could not find any barium in the stomach, œsophagus, and duodenum, but got as much as 158 gram. of barium sulphate from the abdominal glands, brains, and lungs, whose joint weight was 719 grammes.

(b) The soluble portion.

I must state that it is absolutely necessary the fluid shall contain no free chlorine. After making up to a definite bulk, say 500 c.c., aliquot parts are taken and submitted to analysis.

(a)

Fifty to 100 c.c. are nearly neutralised with caustic potash in a porcelain dish, a small coil of clean brass is introduced, and the whole warmed for fifteen minutes. The coil is then washed with water, alcohol, and ether, and after drying (in desiccator) introduced in a combustion tube partly filled with coarse oxide of copper, and drawn out so as to be connected with two capillary tubes. By means of an aspirator a slow current of dried air is drawn through and the tube heated as if an elementary analysis were conducted. Any mercury will condense inside the capillary tubes.

Whilst still hot a drop of water is put on the combustion tube (close to the capillary one) and is so made to break. A minute particle of iodine is now introduced, and with the application of a gentle heat, its vapour mixed with the air current, is made to act on the mercury globules, which will then cause the formation of the characteristic mercuric iodide. As regards this very reliable process, I want to observe that a long immersion of the brass is not advisable because it gets too thickly coated with organic matter.

The quantitative estimation of the mercury I conduct as follows:-Through the hot acid solution, I pass hydrogen sulphide, and collect the precipitate on an asbestos filter. After washing with strong hot hydrochloric acid, the asbestos is rinsed into a beaker and digested with a few drops of nitric acid, which will soon dissolve out the mercury. After filtering through the same asbestos the mercury is precipitated as mercurous chloride by the addition of excess of phosphorous acid, which is allowed to act for twenty-, four hours.

(b) Antimony.

Fifty c.c. of the liquid are put into a clean platinum dish, the excess of acid neutralised with ammonia, and a piece of pure zinc is introduced. After acting for six hours, the liquid is poured off and the antimony will have blackened the platinum, or, if much was present, deposited itself as a spongy mass. The liquid may contain traces of it, as antimony is somewhat soluble in concentrated solution of zinc chloride. Its quantitative determination is conducted as follows:-Though the hot fluid I pass hydrogen sulphide until the liquid has got cold. After standing for three days, the excess of hydrogen sulphide is removed by a rapid current of carbonic acid, a little ether being, if necessary, added to prevent frothing. The precipitate is filtered off and washed with a weak solution of ammonium acetate. It is then rinsed from the filter with solution of caustic soda, which is then saturated with hydrogen sulphide, which will cause the antimony to dissolve. After filtering off from any insoluble matter, the liquid is slightly acidified with hydrochloric acid and heated to boiling. After standing for a day, the precipitate is filtered off, washed, and finally oxidised with nitric acid. To completely get rid of

organic matter, the mass is made alkaline with caustic soda, dried and fused with a mixture of nitrate and carbonate of soda. The mass is now exhausted with proof-spirit, and the insoluble sodium antimoniate washed with proof-spirit to which a few drops of sodium carbonate have been added. It is then digested in a mixture of hydrochloric and tartaric acids, the liquid nearly neutralised with ammonia, and once more precipitated with hydrogen sulphide. The precipitate, which should now possess the characteristic colour of antimony sulphide, is not weighed as such, but I prefer to make it into oxide by treatment with fuming nitric acid, taking all precautions as advised by Bunsen.

To make sure of the absence of tin, I place the uncovered crucible in the reduction flame of a powerful burner, when the antimony will gradually, but completely volatilise. (To be continued.)

METHODS FOR DETECTING OLEO IN BUTTER AND COTTON SEED OIL IN LARD.

BY THOMAS TAYLOR, M.D., MICROSCOPIST, UNITED STATES DEPARTMENT OF

AGRICULTURE.

Abstract of paper read before the Chemical Society of Washington, D.C., March 15th,

1890.

TO DETECT OLEO IN BUTTER.

[I believe that this method will prove useful for all practical purposes under the oleomargarine laws of the United States.]

Dissolve in 20 c.c. of petroleum benzine 140 grains of a mixture of oleo and butter. Heat slightly to secure a perfect solution of the fats. Caseine and animal tissues may be removed by filtering the liquid while it is warm. Fill a test-tube with the filtered solution and place it in ice-water. In from five to twenty minutes the oleo fat will separate from the butter fat, and falls to the bottom of the tube, being sparingly soluble in cold benzine, while the butter fat remains in solution in cold benzine. Separate the oleo fat from the liquid butter by filtration. The fat recovered may be solidified by mechanical pressure, placing it between several layers of filtering paper to absorb the remaining benzine, after which the sheet of solid oleo may be removed from the paper with a palette-knife. The butter may be recovered by evaporating the benzine by means of heat.

TO DETECT COTTON SEED OIL IN LARD.

Dissolve in 20 c.c. of petroleum benzine 140 grains of a mixture of lard and cotton seed oil. Heat slightly to secure a perfect solution of the lard. Remove animal tissues by filtering as above. Fill a test-tube with the filtered solution and place it in ice-water. The fat will be precipitated, and falls to the bottom of the tube by reason of its comparative insolubility in cold benzine, while the cotton seed oil remains in solution. Separate the lard from the cotton seed oil by the use of filtering paper, and subject the recovered fat to pressure as in the case of butter and oleo, by which means the remaining benzine is absorbed. The solidified fat may be removed from the paper with a palette-knife. The benzine is separated from the cotton seed oil by means of heat.

