The samples Nos. 8-14 and No. 36, were from the manufactory of Armour & Co.; Nos. 24-26 from Fitts & Co.; Nos. 27-29 from Wm. J. Moxley; No. 37 from Fairbanks Canning Co., all of Chicago, Ill., No. 3 was sent upon request, from Polar Creamery, Houston, Minn., firm Hostvet & Hourn. The suspected samples were sent to this Station for analysis by private parties.

The three samples of genuine butter were pure beyond doubt. The samples are given the names under which they were sold. Out of eleven samples bought at Chicago grocery stores at random, five were artificial butter.

As will be seen, the determination of the melting point is of no use in discriminating between a natural and an artificial butter. The specific gravity test and Koettstorfer's or Reichert's methods, on the other hand, either singly or combined, are decisive.

Oleo oil, neutral oil, and sesame oil are used in the manufacture of artificial butter; oleo oil, made from the caul fat or kidney fat of beeves, has on an average, a specific gravity of 0.90369; melting point, 27.6° C; by Koettstorfer's method one gramme of the fat requires for saponification 193.4 mgrms. of potash, and by Reichert's method 2.5 grams of the fat gives a distillate the neutralization of which requires 0.09 c. c. of deci-normal soda solution. Neutral oil, made from the leaf lard of hogs, has the following average numbers-specific gravity 0.90530; melting point 38.1° c; Koettstorfer's method 193.3 mgrms. potash, and Reichert's method 0.16 c. c. of deci-normal soda solution.

The specific gravity of genuine butter is usually considered to range from 0.91400-.91200 at 100° F.: seven of the eleven samples of genuine butter reported on above had a lower specific gravity than this limit. The lowest result found was 0.91107.

The samples of artificial butter which were not oleomargarine, — i. e., prepared only from oleo oil, neutral oil and a little cream or milk, used in churning the mixture-contained genuine butter approximately in the following percentages:

The sample No. 30, containing 80 per cent. of natural butter, must have originated in some creamery, and would be more properly called adulterated butter, it would not pay for a manufacturer of artificial butter to use such a large quantity of natural butter. As a rule, not more than 50 per cent. of pure butter enters the artificial product.

NOTES ON BARON HÜBL'S METHOD OF ANALYSIS OF FATS.*

BY F. W. A. WOLL, M. S.

As a supplement to the foregoing article an account is given below of investigations carried on in the fall and winter of 1886-87 in regard to the method proposed by Baron Hübl for the chemical analysis of fats, with special reference to its applicability in butter analysis and for discrimination between natural and artificial butter. For this purpose 53 samples in all were analyzed, of which number 25 were natural butters, 10 butterine, 5 oleomargarine, 4. oleo oil, 2 neutral oil, and the rest commercial oils.

The analyses were carried on essentially in the manner prescribed by Hübl in his original article. For the sake of brevity only an outline of the method of proceeding used will here be given. Eight-tenths to one gramme of the pure filtered butter fat is weighed into a 250 c. c. flask while in the fluid state. Ten c. c. of chloroform are then added, and when the fat is dissolved, a standardized iodine solution prepared as directed by Hübl is dropped into the chloroform solution from a burette, with constant shaking, until the solution remains brown. Sufficient iodine solution to preserve

* Reprinted from Agricultural Science, Vol. I, p. 79-83.

+ Dingler's Polytechn Journal, CCLIII, 281; Fresenius' Analytische Zeitschrift, xxv, 432 (abstract).

the color for two hours must be used. With a little experience one may easily see when a sufficient excess of iodine has been added. After two hours' standing the amount of unabsorbed iodine in the solution is determined by a one-tenth normal hyposulphite of sodium solution; when the end of the reaction is nearly reached, a tight cork is inserted in the flask, and its contents are shaken vigorously in order to dissolve the iodine contained in the layer of chloroform at the bottom of the flask. The results are calculated as is usual in similar reactions. The iodine number of a fat expresses, according to Hübl, the number of grammes of iodine absorbed by 100 grammes of the fat.

The samples analyzed by the writer were partly the same as before reported upon,* partly new ones, of pure butter procured from the dairy of the Station; these samples, from Nos. 43 to 54, were all churned from the mixed milk of the herd. The other samples are given the numbers under which they were received at the Station.

Of the constituents of butter fat, olein, C3 H5 (C18 H33 O 0)3 is the only compound of an acid of the ethylene gas series, CnH2n-2 O2, and is thus the only constituent of butter fat ans similar bodies composed of only tri glycerides which forms an addition product with the halogens, in this case with iodine. We have consequently in the Hübl method a rapid and easy means of ascertaining the amount of olein present in any butter, be it natural or artificial, upon the supposition now generally accepted that these fats contain only tri-glycerides. In the table below are given the percentages of olein corresponding to the iodine number found for each sample, calculated by a simple equation:

* See preceding article.

It will be seen that the iodine number found for pure butters varies from 25.7 to 37.9, with an average for twentyfive samples of 30.83. The samples from the Station's dairy, Nos. 43-54 inclusive, had iodine numbers varying from 26.5-37.9, with an average of 30 0, the samples with the highest iodine number being the first ones in the series, taken in the month of October. According to Blyth,* butter contains 42.21 per cent, of olein; this corresponds to an iodine number of 36.38, a number which lies within the range found by other writers, and also within the results recorded above. The figures in the third column show that the percentage of olein in natural butter varies considerably. This might be expected from what we know about butter in other respects; its specific gravity, melting point, per

* Blyth: Foods, London, 1882, page 285.

centage of volatile and of insoluble fatty acids, all change according to the circumstances which influence the cow during the formation of the milk; food, temperature, treatment, etc., being important factors, as well as the breed, milking qualities and individuality of the cow. According to Gottlieb,* natural butter contains 30 per cent. of olein; Muter found 42.22 and 36.37 per cent. of olein in two samples of butter. Braconnot and Boussingault state that olein ("butter-oil") is present in large quantity in summer butter, up to 60 per cent.; "margarine" (palmitin and stearine), on the other hand in winter butter, where olein may be present to the extent of only about 35.0 per cent., which corresponds with the above figures. The percentages found by the writer vary from 29.8 to 44.0, with 35.9 per cent. as the average of twenty-five samples.

Butterine, as the term is used in the United States, has come to mean oleomargarine or artificial butter with different quantities of natural butter admixed. The results of analyses of ten samples of butterine are given below, compared with the results found by Reichert's method.§ The second column gives the percentage of pure butter found in the different samples by the same method; in the two next columns will be found the iodine number observed and the one calculated on the basis of the percentage of butter found by Reichert's method, using the following formula:

where n means the iodine number of the mixture and x the percentage of genuine butter present. In the next to the last column the amount of natural butter present in each sample is calculated from the last equation on the basis of tho observed iodine numbers. As will appear at a glance, there is a considerable difference in the percentage of butter found by the two methods. Reichert's method gives the

* Annalen der Chemie und Pharmacie, LVII, 39.

The Analyst, II, 73.

Ann. der Chemie und Physik, VIII, 96.

§ See previous article.