbing the rest. To the calcic chloride tube is connected the hard glass reduction tube drawn out to a straight capillary and ending in a capillary point. This tube is supported throughout its length by three adjustable brass hooks (No. 6 gauge, 4.1 mm.). The middle hook, which has a shank 10 cm. long, is fastened to a stand by the ordinary double clamp. On the top of the shank and about two thirds the distance from the clamped end are soldered two ordinary screw connectors at right angles to the shank. The other hooks, which are somewhat longer than the first, are bent at right angles in the plane of the table and their shanks 11 Amer. Chem. Jour., 20, 189 (1898). pass in opposite directions through the connectors, in which they can be clamped by the screws. The end hooks are thus capable of being raised or lowered or of being moved laterally, so that the combination of the three hooks will support any tube and capillary. We found it best to protect the capillary from direct contact with the flame, and at the same time to secure a more uniform heating by enclosing it in a brass tube or collar which is slipped over the capillary and rests on the two anterior hooks of the support. This collar is 5 cm. long, with an outer diameter of 6 mm. and an inner of 4 mm. It is heated towards the anterior end by the tip of a flame 5 to 6 mm. high from a burner with a good air supply and protected by a conical chimney. We assume that under these conditions the capillary is heated, through a space of 3 cm., at about 500°. The Preparation of Standard Mirrors. A standard solution of antimony was made by dissolving 2.3068 gr. of recrystallized tartar emetic in water and making up to one liter. This solution (I) contains 1 mg. of antimonious oxide in each cubic centimeter. Of solution I, 10 c.c. were diluted to a liter, giving a solution (II) containing 0.01 mg. per c.c. The tartrate offers the most convenient and accurate solution, and we satisfied ourselves that the presence of the tartrate ions had no effect whatever on the deposition of the mirror. The strength of solution I was also checked by analysis. The zinc used in the reduction flask was the same as that used in the generator, but in smaller pieces, averaging perhaps 1 cm. in their longest dimension. The weight used was from three to five grams. This zinc, which is obtained of the New Jersey Zinc Company of New York, contains not over 0.019 per cent of lead and not more than 0.013 per cent of iron; hence the evolution of hydrogen by its contact with dilute sulphuric acid is slow. Platinum or other sensitizing agents, either in form of foil or as a deposit on the zinc, cannot be used, as we have proved by trial, since they show a tendency to hold back antimony even greater than in the case of arsenic. 13 We had recourse, therefore, to hydrochloric acid, in a dilution of one to ten, of which we use exactly 20 c.c. for each run. The hydrochloric acid, which is obtained of Messrs. Baker and Adamson, of Easton, Pa., contains no antimony, and only about 002 mg. of arsenious oxide per liter, an 13 The discussion as to the effect of other metals on the evolution of arsenic hydride has been revived in recent years through the endeavors to increase the delicacy of the Marsh test for arsenic. This question, as applied to arsenic and antimony, will be taken up by one of us in a future paper. amount which is inappreciable in our work on antimony.14 By use of hydrochloric acid and purification of the hydrogen from the generator, there is no necessity for purification of the hydrogen from the reduction flask, provided sulphur compounds reducible to hydrogen sulphide are absent. Whenever the hydrogen contains hydrogen sulphide, the deposit in the heated tube is more or less reddish-yellow in color, due presumably to the presence of antimonious sulphide. The fused sticks of calcic chloride used for drying (Merck) dissolved clear in water, and the solution showed an alkalinity of not over 0.3 per cent. We have seen no reason to believe that hydride of antimony is held back by this preparation, nor has there been any loss of antimony observed from an unduly moist condition of the chloride. We have, however, kept the chloride as active as possible by refilling the rear half of the tube with fresh chloride after every half dozen runs, as explained, or when the accumulation of moisture becomes noticeable. When not in use, the tubes are kept stoppered. The selection of hard glass tubing is a matter of the highest importance for the success of the method. We used at first a German glass, source not known to us, which had been used in arsenic work without apparent disadvantage. This glass had a slight brown color after long ignition. Antimony mirrors deposited on this, under the conditions about to be described, were often white and not clearly defined. We next tried a sample of American glass, but this gave the brown color on ignition still more, and the deposits of antimony were entirely white. Next a Jena glass, which on ignition gave the wellknown opaque appearance, and on which the mirrors were also entirely white. Finally we resorted to a glass of Kavalier, which did not give any color or opacity on long heating. The deposits of antimony on the capillaries drawn from this glass were satisfactory, and it was used in the preparation of the standard mirrors.15 14 This would mean not over 0.00004 mg. of arsenic in the amount of hydrochloric acid used, which is beyond the limit of the delicacy of the process as applied to arsenic. 15 A cursory qualitative examination of the different samples of tubing showed no marked points of difference except that the fourth tubing contained no barium, while the others did. The amount of barium in the others was proportional to the degree of change in appearance of the mirror produced on the samples. Lack of time prevents an investigation on this point, and an opinion as to the influence of the barium, if any, would be mere conjecture; as, for example, whether the barium oxide could act catalytically in causing an oxidation of the antimony, since it seems probable that the white deposit is due to an oxide of antimony. In the absence of any definite knowledge on the matter one can only determine the availability of a sample of glass by actual trial. |