123

KCN in grams.

ELECTROLYTIC SEPARATIONS.

BY EDGAR F. SMITH AND LEE K. FRANKEL.

Continued from page 78.)

MERCURY FROM NICKEL.

With these two metals the current was allowed to act for sixteen hours. The

results are as follows:

MERCURY FROM COBALT.

Our experience with these two metals was so unexpected that we append the poor as well as the good results which were finally obtained :

Total dilution.

KCN in grams.

Current in c.c.
O-H gas per
minute.

The current acted for sixteen hours upon reducing the quantity of cobalt, and operating with the conditions, in other respects, the same as before (except in 11 and

13), we obtained :

:

0.2425

Inspection of these figures discloses the fact that nothing approaching a separation of the two metals appears probable until in experiment (11), where not only the quantity of cobalt is reduced, but also that of the potassium cyanide. The result is then surprisingly close (+09 per cent.). This would seem to be due rather to the reduction of the quantity of cyanide, inasmuch as by its increase in experiment (12), we again have a minus error of 401 per cent., and in experiment (13), by reducing the quantity of cyanide to 3 grams., the result is satisfactory. As the quantity of cobalt in this instance was but half of that in experiment (12), the favourable result might be attributed to this. Hence the following trials were made :

Total dilution.

Cobalt was not found in the mercury deposit, nor mercury in the cobalt solution. Evidently the quantity of cyanide present exercises a marked influence upon the separation. Returning to the separation of cadmium from cobalt, it will be observed that the amount of cyanide present there was 4 grams., while the quantity of metal was less; yet the separation proved satisfactory.

SILVER FROM COPPER.

It yet remains for us to record some experiments upon the separation of these two metals in cyanide solution. Reference to a former paper (American Chemical Journal, 11, 264, and Journal of Analytical Chemistry, 3, 254) will show that our attempts in this direction were at that time fruitless. It was after the successful separation of cadmium from copper in cyanide solution that we were impressed with the idea that the separation of silver from copper ought to occur, since silver deposits so readily, even when exposed to a very feeble current. In the communication to which we refer, the current strength recorded was 1 cc. O-H gas per minute. Since it was by carefully reducing this in other cases that we obtained good separations, we instituted a new series of experiments with silver and copper, acting upon the mixture with a much weaker current:

0.1788

41

100 p. c.

-0.05 per ct.

-0.20

We next dissolved 0·1732 gram. pure metallic silver in nitric acid. To its solu tion, after evaporation, there were added 100 per cent. copper and 4 grams. potassium cyanide. On electrolysing, with a current of 0·15 c.c. O-H gas per minute, the resulting silver weighed 0.1725 grams. The latter contained no copper, nor could silver be detected in the copper solution.

A silver ten-cent piece, weighing 1.2236 grams., was brought into solution and diluted to 100 c.c. Of this solution two portions (25 c.c. each) were electrolysed in the presence of 4 grams. potassium cyanide, with a current of 0'4 c.c. O-H gas per minute. The silver found was in

TWO CONVICTIONS FOR THE SAME SAMPLE OF MILK.-Mr. B. S. Weston, Sanitary Inspector for St. Matthew, Bethnal Green, submitted a sample of milk purchased from the Farmers' Direct Supply Association to the Public Analyst, Mr. A. W. Stokes, F.I.C. A certificate was returned stating the milk contained 10 per cent, of added water, and that 33 per cent. of the cream had been abstracted from the sample. Two summonses were taken out, one under section 6 and one under section 9 of the Adulteration Act. On February 28th, at Worship Street, Mr. Bushby heard the case. The defendant appeared by his Solicitor (Mr. Young), but the case was given against him, and a fine of £5 was imposed for the abstraction of the cream and £2 for the addition of water, costs being given in both

cases.

REVIEWS, NOTES, ETC.

JOURNAL OF THE ROYAL AGRICULTURAL SOCIETY. - We have received the first number of the new series (the third) of this journal. It is to be issued quarterly, and being well printed and containing much that is vitally important to all agriculturists, it should certainly be appreciated by all interested in the subject. We observe very interesting reports from the society's consulting naturalist, and also from the botanist; while the analyst exposes a series of frauds in feeding meals, adulterated samples of which seem to be very prevalent of late.

EXERCISES IN PRACTICAL CHEMISTRY; AN INTRODUCTION TO QUALITATIVE AND QUANTITATIVE ANALYSIS, BY DR. W. R. HODGKINSON, F.R.S.E. London: George Kenning, Great Queen Street.

THIS is a little brochure in pamphlet form of 64 pages, giving the usual reactions and tests of the various metals and acids, and a few of the gases, with tables for quantitative analysis. It possesses no special feature, but is well and concisely put. Like many books of its class, it is, in our opinion, sorely deficient in a really systematic acid course, too much being left to the student's initiative in this matter. So far as qualitative work is an introduction to quantitative, then the title is correct, but we have looked in vain for any other attempt to explain quantitative analysis.