Dr. Van Hamel Roos calls attention to the following novel scheme for the sophistication of coffee berries:2

The microscopical examination of a sample, rendered suspicious by its dark color, showed the structure of genuine coffee, but the fat globules, which are always abundant in pure coffee, were almost entirely absent. The ether extract from pure coffee is 13 to 14 per cent. In this sample it was less than 1 per cent. It is evident that the roasted coffee had been treated for the manufacture of coffee extract, after which the grains were roasted a second time with the addition of a little sugar to cover the berries with a deceptive glazing. The dark color of the beans was due to the second roasting.

1 Address given by the customs authorities is 159 Front street, New York City 2 Revue Intern, des Falsifications, 4,10, 166,. May 15,1891.

Owing to lack of time no chemical analyses of artificial coffees were made in connection with this report. A large number of analyses have been published in the journals, from which those given in the following table have been taken:

Imitation Coffee Beans

1 From a tabulation by C. Kornauth, Rev. Internat. Scient. et Pop. des. Falsifications des Derivees Aliment., 3,195-196. 2 Zeit. angew. chem., 1889, 310-311; Chem. centralbl., 1889,154. 3 Original article, Zeit. f. Nahrungsmitteluntersuch. u. Hygiene, 3,221-222; Chem. Centralbl., 1890,135.

Kornauth (loc. tit.) states that he has analyzed artificial coffee beans which were very similar to the genuine, having a specific gravity of 1.26. Specific-gravity determinations of the samples examined in connection with this report gave the following numbers: Serial No. 8491, 1.195; No. 8933, 1.073; No.8859, 1.198; No. 8883, 1.111; No. 8951,1.119; No. 8952, 1.183; No. 8953, 1.194; No. 8955, 1.211; No. 895C, 1.174 (light-colored grains) and 1.131 (dark-colored grains), and No. 8957, 1.118. In making these determinations a solution of sodium chloride was employed. Twenty imitation coffee beans were immersed in this solution, the density of which was then gradually changed until ten beans floated near the surface and ten at the bottom of the solution. The specific gravity of the modified solution was then taken and recorded as the specific gravity of the imitation coffee.

These specific gravities are only an approximation, owing to variations in the density of artificial coffees even from the same sample.

Kornauth, in the article cited, states that he supposes the imitation coffee to be composed of grape sugar and dextrin, which are mixed with fat and sugar and the mass pressed in forms, roasted, and glazed. He states that imitation coffee beans sink in 40 per cent (by volume) alcohol, while the genuine beans float.

Stutzer and Reitnair suppose that the imitation coffee beans are composed of roasted sugar; Konig, of wheat bran; Frike, of sugar and lupine flour; Hanausek, of wheat bran and the refuse from white peppers; Pavlicek, of sugared bran, and Portele supposes them to be composed of sugar, cereals, and legumes. Fricke considers lupine seeds harmful. These opinions are from Kornauth's article cited above. The observations of Portele agree more closely with the results of the inves tigations of this laboratory.

A recent journal1 reports the following as the composition of an imitation coffee seized by the Government of Roumania: Coffee grounds (spent coffee), chicory, and pease. The mixture had been molded in a special machine in imitation of coffee beans.

A factory for the manufacture of imitation coffee was recently seized at Lille, France, by the French Government.2 The capacity of this fac-tory was 40 to 50 kilos of imitation coffee per day. The composition of the product, as shown by the testimony at the trial of the manufacturers, was as follows: chicory, 15 kilos; flour, 35 kilos, and sulphate of iron, 500 grams.

L. Jammes3 examined a sample of imitation coffee and found it to be composed of acorns and cereals.

Detection Of Imitation Coffees

Roasted imitation coffees may usually be very easily detected. As a rule, genuine roasted coffee will float on water, and the artificial product, roasted, will sink; there are, however, exceptions to this. Coffee that has been "overroasted" will sometimes sink in water. The artificial coffees examined in connection with this report, with one exception, sank in water. These remarks apply to whole coffee. Kornauth4 states that imitation coffee beans sink in 40 per cent (by volume) alcohol. Certain coffees will also sink in alcohol of this density; hence this alone is not a sure test.

In examining roasted coffee for the imitation product a portion of the sample should be thrown on 40 per cent (by volume) alcohol, and those beans which sink should be subjected to a further examination. The genuine coffee bean always has a. portion of the line membrane with which it was originally invested still adhering in the cleft. This test alone will distinguish the genuine from the imitation coffee beans, On the examination of a section of an imitation coffee bean it may he seen that the structure is uniform, while that of the genuine bean is not. The imitation coffee generally contains starch, a substance which is never present in the genuine beans. If starch is absent the sample should be examined microscopically for chicory or similar roots.

In the examination of aground sample tests should be made for starch and chicory. A portion should be thrown on cold water. Chicory, If present, will quickly color the water, while cereals will sink, often imparting little, if any, color to the water. It should be noted that coffee, deprived of its oil, will sink in Water and that cereals or chicory treated with oil will float. The cold-water test should always be made, since valuable indications as to the purity of the samples may usually be obtained. A low percentage of ash indicates an adulterated sample. The microscope must be employed for the final identifications of the constituents of the sample.

1 Revue Intern, dee Pal 4 e Annee, N. 11, 1SS.

2 Revue Intern, del Fals., 4e Annee, No. 11, 185.

3 Chem. Centralbl., 1891, 1, 935.

4 Rev. Internat. Scien, et Pop. des Falsifications, 3, 195.