TABLE VII.-Sugar in sample No. 123, unsweetened milk.

P. H. Smith and E. B. Holland: Inverted with one-tenth its volume of concentrated hydrochloric acid, neutralized with sodium hydrate solution and proceeded as under reducing sugar. Calculations after deducting for the lactose by association tables, pages 35 and 37, Bulletin 46, Revised edition. Dextrose to sucrose by factor 0.95.

James M. Doran: Trials with some of the well-known methods have proved unsatisfactory. With proper conditions, however, in the inversion and temperature it is thought that both sucrose and lactose can be determined accurately. G. E. Patrick: Lactose corrected for error caused by cane sugar.

C. A. Browne: You will note a variation in the lactose in sample No. 9 by the different methods, which I attribute to the varying degree of inversion of sucrose produced by the different methods of procedure. The Walker method requires only two minutes instead of six as by the Soxhlet, and uses much less alkali than the Allihn with decidedly less inversion in each case. In sample No. 123, where no sucrose was present, the results by the three methods are in very close agreement.

The error due to the inversion of sucrose in the gravimetric determination of sugars in sweetened milks seems to be as pronounced as in the analysis of sugar cane or beet products. We had no time to determine what correction should be made for this, but the experiments performed upon mixtures of lactose and sucrose (see following tables) indicate that an error of at least 0.50 per cent is involved in the case of the determinations by the Allihn method. Weighing the cuprous oxid gave accurate results with all samples, there being no contamination of the precipitate with organic or with mineral matter, as is often the case with molasses.

Effect of inversion of sucrose on the determination of lactose (M. H. Wiley).

The Allihn method was used in all determinations, the dextrose being converted

to lactose by the formula

of water of crystallization.

D

=

lactose. The lactose used had one-half molecule

0.675

Analysis of an artificial mixture of sucrose and lactose as found in condensed

Per cent sugar after inversion as dextrose equals 50.44; per cent sugar before inversion as dextrose equals 8.72; per cent reducing sugars as dextrose from sucrose equals 41.72.

8.72

Therefore 0.692=12.60 per cent lactose (one-half molecule of water).

41.72X0.993=41.47 per cent sucrose.

DISCUSSION OF RESULTS.

The determination of sugars in condensed milk as reported this year indicates a very promising progress. The results as a whole show a fair agreement when one considers the variation in the total solids, which alone is sufficient to cause differences that would exceed those shown in the data reported. The determinations of sugars in unsweetened milk appear to be more uniform than those in sweetened milk. Polariscopic determinations for sucrose are generally lower than those obtained by gravimetric methods and may be attributed to the inversion of a part of the sucrose. The effect of inversion on the determination of lactose, brought about by the addition of sucrose, is interesting, and in consequence a higher lactose reading would be expected. It is evident that in future work along this line it will be necessary to determine the factor for correction in milk sweetened by sucrose.

It is recommended—

RECOMMENDATION.

(1) That the study of methods of analysis for condensed milk be continued and that special attention be given to the following methods for determining the fat and sugar content of condensed milk.

(a) Double extraction method for fat as described in the proceedings of the association for 1906.

(b) Gottlieb method as described in the proceedings of the association for 1906.

(c) Walker method for the gravimetric determination of sugar (J. Amer. Chem. Soc., 1907, 29: 541).

(d) Allihn method for the gravimetric determination of sugar.

(e) Polariscopic method for comparison with the gravimetric method for the determination of sugar.

G. E. PATRICK. In our laboratory we always correct the lactose result for the error caused by the presence of sucrose in analyzing sweetened condensed milks. We have not yet made a table of corrections, but have thus far determined the error in each sample by making a water solution of the two sugars corresponding to the results found, and in this solution-and a similar one of lactose without sucrose-finding the amount of the error in milligrams of copper oxid, which we then deduct from the copper oxid obtained in the analysis. In our experience thus far the correction has been from 12 to 15 milligrams by the Soxhlet method, and 8 to 12 milligrams by the Walker method on the charges prescribed by the official method for sweetened condensed milk.

P. H. WALKER. I have made a few determinations with mixtures of sucrose and lactose which would indicate that the correction needed would be very small. The lactose used, however, was the same as that employed in making up the tables published in the Journal of the American Chemical Society and had been standing for about eighteen months. Not sufficient work has been done to make up tables for this correction.

TWO METHODS IN CHEESE ANALYSIS.

By G. E. PATRICK.

The two following methods for the determination of moisture and the separation of fat in cheese analysis have been used with great satisfaction for several years in the Dairy Laboratory of the Bureau of Chemistry. They are presented to the association with the suggestion that it might be wise to include them in the methods, as in the writer's opinion they are far superior to the procedures there outlined.

