The following analyses by Leach illustrate the difference in composition between whole-milk and skim-milk cheese:

"Filled cheese" is the common name for a product in which a foreign fat, such as oleo oil or lard, has been used. Formerly filled cheese was made in the United States for the export trade but owing to drastic laws it is now practically unknown.

ANALYSIS OF CHEESE.

Sampling. If the cheese is spherical or drum-shaped, cut a narrow, wedge-shaped segment, reaching from the circumference to the center, if brick shaped, cut a thin slice through the middle; the rind may be included in the sample or rejected according to the purpose of the analysis. The sample is prepared for analysis by chopping, grating, or kneading until uniform, taking care to weigh before and after the treatment if loss of moisture is sustained.

A less accurate, but for many purposes quite satisfactory, method of sampling is to remove cores from different parts of the cheese with a trier using the rind to stop up the holes. By this procedure the cheese is not materially damaged.

Determination of Water.-Dry 5 grams of the sample in a flat-bottomed metal dish 5 cm. in diameter in a boiling-water oven until the weight is practically constant. If ash is subsequently to be determined in the same portion a platinum or porcelain dish must be used.

Winton* has found that reasonably constant weight is obtained in drying American full cream and skim-milk cheese only after 12 to 18 hours while in drying Roquefort losses, evidently other than water, continued for days. For practical purposes drying for 12 hours is sufficient.

Determination of Fat.-Lythgoe's Modified Babcock Method.-Weigh accurately about 6 grams of the sample in a tared beaker. Add 10 cc. of boiling water, and stir with a rod till the cheese softens and an even

*U. S. Dept. of Agric., Div. of Chem., Bul. 35, 1892, p. 13.

emulsion is formed, preferably adding a few drops of strong ammonia to aid in the softening and emulsionizing, and keeping the beaker in hot water till the emulsion is tolerably complete and free from lumps.

If the sample is a full-cream cheese, which is usually evident from its taste and appearance, a Babcock cream-bottle is employed. The contents of the beaker, after cooling, are transferred to the test-bottle as follows: Add to the beaker about half of the 17.6 cc. of sulphuric acid regularly used for the test, stir with the rod and pour carefully into the bottle, using the remainder of the acid in two portions for washing out the beaker. Finally proceed as in the regular Babcock test for milk. Multiply the fat reading by 18 and divide by the weight of the sample taken to obtain the per cent of fat.

Short's Method.*-Grind to a uniform powder 2 to 5 grams of the sample, and about twice its weight of anhydrous copper sulphate. Place a layer of anhydrous copper sulphate about 2 cm. thick on the bottom of the inner tube of a Johnson or Knorr extractor, add the ground mixture, and rinse the mortar first with a little anhydrous copper sulphate and finally with ether. Extract for 16 hours, evaporate the ether from the extraction-flask, and dry the fat in a boiling-water oven to constant weight.

Werner-Schmidt Method.-Boil 2 to 3 grams of the sample in the Werner-Schmidt tube (p. 126) with 5 cc. of water and 10 cc. of concentrated hydrochloric acid till, with constant shaking, all but the fat is dissolved. Cool, add 25 cc. of ether, and shake the tube well. Draw off as much as possible of the ether, after separation, in the usual manner, and extract with four or five additional portions of the solvent.

Distil off the ether from the combined extractions, and weigh the fat. Determination of Total Protein. Calculate from the nitrogen, determined by the Kjeldahl or Gunning method on 1 to 2 grams of the sample, using the factor 6.38. Van Slyke adds a piece of copper sulphate the size of a pea during the digestion.

Separation and Determination of Nitrogen Compounds. Methods of Van Slyke.t-Twenty-five grams of the sample are mixed in a porcelain mortar with an equal volume of clear quartz sand. Transfer the mixture to a 450-cc. Erlenmeyer flask, add about Ico cc. of water at 50° C., and keep the temperature at 50° to 55° C. for half an hour with frequent shaking. Decant the liquid through an absorbent-cotton filter into a * U. S. Dept. of Agric., Div. of Chem., Bul. 35, pp. 15, 17, 225.

† Van Slyke, N. Y. Exp. Station, Bul. 215.

500-cc. graduated flask. Treat the residue with repeated portions of 100 cc. each of water, heating, shaking, and decanting as above till the filtrate, or water extract, at room temperature amounts to just 500 cc. exclusive of the fat floating on top, and use aliquot parts of this water extract for the various determinations.

Water-soluble Nitrogen.-Determine the nitrogen by the Gunning method in 50 cc. of the above water extract, corresponding to 2.5 grams of cheese.

Nitrogen as Paranuclein.—Add 5 cc. of a 1% solution of hydrochloric acid to 100 cc. of the above water extract (corresponding to 5 grams of cheese), and keep the temperature at 50° to 55° till the separation is complete, as shown by a clear supernatant liquid. Filter, wash the precipitate with water, and determine the nitrogen therein by the Gunning method.

Nitrogen as Coagulable Protein.-Neutralize the filtrate from the preceding determination with dilute potassium hydroxide, and heat at the temperature of boiling water till the coagulum,* if any, settles completely. Filter, wash the precipitate, and determine the nitrogen therein.

