TEE ANALYST. 355 N O T E O N H O N E Y . By HENRY LEFFMANN. SOME time ago I was consulted by a manufacturer of table accessories in regard to a sample of honey which had been reported by one of the State chemists as adulterated. The manufacturer did not have any portion of the sample which had been analysed, but brought me some samples which he believed to be from the same lot, and said that the material was from California, and was packed as received, being believed to be pure. Inquiry elicited the fact that the sample purchased by the State’s agent had been found to give the following polarimetric results on the ordinary sugar-scale : Before hydrolysis ... ... ... ... ... 9.0 After hydrolysis ... ... 3.0 ... ... ... The sample submitted to me gave results of the same import, although not exactly the same figures, namely : Before hydrolysis ... ...... .*. ... 13.5 After hydrolysis ... ... ... ... ... 9-5 Dextrorotatory honey is usually looked upon with suspicion. When the sample remains strongly dextrorotatory after hydrolysis, the addition of commercial glucose is assumed. The following figures are reported on good authority as given by samples of undoubted purity : I t is, however, established that pure honey may be dextrorotatory. Before hydrolysis ... .., ... 8.2 7.2 7.3 After hydrolysis . . , ... ... 2.8 3.3 2.6 Polarimetric data are so technical that it is desirable to have some chemical tests for the presence of glucose as such. Most of the data that are herewith pre- sented are principally of value as embodying the application of German methods to American samples, and, as will be seen, confirming the claims of the German chemists.Commercial glucose is by no means the simple substance that it was supposed to be when first manufactured. It contains, in addition to dextrose. considerable dextrin and maltose, some nitrogenous matter: and a notable amount of unfermentable optically active carbohydrates of uncertain nature, among which is one called gallisin. As strong alcohol precipitates dextrin from glucose, the suggestion to test with this would naturally occur; but although honey is largely invert-sugar, it356 THE ANALYST, frequently contains considerable amounts of dextrinous matter that precipitates with strong ethyl-alcohol. Bechmann suggested the use of methyl alcohol, and tests I have recently made have shown the value of this method.The refined methyl alcohol, termed '' Columbian spirit," is preferable. Two methods are employed. The simplest consists in mixing the honey with an equal weight of water and then adding a large volume of the alcohol. The quantities used have been 8 grammes of honey, 8 C.C. of water, and 100 C.C. of alcohol. The characteristic of glucose is to give a sticky precipitate which may become somewhat granular, and which adheres to the wall of the vessel. Pure honey gives a light flocculent precipitate, which does not adhere to the glass. For the second test a solution of 20 grammes of the honey in 100 C.C. of water is prepared. Five C.C. of this solution are mixed with 3 C.C. of a 2 per cent. solution of barium hydroxide and 17 C.C.of methyl alcohol, and the mixture shaken with as little contact with air as possible, Pure honey gives but little precipitate, but in the presence of glucose or glucose syrup a considerable precipitate is formed. These tests applied to the suspected honey and to samples believed to be pure, and also to mixtures of such samples with commercial glucose, gave very cheracter- istic results. The barium method is unsatisfactory on accouut of the action of the air, which cannot be easily avoided. An attempt to substitute a solution of barium carbonate in water containing carbonic acid did not succeed. Apparently the direct action of the hydroxide upon the carbohydrates is needed. I t seemed probable that useful information might be obtained by comparing the condition of solutions before and after fermentation, including the polarimetric data, especially as applied in the examination of wine.The following experiments were made : Three solutions were prepared by dissolving 25 grammes of material in water made up to 250 C.C. The polarimetric reading of each was taken, and the solutions were then fermented at room temperature (over 20" C.) for six days, and again examined. A few drops of a solution carbonate had been added to each. days. The results were as follows : Pure honey ... ... ... Commercial glucose ... One-half honey, one-half glucose of disodium hydrogen phosphate and ammonium The fermentation was quite active for several Polarimetric Reading. Fermented Solution. S p e n S o i i c i s Before After & Fermentation. Fermentation.Gravity. per Cent. ... - 6.0 3.0 1,003 1.6 ... 65.0 43.7 1,014 4.2 ... 25.6 22-5 1,009 2.8 The above data are too few to form a basis for argument, but they tend to show that the fermentation method may aid in elucidating the nature of sample under dispute. Some experiments were made with a view of determining the polarimetric character of solutions of samples in methyl alcohol. Crude methyl alcohol was used. In each case 8 grammes of the sample were treated with 100 C.C. of methyl alcohol, well mixed, and allowed to stand for more than a day. The clear liquid was thenTEE ANALYST. 357 decanted and the reading taken. ture. The results were : All samples were examined at the same tempera- Precipitate PolarimePrio Sample. with Methyl Alcohol. Reading. ... 1.0 Pure honey ... ... ... ... Slight, flocculent ... Suspected honey ... ... ... Marked, granular ... ... 22.0 Pure honey (glucose equal parts) ... Marked, very sticky ... ... 31.0 Glucose ... ... ... ... Marked, sticky ... ... 10.5