POLARISCOPIC AND CHEMICAL METHODS FOR DETECTING INVERT SUGAR.

Polarization has been used to some extent by chemists in this country for detecting inverted sucrose in honey. If the polarization of a honey exceeded -20° V. at 20° C. it was supposed to indicate the addition of invert sugar. The original authority for this statement. has not been found. It is hardly necessary to state that the test is altogether valueless; fifteen of the one hundred pure honeys, as has already been noted, gave a direct polarization exceeding -20° V. The non-validity of this test is also further shown by the fact that invert sugar sirups may polarize considerably below -20° V., as in the analysis of the Herzfeld sirup. While an 80 per cent solution of pure invert sugar (26 grams to 100 cc, normal weight) should polarize at about 26° V. at 20° C., the invert sirups as made by the Herzfeld and by similar processes usually polarize considerably below -20° V., owing to the destruction of some of the levulose, as previously stated.

Another test is proposed by König," which is based upon the lowering of the percentage of non-sugars by the addition of invert sugar. He says: "If the honey contains less than 1.5 per cent of soluble nonsugars (calculated from the difference between soluble solids and total sugar), it is to be regarded as an indication of the addition of artificial invert sugar, cane sugar, or grape sugar."

This test, however, has but very little value. Seven of the pure honeys in the table contained less than 1.5 per cent of soluble nonsugars, this being very frequently the case with honeys of low dextrin content, as the alfalfa. Then, again, the invert-sugar préparation itself may contain considerably more than 1.5 per cent of non-sugars, as in the Herzfeld sample, which contained 5.94 per cent of dextrin and undetermined matter. A third objection is that even with the purest preparation of invert sugar it would be impossible to detect adulteration if very low grade honeys of the honeydew type were used for mixing. Over 80 per cent of pure invert sugar could be mixed, for example, with the white oak honey No. 86, and the percentage of soluble non-sugars would still be maintained above 1.5 per

cent.

COLORIMETRIC TESTS FOR INVERT SUGAR.

The uncertainty and practical impossibility of basing a judgment. upon the polariscopic or chemical constants of a honey as to the presence of added invert sugar have led chemists to make use of colorimetric tests, the reaction depending upon the presence, either in the honey or in the invert-sugar preparations, of some compound which even in traces would give some characteristic color reaction.

"König's Untersuchung landwirtschaftlich und gewerblich wichtiger Stoffe, 3d ed., p. 596.

20230-Bull. 110-08-5

LEY'S AMMONIACAL-SILVER TEST.

One of the best known of such colorimetric tests in Germany is that of Ley, which is thus described:

(a) Preparation of the reagent.-Ten grams of silver nitrate are dissolved in 100 cc of water and the solution treated with 20 cc of 15 per cent sodium hydrate solution. The precipitate of silver oxid is collected on a filter, washed with 400 cc of distilled water, and then dissolved in 10 per cent ammonia. The solution is afterwards made up with the ammonia to a weight of 115 grams. The reagent as thus prepared is preserved in a well closed bottle and, if protected from the action of light, is quite stable.

(b) Execution of the test.-A solution of the honey to be examined is made by dissolving 1 part of the same in 2 parts of water. Five cubic centimeters of the filtered solution are placed in a test tube and 5 drops of the silver reagent added. After mixing, the test tube is closed with a plug of cotton and placed in a boiling water bath. The operation should be carried out away from the direct sunlight, and the tube placed in the bath immediately after the addition of the reagent, otherwise color changes may set in which are the result of the action of light. After five minutes the test tube is removed from the bath and the color of the liquid observed.

(c) Results of the test.-Solutions of natural honeys show, after the treatment described, a dark color; the body of the liquid is not directly transparent and is more or less fluorescent in reflected light. Heather honeys especially show the latter characteristic. On shaking the tube the surface of the liquid becomes a clear brownish red and leaves a brownish-green or yellowish-green tint upon the surface of the glass. This greenish color tone is one of the characteristic features of the reaction. Honey substitutes or mixtures of the same with natural honey after similar treatment appear a dirty brown or black and do not exhibit on shaking the characteristic greenish tint above referred to.

