Chapter IV: Butter

=Butter Fat.=—The fat contained in suspension in the milk of mammals differs considerably in chemical composition from the body fat of the same animals. The latter consists mainly of glycerides of the higher fatty acids, notably stearic, palmitic, and oleic acids in varying proportions, whereas in the case of milk fats there is in addition to compounds of these fatty acids a large proportion of glycerides of volatile fatty acids.

It is upon this characteristic feature of butter fat that many of the methods of detecting foreign fats in butter are based.

A specimen of butter fat examined by Brown yielded 86·40 per cent. of insoluble fatty acids and 8·35 per cent. of soluble volatile fatty acids. The former included 32·85 per cent. of oleic acid, 1·83 per cent. of stearic acid, 38·61 per cent. of palmitic acid, 9·89 per cent. of myristic acid, and 2·57 per cent. of lauric acid. The soluble volatile acids were made up of 5·45 per cent. of butyric acid, 2·09 per cent. of caproic acid, 0·49 per cent. of caprylic acid, and 0·32 per cent. of capric acid.

The low proportion of stearic acid found was in accordance with the direct estimations made by Hehner and Mitchell, who obtained only small amounts in their examination of the fatty acids from a large number of samples of butter fat of different origin. It was interesting to note, however, that after the fatty acids had been exposed to the air for some weeks, there was apparently a gradual formation of stearic acid.

The general composition of butter fat, as shown by the examination of Brown (_supra_), is as follows:—

+----------------------+-----------+
| Glycerides of— | Per Cent. |
+----------------------+-----------+
|Dihydroxystearic acid | 1·04 |
|Oleic acid | 3·95 |
|Stearic ” | 1·91 |
|Palmitic ” | 40·51 |
|Myristic ” | 10·44 |
|Lauric ” | 2·73 |
|Capric ” | 0·34 |
|Caprylic ” | 0·53 |
|Caproic ” | 2·32 |
|Butyric ” | 6·23 |
+----------------------+-----------+

Naturally, different specimens of butter may show wide variations in the proportions of their different constituents, but the above may be taken as typical and as an illustration of the more complex character of milk fat than that of a body fat.

The glycerides in butter fat are probably present in the form of mixed glycerides, though, as yet, certain proof of this is wanting.

=Butter.=—Commercial butter consists of butter fat, water, casein, and salt, the proportion of fat usually ranging from about 84 to 87 per cent.

The following results obtained by Vieth (_Analyst_, 1891, xvi. 1) show the average composition to be expected:—

+---------------------------+---------+---------+---------+---------+
| | Fat, | Curd, | Salt, | Water, |
| Butter. |per cent.|per cent.|per cent.|per cent.|
+---------------------------+---------+---------+---------+---------+
|Danish (17 samples) | 83·41 | 1·30 | 1·87 | 13·42 |
|English (72 samples) | 86·85 | 0·59 | 1·02 | 11·54 |
|French (fresh, 108 samples)| 84·77 | 1·38 | 0·09 | 13·76 |
| ” (salted, 5 samples) | 84·34 | 1·60 | 2·01 | 12·05 |
|Kiel (4 samples) | 85·24 | 1·17 | 1·35 | 12·24 |
|Swedish (25 samples) | 83·89 | 1·33 | 2·03 | 13·75 |
+---------------------------+---------+---------+---------+---------+

_Water._—Statutory regulations are in force in many countries to limit the permissible amount of water in butter. In Great Britain and Ireland pure butter must not contain more than 16 per cent. of moisture, while the so-called “milk-blended butter” is not allowed to exceed the limit of 24 per cent.

The following limits are in force in other countries: _Belgium_, 18 per cent. unless declared; _Germany_, 18 per cent. for unsalted butter and 16 per cent. for salted butter; _United States_, 16 percent.; _Canada_, 16 percent.; _Queensland_, 16 per cent.; _Victoria_, 15 per cent.

These regulations have been found necessary, owing to the readiness with which a large excess of water (over 25 per cent.) may be churned into a butter without making it appear abnormally moist or interfering with its sale.

The effect of “salting” butter is to reduce the amount of moisture, as is shown in the results of Vieth, quoted in the table given above.

Richmond considers that the best proportion for a butter that is meant to keep well is 13·5 per cent.

_Salt._—There is no regulation as to the amount of salt that may be added to butter, and the proportion will depend upon the popular taste; sometimes as much as 10 per cent. may be found.

