same distance from the center of the block and equally inclined, as indicated. When the Nessler tubes are twenty-nine mm. in diameter, these holes should be about thirty-one mm. and be lined with a strip of woolen cloth to afford a snug fit for the tubes. The pivot B is a knife edge, made by grinding at its ends, a piece of a double three-inch saw file about fifty-five mm. long, and is merely hammered into the small hole cut for it. The knife edge is supported by a pair of plain two-inch angle irons (such as are kept in stock at a hardware store); in each of these, the hole in the upright limb is bored out and filed to the shape shown in the figure. These irons, D, are screwed to the wooden base, E, 170X170X 20 mm.

The motion of the block, A, is checked by two round-head screws fastened into the under side of A, and by two flat-head wood screws partly screwed into the block E. By screwing the latter up or down you may adjust the apparatus so that when one tube is empty and the other is being filled, as soon as the latter collects fifty cc. of water, it topples over and immediately brings the other tube under the point P, to collect the distillate. While the second tube is being filled, you may hold the block A, remove the first tube and empty or replace it by another tube of same size and weight; when the second tube has collected fifty cc. it will topple over and again bring the mouth of the empty tube under P. The screw which serves to fasten each angle iron in place also passes through one end of a slip of thin sheet brass 15X30 mm., the other end of which is bent up at right angles, so serving to keep the pivot from shifting to one side, but not binding upon it. This piece of brass is not shown in the figure, since it would cover the pivot end.

The Nessler tubes used are 29 mm. in diameter and 195 mm. long, and weigh 49 grams; should the tubes be heavier, it would be necessary to bore the hole for the pivot at a point a little higher up on the center line of the block A. It is easy to adjust this apparatus so that the amount required to topple the tubes will not vary as much as one-half cc. from the fifty cc.

UNIVERSITY OF VIRGINIA, February 2, 1898.

BORIC ACID DETERMINATION.'

BY THOMAS S. GLADDING.

Received February 7, 1898.

OR the determination of boric acid I find the f~"
method gives satisfactory results :

FOR

a uttle

Five cc. of sirupy eighty

One gram of the substance in which the boric ac termined is washed into flask B (150 cc. capacity ninety-five per cent. methyl alcohol. five per cent. phosphoric acid are added. Flask A is filled twothirds full of ninety-five per cent. methyl alcohol and placed in the water-bath E. Flask B is now connected with the condenser D, and flask C placed in position to receive the distillate. Heat is now applied to the water-bath E and, when the methyl alcohol is boiling, flask A is connected to the tube which passes to the bottom of flask B. A current of methyl alcohol vapor is thus continually passing through the liquid in flask B, and carries over the boric acid. Heat is applied under flask B and so regulated that the liquid remains between fifteen and twenty-five

CC.

The distillation is carried on in this way for about one-half hour, the distillate finally amounting to about 100 cc. A mixture of 40 cc. glycerine and 100 cc. water is now carefully neutralized, using phenolphthalein as an indicator, and then added to the distillate, which is then titrated with standard soda. The distillation must be continued until no more acid is obtained. Usually, it is complete in thirty minutes.

1 Read before the New York Section, February 4, 1898.

A blank should be run using all the reagents, and any acidity found must be deducted from the final results.

The following results were obtained with borax and boric acid: One gram borax gave 36.57 per cent. boric acid, the theoner cent. being 36.65 per cent. ; one gram boric acid gave acid.

ACT

ystallized boracic acid, without the addition

er acid, was found to yield all the boric acid
distillation of borax was found to yield 19.50
out of a total of 36.65 per cent. present, or
fifty per cent. of the whole amount.
eful in some analyses.

This

by means of aspiration bottle, upon flask C, 1 loss by possible leakage.

THE JOHN HARRISON LABORATORY OF CHEMISTRY,
No. 29.]

HUR MONOCHLORIDE UPON MINERALS.

BY EDGAR F. SMITH.

Received March 7, 1898.