The reaction of Ley was applied to nearly all of the 100 pure honeys in the list. It was found that, as a rule, much more dependence was to be placed upon the brownish body color of the liquid on shaking than upon the greenish after-tone; the latter, in fact, was very uncertain in many of the samples, whereas a characteristic brown varying in intensity from light red to purplish was produced in all the honeys, even those of a pronounced honeydew type. The reaction failed completely with all samples of invert sirup, whether prepared by hot inversion, as in the method of Herzfeld, or by inversion at a low temperature. The character of the color reaction with a number of pure honeys, with different invert-sugar preparations, and with mixtures of the latter with pure honey is shown in the following table:

a Pharm. Zts., 1902, 47: 603,

Ley's colorimetric test for invert sugar applied to pure honeys and honey substitutes.

The reaction of Ley has been tested recently by Utz who gives the color appearance with sixty-one different honeys. Utz places most dependence upon the yellowish green after-tone, but notes that this failed completely with a number of pure honeys. Ley also found among fifty-nine honeys five which failed to give the test. The test, therefore, is not an infallible one. As was previously stated, the greenish after-tone failed with so many of the American honeys that no special reliance could be placed upon this phase of the reaction. The brownish color of Ley's reaction was found to have a distinctive value, however, as a confirmatory test in honey examination, though the chemist should thoroughly familiarize himself with the test upon samples of pure and artificial honey before employing it in commercial work. For honeys which have been heated to boiling for any length of time, Ley's test was found to be valueless.

a Zts. angew. Chem., 1907, 20:993.

ANILIN-ACETATE TEST.

The use of concentrated anilin acetate solution has been found in this laboratory to be of great value in testing for invert sugar. The reaction with this reagent is dependent upon the fact that in making invert sugar, by the Herzfeld and similar processes, a small amount of furfural is produced at the high temperature of inversion, which is sufficient to impart a red or pink coloration to the anilin acetate solution when the latter is brought into contact with the honey. The method of conducting the test is as follows:

(a) Preparation of reagent. This should be freshly prepared each time before using. Five cubic centimeters of chemically pure anilin are shaken up with 5 cc of water and sufficient glacial acetic acid added (2 cc) to just clear the emulsion.

(b) Execution of test.-Five cubic centimeters of a concentrated solution of the honey (1:1) are treated in a test tube with 1 to 2 cc of the anilin reagent. The latter is allowed to flow down the walls of the tube so as to form a layer upon the honey solution. If, when the tube is gently agitated, a red ring forms beneath the anilin solution, this color becoming gradually imparted to the whole layer, artificial invert sugar is present.

The above test was applied to all of the pure honeys listed in the table, but with negative results. It may be said, therefore, that there are no substances in natural honey which will react with anilin acetate. It was found possible to detect by this method as small an amount as 5 per cent of Herzfeld's invert sirup in a mixture with pure honey. With natural honeys which have been boiled or heated to a high temperature for any great length of time, traces of furfural are formed just as in the invert-sugar sirups, and these honeys will then react with the anilin reagent. The boiling of honeys, however, is a most questionable operation at best, the flavor of the honey being completely destroyed by this process of cooking. Careful producers and bottlers of honey take extreme pains in putting up their products not to expose the extracted honey to a temperature exceeding 80° C. (176° F.). A positive reaction of a honey with anilin acetate may, therefore, be regarded as an almost certain indication of adulteration with invert sugar. The two tests with the Ley reagent and with the anilin acetate give good confirmatory reactions, the experience in the Bureau of Chemistry having shown that honeys responding to the one test would almost invariably give a positive reaction with the other.

CONCLUSIONS.

The results of the foregoing work upon the composition of American honeys and the methods of detecting adulteration have been applied to the analysis of a large number of commercial honeys. The results of these examinations can not be published as yet, owing to certain requirements of the food and drugs act, June 30,

1906. It may be said, however, that 20 per cent of the large number of samples analyzed were found to contain commercial glucose varying in amount from small quantities (when it was apparently added to prevent granulation) to as much as 80 per cent; 12 per cent of the samples exceeded 8 per cent sucrose, and 8 per cent of the samples gave the reactions for artificial invert sugar with both ammoniacal-silver solution and anilin acetate.

Bottled honeys are frequently labeled by some floral designation, and these honeys, even when within the standard, often show a suspicious variation from the analyses of such honeys of known floral origin, especially as regards sucrose content. The microscopic examination of these honeys frequently shows an almost complete absence of pollen from the flower in question. Bee keepers and bottlers of honey are sometimes exceedingly careless in the labeling of their products and apply the name of almost any flower, thinking that they keep within the requirements of the law so long as their product is pure honey. This is a mistake, as the product should be labeled exactly-according to its origin. The composition of different honey types and the possibility of establishing special standards for these are questions which will require further study.'