1. WATER DETERMINATION.

Sift a quantity of asbestos by rubbing it through a wire screen-a window screen free from rust serves the purpose well. (Iron rust in the asbestos works mischief in the subsequent ash determination.) In a flat-bottomed platinum dish (aluminum may be used if the ash is not to be determined) of 6 to 7 cm diameter place about 2 grams (a little more rather than less) of the sifted

a U. S. Dept. Agr., Bureau of Chemistry, Bul. 105, p. 105.

asbestos, pressing it down well over the bottom, then lay in the dish a thick glass rod cut off square at one end, to serve as a pestle. Ignite and weigh, then weigh into the dish about 5 grams of the cheese sample (prepared as usual) and with the glass rod rub and press the cheese and asbestos together most intimately, until all cheese particles have disappeared and the mass is homogeneous; finally, loosen the mass up into as fluffy a condition as possible, and place in the drying oven. Dry until loss of weight ceases, each drying to be for only an hour or an hour and a half. Usually two or three dryings suffice.

2. SEPARATION OF THE FAT FOR EXAMINATION.

Rub up thoroughly in a mortar 75 grams of the cheese with 90 to 100 cc of ether, throw the mass upon a cloth filter, press out the ether-fat solution as much as possible, return the mass to the mortar and repeat the treatment twice again (the second time less ether will suffice); filter the ether-fat solution through paper, recover the ether by distillation, filter the fat if it is not clear, then dry on the steam bath and finally in the oven until the ether is expelled, weighing every hour in order to guard against oxidation.

Petroleum ether boiling below 40° C. may be used, but has the fault of being expelled rather slowly from the fat. If only the neutral fat, free from uncombined fatty acids, is desired, add to the original extract two or three drops of phenolphthalein solution and then soda solution dropwise, with constant shaking, until the aqueous portion remains pink; then proceed as already described.

REPORT OF COMMITTEE C ON RECOMMENDATIONs of refEREES. By L. M. TOLMAN, Chairman.

It is recommended that

(1) MEAT PROTEIDS.

(1) The modified tannin-salt method for the separation of proteoses and peptones from the simpler amido bodies be adopted as a provisional method. (For statement of method, see referee's report, p. 51.)

The committee recommended that this method be further studied for another year before its adoption, and this motion was passed.

(2) That the following method for the determination of the acidity of meat products be adopted as a provisional method:

Determine the acidity by titrating a dilute solution of the meat, or meat preparation, with N/10 alkali, using phenolphthalein as indicator. But if the color prevents the use of this indicator, use litmus paper, with the understanding that the results are lower than when phenolphthalein is employed.

Adopted.

(3) That inasmuch as the best modification of the Folin colorimetric method for the determination of kreatin and kreatinin in meat products has not been determined, this point be studied another year.

Adopted.

(4) That the referee for 1908 investigate methods for the estimation of organic phosphorus and of xanthin bases.

(1) That methods for the detection of caramel and other coloring matter in vanilla extracts be studied.

Adopted.

(2) That the quantity and quality of soluble matter in vanilla beans obtained in the preparation of standard vanilla extract be determined by the referee in 1908.

Adopted.

(3) DETERMINATION OF MOISTURE IN FOODS.

It is recommended that the work on this subject be continued another year. Adopted.

(4) COLORS.

It is recommended that further work be done looking to a more comprehensive examination of the so-called “vegetable colors," some of which would seem to be no more properly classed as vegetable colors than are the anilin dyes. Adopted.

(5) VEGETABLE PROTEIDS.

It is recommended that special attention be given to the study of methods for the determination of proteid and nonproteid nitrogen.

Adopted.

(6) COCOA AND COCOA PRODUCTS.

It is recommended that the methods submitted by the referee" be adopted as provisional until a fuller report on the subject can be made, except that for section 13 on sugar the method proposed by Dubois for the determination of sugar in chocolate be substituted.

Adopted.

REPORT OF THE COMMITTEE ON THE REVISION OF METHODS.

J. K. HAYWOOD, Chairman.

The following report was presented by Mr. L. M. Tolman, acting for the chairman:

In accordance with the recommendations made in the report of the committee on revision of methods in 1906, and the suggestions of the secretary of the association made at that time, the revised methods were printed as Bulletin No. 107 of the Bureau of Chemistry and forwarded to the members of the association for criticism about three weeks before the 1907 meeting. Copies of this bulletin were sent to the following lists: Officers and referees of the association; directors and chemists of State agricultural experiment stations; superior officers of State boards of health, and chiefs of laboratories in the Bureau of Chemistry, Department of Agriculture.

The following letter was sent with each copy :

SEPTEMBER 12, 1907.

DEAR SIR: Inclosed you will find a copy of the revised methods of analysis of the Association of Official Agricultural Chemists recommended last year by the committee on revision and provisionally adopted by the association. Final action upon these revised methods will be taken this year.

You are respectfully requested to look these methods over at once, especially those with which you are most familiar. If you have any criticism to make of the changes incorporated, or can point out any errors, please report same to J. K. Haywood, Bureau of Chemistry, U. S. Department of Agriculture, Washington, D. C., chairman of revision committee, so that the letter will reach him before October 5, if possible, certainly not later than October 7. All criticisms will be considered by the committee at a meeting to be held in Washington on Octo

"U. S. Dept. Agr., Bureau of Chemistry, Bul. 107 Revised, p. 254.