Nitrogen as Caseoses.-Treat the filtrate from the preceding with 1 cc. of 50% sulphuric acid saturated with C. P. zinc sulphate, and warm to about 70° C. till the casesoses settle out completely. Cool, filter, wash with a saturated solution of zinc sulphate acidified with sulphuric acid, and determine the nitrogen in the precipitate.

Nitrogen as Amino Acids and Peptones.-Place 100 cc. of the water extract in a 250-cc. graduated flask, add 1 gram of sodium chloride and a solution containing 12% of tannin, till the addition of a drop to the clear supernatant liquid does not further precipitate. Dilute to the 250-cc. mark, shake, pour upon a dry filter, and determine the nitrogen in 50 cc. of the filtrate, which gives the amount of nitrogen in the amino-acid and ammonia compounds. Deduct from this the amount of nitrogen as ammonia separately determined, and the difference is the amino-nitrogen.

Nitrogen as peptones is obtained by subtracting the sum of the amounts of nitrogen as paranuclein, coagulable proteins, caseoses, amino-bodies, and ammonia from the total nitrogen in the water extract.

Nitrogen as Ammonia.-Distil Icc cc. of the filtrate from the above tannin-salt precipitation into standardized acid, and titrate in the usual

manner.

According to Van Slyke a precipitate at this point is rare in cheese.

Nitrogen as Paracasein Lactate.-Treat the residue insoluble in water in obtaining the water extract, with several portions of a 5% solution of sodium chloride, to form a 500-cc. salt extract of the same, in an analogous manner to that employed in preparing the water extract. Determine the nitrogen in an aliquot part of this salt extract.

Determination of Lactose.-Rub up ten grams of the sample with water at 40 to 50° C., and decant onto a filter. Repeat the operation until the filtrate measures 100 cc. Determine lactose in an aliquot by one of the methods described under milk.

Except in the case of whey cheese, lactose is present only in minute amount if at all. Commonly lactose, lactic acid, and other minor constituents are determined by difference subtracting the sum of the water, fat, total protein, and total ash from 100.

Determination of Ash.-Ignite at dull redness 5 grams of the cheese in a tared platinum or porcelain dish until a white ash is obtained, using preferably a muffle furnace, cool in a desiccator, and weigh.

Determination of Sodium Chloride.-Treat the ash with water acidified with nitric acid, filter, wash, and precipitate the chlorine in the filtrate with silver nitrate solution. Filter on a Gooch crucible, heat cautiously to melting, and weigh. Calculate the equivalent sodium chloride.

The chlorine may also be determined volumetrically, without filtering, by Volhard's method.

Acidity.-Jensen and Plattner Method.*-Rub up 10 grams of the sample in a mortar with water at 40 to 50° C., and decant onto a filter. Repeat the rubbing and decanting until the filtrate measures 100 cc. Titrate an aliquot with N/10 sodium hydroxide, using phenolphthalein as indicator; I cc.=0.009 gram lactic acid.

Although the acidity is usually calculated as lactic acid, other acids or acid-reacting substances may be present.

Detection of Foreign Fat.-The cheese fat, separated in the manner described below, is subjected to the various processes detailed under butter, in precisely the same way, the fat of cheese being identical with that of butter. The most ready means for judging its purity consists in determining the refraction with the butyro-refractometer, and the Reichert number.

The distinction of cow's milk from goat's milk fat is best accomplished, according to Hals and Sunde † by the ratio of the Polenske to the Reichert*Zeits. Unters. Nahr. Genussm., 12, 1906, p. 193.

†Tidskr. Kem. Farm. Ter., 5, 1908, p. 8.

Meissl number which for goat's milk is about 1:5 and for cow's milk I: 8 to 19.

Separation of the Fat for Examination.-Place a quantity, say 25 grams, of the finely divided sample in a large Erlenmeyer flask, add about 100 cc. of petroleum ether, cork the flask and allow it to stand for several hours with frequent shaking. Decant the petroleum ether through a filter, evaporate off the solvent by the aid of heat, and the residue will be found to consist of nearly pure fat.

Or, wrap a sufficient portion of the finely divided sample in a muslin cloth, place this in a dish, and heat on the water-bath. The fat which runs out is afterward filtered and dried at 100°.

Sufficient cheese fat may usually be obtained for the refractometer reading from the neck of the test-bottle, after completing the Babcock test, and, usually (except in the case of skimmed-milk cheese), for the Reichert number.

Detection of Skimmed-milk Cheese.-In a cream cheese the fat should greatly exceed the protein; in a whole-milk cheese the per cent of fat should at least equal that of the protein, and is generally in excess. If the fat is considerably less than the protein, the cheese has undoubtedly been made from skimmed milk. The analyses on page 199 illustrate these points.

PROTEIN PREPARATIONS.

Casein is the basis of a variety of preparations, some of which are intended for the use of invalids and people of weak digestion, and others, from their compactness, for travellers and campers. Among these foods are the following:

Nutrose. This is a caseinate of sodium formed by the action of the alkali upon dried casein. It is soluble in water.

Eucasin is a caseinate of ammonium, a soluble powder somewhat similar to nutrose.

Plasmon. This is a yellowish powder, prepared by treatment with sodium bicarbonate of the curd precipitated from skimmed milk. The compound is kneaded in an atmosphere of carbon dioxide, and reduced to a soluble powder.

The following analysis of plasmon was made by Woods and Merrill:*

*Maine Agric. Exp. Sta., Bul. 178, p. 101.