No associate referee on the subject of honey was appointed and no report on this subject was presented.

Messrs. C. S. Hudson1 and S. F. Sherwood2 (Bureau of Chemistry, Washington, D. C.), presented a paper on "The Occurrence of Melezitose in Honey".

E. T. Wherry (Bureau of Chemistry, Washington, D. C.), presented a paper on "Crystallography of Melezitose".

1 Present address, 204 Academy Ave., Trenton, N. J.

2 Present address, Bureau of Plant Industry, Washington, D. C.

3 J. Am. Chem. Soc., 1920, 42: 116.

• Ibid., 125.

THE DETERMINATION OF ASH IN CANE SIRUPS AND

MOLASSES.

By F. W. ZERBAN1 (Louisiana Sugar Experiment Station, New Orleans, La.), Associate Referee on Sugar House Products2.

Since there is a possibility that the percentage of ash in sirups and molasses may, in some way, play a part in the official food standards to be established for these products, it seemed advisable to carry out a comparative study of the methods for the determination of ash, as adopted by the association. There are at present three such methods3, all of them official. The two principal questions to be settled are: First, which of the three methods yields the most concordant results in the hands of different analysts; and second, how well the results obtained by the three methods agree with one another. The first of these questions appears to be the more important from the standpoint of food control, although the second should also be settled as well as can be done in a matter involving the determination of a constituent as vaguely defined as "ash". For the present, we shall call "ash" that part of a substance which remains after heating it in the presence of air to a temperature high enough to destroy its organic constituents or convert them into carbonates, but not sufficiently high to volatilize the inorganic constituents to any extent.

In order to get an answer to the first question, and at the same time some information on the second, three samples of genuine cane products were sent out to a number of chemists who had expressed their willingness to cooperate. One of the samples was a cane sirup; another a first centrifugal molasses; and the third a final molasses.

Collaborators were asked to determine the ash in each of these samples by all of the following methods:

Method I.

Place 5 grams of the sample in a 50-100 cc. platinum dish (or a smaller quantity, if only small dishes are available), add a few drops of pure olive oil, and heat slowly over a flame until swelling ceases. Then place the dish in a muffle and heat at a very dull red heat, until all the carbon is completely burned off. Repeat the heating in the muffle, until constant weight is obtained. Record the result.

Take up the residue with a little ammonium carbonate solution, reevaporate, and heat again in the muffle at very dull red heat to constant weight. Record the result.

Method II.

Place 5 grams of the sample in a 50-100 cc. platinum dish (or a smaller quantity, if only small dishes are available), add a few drops of pure olive oil, and carbonize the mass over a free flame at the lowest possible temperature. Dissolve the soluble salts in hot

1 Present address, Penick & Ford Ltd., Inc., Marrero, La.

For the year ending November, 1918.

* Assoc. Official Agr. Chemists, Methods, 1916, 128.

water, and ignite the carbonaceous residue as directed in Method I. Then add the solution of soluble salts, evaporate to dryness at 100°C., and heat in a muffle at very dull red heat, until constant weight is obtained. Record the result.

Take up the residue with a little ammonium carbonate solution, reevaporate, and heat again in the muffle at a very dull red heat to constant weight. Record the result.

Method III.

Place 5 grams of the sample in a 50-100 cc. platinum dish or silica dish (or a smaller quantity, if only small dishes are available), add 0.5 cc. of concentrated sulphuric acid per 5 grams of sample (if less was taken, reduce the amount of sulphuric acid accordingly), heat gently over a free flame until the sample is well carbonized. Then heat in the muffle at low red heat to constant weight. Record the result.

Method IV.

Same as Method III, except that 1 cc. of sulphuric acid per 5 grams of sample is used.

Method V.

Same as Method III, except that 2 cc. of sulphuric acid per 5 grams of sample are used.

Collaborators were asked to make duplicate determinations in each case, and if the results differed by 0.2 per cent or more, to run a third determination.

Ten analysts, to whom the writer wishes to express his thanks, reported results on all or part of the samples, and the results are given in Tables

TABLE 2.

Determination of ash in first molasses.

ANALYST

I*

It

METHOD METHOD METHOD METHOD METHOD METHOD METHOD
II*
III

IV

V

1 to 3. Only the averages of duplicate determinations are shown, and where triplicates were run, the average of the figures nearest each other. On the whole, duplicates made by the same analyst showed very good agreement. Differences of 0.1 per cent or more occurred in 12 cases out of 59 in the sirup analyses, 14 out of 62 in the analyses of the first molasses, and 16 out of 55 in the analyses of the final molasses. Duplicates differed by 0.2 per cent or more in only one case in the sirup analyses, and in three in the analyses of the final molasses. In other words, the agreement is not so good in the more impure products.

The tables show clearly that the agreement between different analysts is not nearly so good as that between duplicates of the same analyst.

TABLE 4.

Maximum variations between the results obtained by the various analysts.

These maximum variations are, of course, more or less accidental, but they show, nevertheless, that the deviations between results are smaller in the higher grade products, as had been found to be the case for the individual analyst.

It is further noted that the results of one and the same analyst are usually in one direction from the average, either generally high or low, or near the average. This would seem to indicate that each analyst did all of his work under practically identical conditions, but that these conditions differed with different analysts. In one case, that of Roberta Hafkesbring, this was probably due to the fact that silica dishes were used instead of platinum, and this may be responsible for her results being quite below the average. Contact with silica at high temperatures may drive out sulphuric acid under formation of silicates. This point should be further investigated. But in the other cases, where platinum was used, the differences were most probably due to different temperatures being used for the incineration. It will therefore be most important to continue this cooperative work under conditions where the temperature is kept constant and at the same time measured, and to make determinations at different temperatures, in both platinum and silica dishes. From the results obtained, some important conclusions on the accuracy of each method, in the hands of different analysts may be drawn.