_Curd._—This term represents not only the solid nitrogenous substances derived from the casein of the milk, but also the milk sugar, etc., and its amount is roughly estimated by subtracting from 100 the percentages of butter fat, moisture, and ash which have been separately determined. The proportion thus found varies from a fraction of 1 per cent. to about 1·75 per cent., the average being a little over 1 per cent.

More accurate estimations of the true casein are obtained by determining the amount of nitrogen, and calculating the quantity of casein from the result.

=Keeping Properties of Butter.=—Rogers and Gray have studied the effect of the acidity of the cream upon the flavour of butter (U.S. Dept. Agricult., 1909, Bull. No. 119). They show that butter is liable to develop unpleasant flavours even when stored at temperatures as low as -10° F., and that the amount of alteration increases with the acidity of the cream from which the butter was prepared.

No micro-organisms to which the more rapid deterioration of butter from very acid cream could be attributed were detected, nor did the changes appear to be due to the action of enzymes.

When the butter had been made from sterilised cream acidified with various acids there was a gradual development of unpleasant flavours, and this result indicated that the acid normally produced in milk by the lactic acid bacteria had an influence in bringing about slow decomposition of unstable compounds in the butter.

Butter made from sweet sterilised cream was found to show much less tendency to change on storage than that made in the ordinary way, but the flavour would as a rule be regarded as too mild.

Flavours derived from the wood of the churn, or other external sources, would also be much more noticeable in the case of such butter than in that made from soured cream.

This lack of flavour has been shown by Storch to be due to absence of the products of certain bacteria which develop in the souring of the cream, and various species have been isolated, each imparting characteristic properties to the butter.

Thus Conn has demonstrated the practical advantages of inoculating the cream with cultures of specific bacteria, and has shown that in this way it is possible to give to the butter the particular flavours which previously had only been naturally produced at certain periods of the year. At the same time butter thus artificially ripened was found to keep better and to retain its flavour longer than that prepared by the older method. This method of artificial inoculation with specific organisms has been successfully employed on a large scale in America.

Burr and Wolff (_Milchwirtschaftl. Zentralbl._, 1910, vi. 241) have recently studied the effect of the parchment wrappings of butter upon the keeping qualities, and have found that under certain conditions parchment paper affords a suitable medium for the growth of moulds.

The main factors influencing this growth are a high percentage of moisture in the butter and the access of air. The presence of 1 to 1·5 per cent. of salt checks the growth, whereas unsalted butter is readily attacked.

A method of improving the keeping qualities of butter has been based upon the sterilising action of ultra-violet rays upon the water used for washing the butter in the dairy.

It has been pertinently pointed out by Dornic and Daire (_Comptes Rend._, 1909, 149, 355) that it avails little to prepare butter from sterilised cream if the water used for washing it contains micro-organisms that tend to produce rancidity.

The apparatus employed by Dornic and Daire consists of a tank lined with glass, and provided with glass partitions over which the water passes, and receives the rays from two quartz electric lamps, which are inserted through holes in the cover of the tank.

About 3000 litres of water can thus be rendered practically sterile in a day, the number of bacteria being reduced by the treatment to an insignificant quantity.

The cream, or the butter itself, might be rendered sterile by this process, but an unpleasant flavour produced by the action of the ozone formed by the lamps prevents the method being successfully employed in practice.

A French patent has, however, been taken out (No. 400,921 of 1909) for sterilising butter or other fats in this way. The substance is spread in a thin layer upon an endless band or revolving drum, the movement of which carries it past a series of lamps emitting ultra-violet rays.

=Rancidity of Butter.=—There appears to be little doubt that micro-organisms may play a part in the production of rancidity in butter, although the presence of light and air are probably the chief factors in the changes that take place.

Laxa, who studied the question with especial reference to butter (_Arch. Hyg._, 1902, xli. 119), found that several mould fungi, and at least one bacillus (_B. fluorescens liquefaciens_), were capable of growing upon media containing butter fat and of breaking up the glycerides, with the liberation of free fatty acids, which were then in turn decomposed.

The acidity of the fat increases with the progress of the rancidity, but not necessarily proportionately, and a specimen of butter fat in which the oxidation product to which the rancid taste and odour are due may have a lower acid value than butter, which shows no trace of rancidity.