ULPHUR monochloride has frequently been applied in

the results obtained by the employment of this reagent have invariably had great attraction and interest for the investigator. There are, however, other directions in which the same reagent may be followed with equal interest; e. g., in the action upon the natural products furnished by the mineral world. To illustrate, mention may be made of the behavior of such substances as arsenopyrite, chalcopyrite, pyrite, and marcasite with the reagent in question. Finely divided arsenopyrite and sulphur monochloride were brought together in a glass tube. After slight agitation, action set in, accompanied by the evolution of much heat, and the almost complete decomposition of the mineral. The tube was then freed from air by the introduction of carbon dioxide. It was sealed and heated to about the boiling-point of the sulphur monochloride (139°) for a period of nine hours. On cooling, beautiful olive-green colored plates or scales separated. After their removal from the tube and separation from the adherent liquid they proved to be deliquescent and

readily soluble in water. Their aqueous solution tested with potassium ferrocyanide and silver nitrate showed the presence of iron in the ferric condition, and also of chlorine. In subsequent decompositions the crystals were filtered out, and washed with petroleum ether; then they were dissolved in water and the iron content determined quantitatively by means of stannous chloride. In this manner it was proved that 32.6 per cent. of iron was present in the ferric condition in arsenopyrite. This is certainly a confirmation of the work previously carried out in this direction on arsenopyrite by Starke, Shock, and Smith.'

The petroleum ether solution from the ferric chloride crystals was distilled, the product diluted with water, and tested with hydrogen sulphide, when the arsenic was precipitated. Its quantity was not estimated.

Chalcopyrite treated in a similar manner with sulphur monochloride, was completely decomposed, with the production of ferric and cupric chlorides. Marcasite and pyrite also gave beautiful crystallizations of iron chloride. The decompositions were in both instances complete and the total iron content determined. On the addition of sulphur monochloride to marcasite and pyrite no action was observed in the cold. With chalcopyrite and arsenopyrite the evolution of heat, as already mentioned, was very great, so that the vessel containing them could not be held in the hand. This difference in behavior evidently indicates a marked difference in the union of the elements concerned. In pyrite and marcasite we deal with iron and sulphur alone and with them there is an absence of marked reactivity, whereas in arsenopyrite and chalcopyrite we have substitution products in which there is perhaps a less intimate union of the sulphur than exists in the pyrite and marcasite. It may be observed here that Brown, while working with marcasite, found that it contained its total iron in the ferrous condition, but when this same mineral was dissolved in sulphur monochloride it gave a mass of iron chloride crystals showing the presence of about forty-one per cent. of iron in the ferric state. This would mean that this reagent undoubtedly acts as an oxidant; that its power is in a large measure due to its chlorine content, 1 This Journal, 19, 948.

2 Proc. Am. Phil. Soc., 33, 1894.

2

so that we may say of sulphur monochloride, it is "chlorine in disguise." This view of its action is corroborated by other decompositions which will be presented later. The experiments upon which the preceding statements are based werecarried out by Mr. F. W. Moore.

The mineral stibnite dissolved immediately and completely in sulphur monochloride with a violent ebullition and the generation of much heat. When the solution cooled crystals separated. These were collected and heated with water, with the formation of antimony oxychloride. Much heat was also evolved when cinnabarite and sulphur monochloride were mixed. To effect complete decomposition the mixture was heated in a sealed tube. On cooling long needles were observed throughout the liquid. These proved to be mercuric chloride. Chalcocite was also completely decomposed, with the production of a crystalline powder, which on examination was found to be cupric chloride. This is additional evidence of the oxidizing character of the solvent employed, and explains the reason for the obtainment of so much ferric chloride when using marcasite, which contains its iron in the ferrous condition. Tetrahedrite, although of a complex nature, yields in the cold to the influence of the sulphur monochloride, and crystals of ferric chloride appeared in a very short time, although to effect the complete decomposition of the mineral it was necessary to heat it with the reagent to 140° C. Sphalerite was not attacked in the cold, and at 150° its decomposition was very slight, but at 250° C. complete solution occurred. On cooling, a mass of anhydrous zinc chloride separated. This was quickly collected and treated with water when it dissolved with a hissing noise. The zinc in solution equaled 67.5 per cent.

Galenite, furthermore, was not affected in the cold, but at 250° C. was changed completely to lead chloride. Finely divided molybdenite was not attacked in the cold, nor was it entirely broken up after heating to 300° C. Many of these experiments show that, where the sulphur estimation of a sulphide is not desired, this method of decomposing such minerals may be of use in analysis. This is particularly true in the case of tetrahedrite. Mr. C. S. Reeve was kind enough to conduct these experiments for me.