To this end, the average probable error for each method and each product, as well as for the three combined, has been calculated by the expression 2 S 3 N-1 The averages included in Table 5 were calculated from all the individual determinations made by each particular method, using the formula s 3 N-1 TABLE 5. Average probable errors of the determinations. PRODUCT Sirup.. First molasses. Final molasses. Average. It METHOD METHOD METHOD METHOD METHOD METHOD METHOD V *Without the use of ammonium carbonate. † With ammonium carbonate. It is readily seen that from the standpoint of chance of error, the sulphated ash method is superior to the others in the case of sirup, but this is not generally true in the case of first molasses, and less so with the final molasses. Even if consideration is given to the fact that the errors shown in the table for the sulphate method are in reality proportionately smaller, on account of the correction factor used in reducing sulphated ash to carbonated ash, the sulphated ash method does not seem to offer any advantages over Methods I and II with ammonium carbonate in point of chance of error. Apparently, these are the two methods by which different analysts show the closest agreement. Passing now from the question of relative accuracy to that of absolute accuracy of the different methods used, there is, of course, no infallible standard by which to judge which of the first four methods (Methods I and II, both with and without the use of ammonium carbonate) gives the closest approach to the true ash content. The sulphated ash method will have to be considered separately, since in this case the ash must be calculated from the value actually obtained, by applying a correction factor. Tables 1 to 3 show that Method II, in which the soluble salts are leached out with water and after incineration of the insoluble portion combined with it and incinerated again, has a tendency to give slightly higher results than Method I—the method of direct ashing. This seems somewhat strange, the object of Method II being to facilitate the combustion of the carbon, which would lead to high results if not removed. But it must be considered that in the operation of filtering and evaporat ing, the solution may easily take up dust particles which will increase the weight. In the experience of the writer, Method II has no advantage over Method I in the three samples analyzed, because it was found easy to burn off the carbon without first dissolving the soluble salts. The results obtained by Method II, without the use of ammonium carbonate, agree so closely with those obtained by Methods I and II, with ammonium carbonate, that there does not seem to be any reason for using this very slow and tedious method, except when it is found impossible to get a carbon-free ash by the shorter method. The use of ammonium carbonate in either Method I or II has a tendency to give slightly higher results, as may be expected, the object of its use being the reconversion into carbonates of any alkaline earth oxids formed. Considering that the relative accuracy of Methods I and II, with ammonium carbonate, is about the same, and that both give very closely agreeing results, it would appear that for general purposes Method I, using ammonium carbonate, should be given preference, except when necessity compels the use of the longer procedure. In the experience of the writer, an analysis by Method I, using ammonium carbonate, can be made just as easily and quickly as by the sulphated ash method, provided foam formation is prevented by the use of olive oil or some similar substance. The sulphated ash method is the one most commonly used in this country in commercial work, and will therefore be discussed more in detail. The official method gives a correction factor of one-tenth, to be deducted from the actual result obtained, in order to convert it into a supposedly equivalent amount of carbonated ash. This correction factor of one-tenth has been the object of investigation on the part of a large number of workers, and an important contribution to this field. has lately been made by Ogilvie and Lindfied. They analyzed a number of sugars and molasses, from both beet and cane, by substantially the same methods used in the work here reported, and found that the conversion factor actually varied from 12 to 25 per cent in the case of beet sugars, averaging 15 per cent; from 12 to 17 per cent in that of beet molasses, averaging 15 per cent; from 6 to 22 per cent with cane sugars, averaging 14 per cent; and from 14 to 21 per cent with cane molasses, averaging 18 per cent. Ogilvie and Lindfield also give a summary of previous work on this question, which shows that this problem has been studied quite extensively in connection with beet products, but that very little has so far been done on cane products. Since the correction factor of one-tenth is still generally used in this country, it seems most 1 Intern. Sugar J., 1918, 20: 114. important to ascertain whether this factor is correct for the cane products handled in the United States, because, if a maximum ash figure should be established as the basis of any of the official standards, then too low a correction factor might in many cases prove detrimental to the seller of these products. Let us suppose, for instance, that a maximum of 5 per cent of ash should be fixed for a certain product, and that an analysis shows 6 per cent of sulphated ash. With the correction factor of one-tenth, adopted by the association, the true ash would be found to be 5.40 per cent, ruling out this particular product. But if the true correction factor be, say one-fifth, instead of one-tenth, the corrected ash will be found to be 4.80 per cent, quite within the limit of the maximum. Turning again to Tables 1 to 3, it will be noted that the amount of sulphated ash found increases with the quantity of sulphuric acid used. This is quite in agreement with the results of Schweizer1, who showed that a much higher ash figure is obtained by re-treating the sulphated ash with sulphuric acid, to decompose any sulphids formed. It follows that the correction factor varies with the quantity of sulphuric acid added, and the following table shows what the correction factor is in the case of the three products analyzed in this investigation, taking the results by Method I, using ammonium carbonate, for establishing the factor. Four more samples, different from the above, were analyzed by E. C. Freeland, and the results are given in Table 7. The comparisons are made with results obtained by Method II, without the use of ammonium carbonate. The correction factor is evidently much nearer 20 than 10 per cent, as already pointed out by Ogilvie and Lindfield, and others before them. In the seven products analyzed it is rather constant, averaging about 19 per cent. But this may be accidental, and if the correction factor should in a large number of samples really show such wide variations as found by the authors just mentioned, the sulphated ash method should not be used at all in food control work concerning these products, except where the total ash content is so low that the difference between the highest and lowest correction factors falls within the limit of experimental error. The results reported show that from the standpoint of close agreement between different analysts the sulphated ash method has no advantage over the direct ash method, and, in the writer's opinion it even has no advantage from that of ease of manipulation. It has furthermore been confirmed that the sulphated ash method, as adopted officially by the association', may give results widely divergent from those obtained by the direct ash method. RECOMMENDATIONS. It is recommended (1) That a study be made of the influence of different and known temperatures of incineration on the results of ash determinations in cane sirups and molasses, carrying out the incineration in both platinum and silica dishes for comparison. (2) That a large number of samples of different grades of cane sirups and molasses be used for comparing ash determinations by the sulphate and direct methods, to determine, if possible, the proper correction factor to be applied to sulphated ash. No report on sugar house products was made by the associate referee for the year ending November, 1919. No report on food preservatives was presented because of the death, during the year, of A. F. Seeker, the referee 1 Assoc. Official Agr. Chemists, Methods, 1916, 128. REPORT ON COLORING MATTERS IN FOODS1. By W. E. MATHEWSON (Bureau of Chemistry, Washington, D. C.), Referee. No work on colors was done by the association in 1918. In March, 1919, the following circular letter was sent to the collaborators: Since the last meeting of the association a number of papers and reports have been published of interest in connection with the analysis of colored food products. Willstaetter and Schudel' describe a method for separating basic coloring matters from aqueous solutions, depending on the extraction of the picrates (or dichlorpicrates) by immiscible solvents. The procedure is intended especially for the separation of the anthocyans or glucoside fruit colors, and the authors give the figures in Tables 1 and 2 for the distribution ratios as found when such solutions are shaken with the solvents named. No detailed description has been given, but presumably the solutions extracted are treated with an excess of the acid. Tests for the fruit colors have usually been made in the fruit juices or other mixtures in which they occur, as no satisfactory methods for their preliminary purification have 1 Abstract. 2 Ber., 1918, 8: 782. |