dyed with 12 varieties of indigo, using dilute soda and zinc powder for the reduction in order to avoid the removal of impurities which takes place in the lime-copperas vat. In the case of one of the samples (fine Java indigo), the dyed fabric behaved with acidified alcohol as though it had received a second dyeing with logwood. This anomalous result appears to have been due to the formation of an unstable yellow product of the reduction of indigo. It could not be obtained again with the same sample.

For the detection of the more usual blue dyes the following methods may be trusted :-The stuff is warmed with an acidified 10 per cent. solution of stannous chloride. Prussian-blue remains unchanged. Indigo (vat-blue), indigo-carmine, cotton-blue (triphenylrosaniline trisulphonate) are completely removed from the fibres and yield pale-yellow solutions. Logwood is also removed but gives a rose-red solution. On adding a large excess of hydrogen peroxide to these solutions, the rose-red of logwood is destroyed, cotton-blue gives a blue solution, whilst indigo is not regenerated.

Glacial acetic acid, or concentrated formic acid, dissolves indigo from a fabric. In presence of logwood, either acid in the cold acquires a rose-red colour, which on heating passes into yellowishred, and is soon obscured by the dissolving indigo. Prussian-blue is unaffected, and indigo-carmine is not dissolved if more than 48 hours have elapsed since the dyeing. On mixing the acetic acid solution with ether and then adding water until the ether separates, the indigo is removed from the aqueous layer, which then in the presence of logwood shows a feeble reddish-yellow tint. If now a few drops of concentrated hydrochloric acid is added, the smallest trace of logwood is revealed by the production of a rich red colour in the aqueous layer. Cotton-blue obscures this reaction. In this case, the aqueous layer must be separated and shaken with chloroform, or better amyl alcohol, which abstracts the aniline-blue. Another method of examining the acetic acid solution is to mix 1 c.c. of it with 5 c.c. of chloroform, then add 2 c.c. of water and crystallised sodium carbonate until alkaline. The chloroform layer is then separated and shaken with strong borax solution, which frees it from logwood. It is then mixed with ether and dilute acetic acid, when the aniline blue passes into the acid and the indigo remains in the mixture of ether and chloroform. The alkaline solution contains the indigo-carmine and logwood; this is acidified with acetic acid and shaken with amyl alcohol, which removes the logwood with reddish colour, further intensified by hydrochloric acid.

If a fabric dyed with the above colours is boiled with borax solution the logwood-blue and indigo-carmine are dissolved, whilst Prussian-blue is decomposed. Aniline-blue and vat-blue remain on the fibre, and may be distinguished by warming with ferric chloride. The filtered borax solution, which in presence of indigo-carmine is blue, is tested for logwood by stannous chloride.

As a special test for logwood, the stuff may be treated directly with hydrochloric acid, or boiled with a 10 per cent. solution of alum, or of glucinum chloride to which an excess of ammonium carbonate is subsequently added. Ammonium molybdate gives at once an intense

University of

blue-violet. The absorption spectra of many of these solutions are figured.

A fabric dyed with indigo alone should stand the following tests: -Alcohol should dissolve no colour even on gentle warming; cold saturated oxalic acid and borax solutions, 10 per cent. alam solution, and 33 per cent. ammonium molybdate must remove no colour even on boiling. Stannous or ferric chloride destroys the colour on warming. Glacial acetic acid dissolves all the colour on repeated boiling, and after mixing the solution with ether and water the aqueous layer is colourless and is not coloured by strong hydrochloric acid. The fabric boiled with hydrochloric acid evolves no hydrogen sulphide, and the acid extract warmed with a large excess of alkali and some chloroform evolves no odour of isonitriles. M. J. S.

Detection of Artificial Colouring Matters in Butter, &c. By E. W. MARTIN (Analyst, 12, 70).--The following method serves for the detection of annatto, turmeric, carotin, and the aniline- and naphthol-yellows. Two parts of carbon bisulphide are gradually added to 15 parts of methyl alcohol, with gentle shaking; to 25 c.c. of the mixture, in a convenient tube, 5 grams of the butter-fat or oil is added, and the tube is shaken. In a few minutes the bisulphide falls to the bottom, carrying with it the fat, whilst the foreign colouring matter remains in the alcohol. The natural colouring matters of fats, &c., do not colour the methyl alcohol. M. J. S.

Pettenkofer's Reaction. By F. MYLIUS (Zeit. physiol. Chem., 11, 492-496). The cause of the blood-red colour produced on adding cane-sugar and sulphuric acid to cholic acid (Pettenkofer's reaction) is unknown. The same colour is produced by the bile acids, by choleïc acid, and by the distillation products of cholic acid. On the other hand, dehydrocholic acid and bilianic acid do not give it; in these latter acids hydroxyl is absent, and it is not improbable that the reaction depends on the presence of the hydroxyl-group. The reaction may be obtained by using furfuraldehyde instead of sugar. That it is not the sugar itself on which the reaction depends, but some volatile substance, was considered probable, because the reaction was obtained from the distillate of sugar with dilute sulphuric acid; acetone, mesityl oxide, and phorone, which are formed in this process, do not give the reaction, but furfuraldehyde, which is also formed, does. As a test for furfuraldehyde, it is exceedingly delicate, a drop of a mixture of water and furfuraldehyde in the proportion 20,000 to 1 gives the colour on the addition of cholic acid and sulphuric acid; in other words, the fortieth part of a milligram can be thus detected. The colour can be extracted by ether, which is coloured blue by it.

Other substances, however, besides cholic acid give the same reaction with furfuraldehyde, namely, isopropyl alcohol (to a slight extent), isobutyl alcohol (very intensely), allyl alcohol, trimethyl carbinol, dimethyl ethyl carbinol, amyl alcohol, oleïc acid, and petro

leum. Among the substances which do not give the colour are the following:-Ethyl alcohol, propyl alcohol, capryl alcohol, acetic acid, isobutyric acid, acraldehyde, and benzene. The substances which give the test are seen in general to be unsaturated compounds, or those which by the loss of water become unsaturated; it is considered not improbable that the reaction depends on some hydrocarbon produced from the foregoing compounds when they lose water owing to the action of sulphuric acid.

W. D. H.

Detection of Albumin in Urine. By A. LIEBERMANN (Chem. Centr., 1887, 600).—If albumin is extracted a few times with alcohol, and then washed with cold ether, it gives a deep violet coloration if heated with concentrated hydrochloric acid (sp. gr. 1.196). Positive results were obtained with various pathological products, but experiments with hæmoglobulin, chondrin, and keratin, were unsuccessful. In the case of urine, the liquid is at first boiled with a small quantity of acetic acid, and the precipitated albumin treated as above; the reaction succeeded with 5 c.c. of urine to which 0.1 per cent. of eggalbumin was added. V. H. V.

INDEX OF AUTHORS' NAMES.

ABSTRACTS. 1887.

And also to Transactions, 1887 (marked TRANS.); and to such papers.
as appeared in Abstract of Proceedings but not in Transactions
(marked PROC.).