Some of this water was placed in a bottle, and tests were applied, as with the Ca and Mg solutions, diluting whatever amount was taken, up to 100 c.c., with distilled water before applying the soap test.

EXPERIMENTS ON SAMPLE A (after standing 3 months).

Which figure should be taken as representing the total hardness of this water? The soap solution was entirely used up in making these tests, or they would have been carried farther. A small amount of another soap solution happened to be available, which like the first had deposited a slight sediment, which as before was filtered off, affording a clear, bright solution. With this tests were also made on the Ca solution. The results were:

100 c.c. of distilled water took 1.2 and 1.3 c.c. soap solution. The value of the soap solution was a little different from the first one, but the results show essentially the same as before.

The results by the soap test are apparently altogether unreliable, for very hard waters especially, and even with waters comparatively soft the results for permanent hardness are still more untrustworthy. This last point is alluded to by Mr. Wanklyn (Water Analysis, 7th Ed., p. 98) where be expresses disapproval of any attempt to determine the different kinds of hardness (temporary or permanent) by means of the soap test.

It would seem advisable to abandon the use of the soap test entirely, for although

all chemists are well aware of the roughly approximate character of the results, their clients usually are not, and attribute as much importance to the figure for “hardness by soap test as to any other figure on the report of a chemist whom they may employ.

APPARATUS FOR BUTTER ANALYSIS.

BY H. DROOP RICHMOND.

In a paper, a translation of which appeared in the ANALYST for November and December, 1888, and January and February, 1889, Wollny recommends the use of a 50 per cent. soda solution for saponifying the butter fat in Reichert's process, and the following apparatus has been found useful for storing this solution.

It consists of a bottle holding the solution fitted with an indiarubber cork, through which two tubes pass, one of them, open at both ends, carrying the blowing-ball A, and the other being bent over and terminating in the pipette B. The tube passes down inside the pipette, and terminates at a mark, to which, if the pipette is filled, just over 1 c.c. of soda solution would be discharged. In the upper part of the pipette at tube D, on which is the tap C, is joined at right angles, to the end of which a soda-lime tube is attached (not shown in figure). At the bottom of the pipette a cap (either of glass as in figure, or preferably of indiarubber) is fitted.

The apparatus is used in the following manner: On compressing the ball A, the soda solution is forced up the other tube into the pipette, the air escaping by the tube D (the tap being open). When the pipette is full to the mark, or a little over, the pressure on the ball is released, and the excess of soda solution, together with some air, is sucked back. The tap is now turned off and the cap removed, and the flask or other receptacle placed under the bottom of the pipette; on the tap being turned, the caustic soda solution runs out, and the cap being replaced, the apparatus is ready for use again.

Fig. 1,

Fig. 2.

As long as the level of the caustic soda in the pipette is not allowed to rise as high as the tube D, no particular care need be taken to fill to any place, the excess being invariably sucked back. All the air that enters the apparatus must pass through the soda-lime tube attached to D, and is thus free from carbonic acid. The quantity 1 c.c. has been chosen, as it is sufficient for 2 grs. of butter-fat, the quantity taken for the Reichert process; should 5 grs. be used, 2 c.c. may, of course, be taken.

Another piece of apparatus which will be found useful is this (Fig. 2), which is used when distilling. It consists of a tube, A 1 c.m. in diameter, and about 4 c.m. long, connected by a bulb, B 2 c.m. in diameter, with a tube C about 6 c.m. in diameter and 15 c.m. long, at angle of about 120°; this at the end is also bent at 120°. On this tube C is fitted a small condenser D, the water running in at E and out at E'. The distilling flask is attached by a cork to A, and the end of C is inserted into a condenser, the water being allowed to run freely through D until the fatty acids are melted. When this is the case, the water is removed from D and the distillation proceeded with. The bulb B prevents any spurting that may take place, and the wide tube A allows the condensed liquid to run back into the flask without trouble. If care be taken to remove the water from D, the results are not sensibly different from those obtained with the Wollny apparatus. (Conclusion of the Society's Proceedings.)

ESTIMATION OF SOLUBLE AND INSOLUBLE FATTY ACIDS IN BUTTER BY WILLIAM JOHNSTONE, PH.D., F.I.C., F.C.S.

Of the butter to be examined, 2.5 grammes of the clarified fat is saponified with a known quantity of alcoholic potash in a closed flask with the aid of heat and repeated shakings.

When saponification is complete, the soap solution is transferred to a suitable vessel and titrated with standard acid, and thereby the amount of KHO ascertained which is required to saponify the 2.5 grammes of the following three samples of butter.

The alcohol is now boiled off, excess of acid added so as to decompose the soap, and then gently heated until the fatty acids are melted.

The insoluble fatty acids are now filtered off through an unweighed filter in the usual manner, and then thoroughly washed with boiling water until the washings are neutral, when the filter containing them is set aside in a moderately warm place and allowed to air dry. When the filter is sufficiently dry, it is taken and folded up and transferred to a Soxhlet extraction tube and thoroughly extracted with dry ether, the acids being received in an accurately weighed flask.

When the extraction is complete, the ether is evaporated off, and the flask and contents put in the water oven, and when dry allowed to cool and finally weighed.

When the percentage of insoluble fatty acids have been ascertained, add sufficient standard alkali to the flask, heat gently when saponification rapidly takes place, then wash the soap solution thoroughly into a basin and titrate again with standard acid and estimate the amount of alkali now consumed by the insoluble fatty acids.

The percentage of soluble fatty acid (as butyric) by Reichert-Wollny's process is:

The amount of soluble acids found in the above three samples of butter calculated as butyric acid:

1.78

4.44

2.18

Working in the manner above described, we get an accurate estimation of the amount of the total soluble fatty acids contained in a sample of fat, and it is, therefore, an infinitely superior process to Reichert's, for by the latter only a proportion of the soluble acids are obtained, and that merely of the volatile acids; besides, the Reichert-Wollny method only furnishes comparative results when rigidly performed in the same manner, a process which ought not to be tolerated when there is another method which gives with absolute accuracy the total amount of soluble fatty acids. Reichert's process certainly allows the operator to make a very rough estimation of the volatile acids, but it does not give him any insight as to the composition of the volatile acids, viz., as to the proportion of Butyric, Propionic, Acetic, and Formic acids, all of which are sometimes present in the distillate, and last, but not least, the proportion of nonvolatile soluble fatty acids.

The following equations are extremely interesting, and may very easily be brought about with ordinary glycerine.

What must it be with nascent glycerine in presence of excess of KHO?

RECENT RESEARCHES AND IMPROVEMENTS IN ANALYTICAL

PROCESSES.

TEST TO DISTINGUISH RESORCIN FROM CARBOLIC OR SALICYLIC ACID. H. BORDE Nederl. Tydschr. v. Pharmacie, etc., May, 1889.-If a few drops of a solution of sodic hypochlorite are added to a watery or alcoholic solution of resorcin, a violet colour, rapidly changing to yellow, is produced. On warming or adding excess of the reagent, the liquid gets dark brown. One part of resorcin dissolved in 10,000 parts of water will still show the reaction. Carbolic, salicylic, benzoic and other allied acids do not show the reaction, and at the most turn only slightly yellow on warming. Pyrocatechin, when treated with the reagent, turns a vanishing green; hydroquinon soon gets yellow and red. Another test is to first add a little liquor ammonia and then a few drops of the hypochlorite, when the liquid will give a reddish violet colour, which changes green on boiling. The colouring matter is not taken up by benzol. The reaction is not shared by salicylic acid, benzoic acid or antifebrin, but carbolic acid thus treated turns a greenish blue, which is partly decomposed by benzol. The colours are changed to red by dilute sulphuric acid. L. DE K.

COMPOSITION OF MACASSA OIL. L. VAN ITALLIE.-Nederl. Tydschr. v. Pharmacie, etc., May, 1889.-The seeds of Cussambrium spinosum contain about 36 per cent. of a soft fat easily soluble in ether, petroleum spirit, and benzol, but slightly so in alcohol. On warming, the bulk of the fat melts at 22° C. The author obtained on analysis the following results: Iodine number 53. Insoluble fatty acids, 91-4 per cent. Specific gravity of these acids, ·922 at 15° C.; their melting point, 55° C. One gram. of the fat contained free fatty acid corresponding with 0166 gram. caustic potash. The saponification equivalent was found to be 230. The glycerin was estimated by Hehner's bichromate method, and amounted to 6.3 per cent. The volatile acids were found to be acetic and butyric acids. The presence of oleic acid was proved by dissolving out the lead soap with ether and weighing the lead oleate. Some of the oil was saponified and the acids thrown up with hydrochloric acid. They were then treated with alcohol of 70 per cent. to remove bulk of the oleic acid, and then treated with alcohol of 90 per cent. The insoluble portion proved to be arachidic acid, whilst the neutralised solution gave with magnesium acetate a precipitate which, from its percentage of magnesia and other properties, was found to be a salt of lauric acid. The fat therefore consists of the glycerides of lauric, oleic, arachidic, acetic, and butyric acids. A sample of the oil which had been kept for twenty years had not in the least deteriorated.

L. DE K.

ESTIMATION OF ALKALOIDS IN COCA LEAVES. PROF. VAN DER MARCK. Nederl. Tydschr. v. Pharmacie, etc., April, 1889.-The author first tried Squibb's method. Fifty grammes of the leaves were macerated with 40 grammes of 95 per cent. alcohol and 1.6 gram. of hydrochloric acid. After percolating and distilling off the spirit, the residue was taken up with 30 c.c. of ether, which was then shaken with 30 c.c. of water and 1 c.c. of 10 per cent. acid. The watery fluid was drawn off, and the ether once more