Many of the compounds produced in the development of rancidity are soluble, and may be separated from the fat by washing it with water. Advantage is taken of this fact in the preparation of “process” butter from stale or unsaleable genuine butter.

There is no definite test that can be applied to detect rancidity in butter, and reliance must therefore be mainly placed upon the taste.

=Renovated Butter.=—A product sold in large quantities under the name of renovated or “process” butter is prepared by melting down the fat of old butter, separating it from the curd and water, chilling it upon ice, and re-churning it with fresh milk and whey. By this means the casein is restored to the butter, and a mixture is obtained which answers to the chemical tests for fresh butter, but lacks its fresh flavour.

Several methods are employed for distinguishing process butter from the genuine product, the most satisfactory of which are based upon the physical alteration of the butter fat in the processes of melting and congealing. Thus in the case of ordinary butter the separated fat does not show the crystalline structure of the fat from renovated butter when examined under the microscope, and the difference is still more pronounced when the fats are compared in polarised light.

Other tests upon which a judgment may be based are the greater solubility of the “process” fat in glacial acetic acid (Cochrane); and the appearance and behaviour of the separated curd, which is gelatinous when derived from fresh butter, and granular and flocculent when obtained from renovated butter.

Hess and Doolittle (_J. Amer. Chem. Soc._, 1900, xxii. 150) also observed a difference in the behaviour of the two products when heated, genuine butter foaming, while process butter splutters in the same way as margarine, from which, however, it may be distinguished by other tests.

The trade in process butter is much more extensive in the United States than in this country, and numerous patents have been taken out to obtain a product that shall imitate ordinary butter still more closely.

Among the most recent of these is the process claimed by Roos (U.S. Pat. 854,383, 1907). The fat separated from the old butter is melted and heated to a temperature of 108° to 110° F., and is then termed “butter oil.”

About five parts of it are added to 3 parts of a previously churned mixture of acidified skimmed milk and fresh whole milk, and the whole stirred up until an emulsion is obtained. This is treated with cold water to cause crystallisation, and the resulting crystals are mixed with salt and exposed to the air for some time at the ordinary temperature, after which the mass is thoroughly kneaded to expel the salt and the excess of milk.

=Preservatives in Butter.=—Butter that at one time was heavily salted is now frequently preserved by the addition of a small amount of borax or other compound of boric acid.

The prevalence of the practice was shown by the evidence given by leading representatives of the butter industry before the Departmental Committee on Preservatives in Food (1901). One of these witnesses informed the Committee that his firm had ceased to use heat sterilisation of the cream, as it was found to impair the flavour of the butter.

It was further shown that the butter imported from Australia, Normandy, etc., contained 0·5 per cent. of boric acid; and that this amount was regarded by the trade as sufficient for the purpose.

In their report the Committee recommended that no preservative other than boric acid or borax, or mixtures of the two, should be permitted to be used in butter or margarine, and that the proportion should not exceed ½ per cent.

An attempt was made in France to prohibit absolutely the use of preservatives in butter, but the needs of the trade were too strong, and, as was mentioned above, the butter imported into this country from France is almost invariably preserved.

The presence of boron compounds is proved by melting the butter, separating the aqueous portion, rendering it faintly acid with hydrochloric acid, and immersing a strip of turmeric paper therein. On drying the paper at a gentle heat a purple-red coloration, changing to bluish black on the addition of ammonia solution, indicates the presence of boric acid.

Tests are sometimes applied for other preservatives, such as salicylic acid, sodium benzoate, sodium fluoride, and formalin, but these compounds are not of common occurrence in butter.

Some years ago butter intended for export to tropical countries was preserved by the addition of a fairly large proportion of glucose. This practice was most common in France.

=Physical Characteristics.=—Butter fat has a specific gravity ranging from about 0·907 to about 0·913 at 35° C., whereas the specific gravity of the animal fats used as adulterants is considerably higher.

It has been shown by Skalweit (_J. Soc. Chem. Ind._, 1894, xiii. 54) that these differences are most pronounced at a temperature of 35° C., as is illustrated by the following table:—

+-------------+---------+------------+-------------+
|Temperature, | | | |
| °C. | Lard. | Margarine. | Butter Fat. |
+-------------+---------+------------+-------------+
| 35 | 0·9019 | 0·9017 | 0·9121 |
| 50 | 0·8923 | 0·8921 | 0·9017 |
| 60 | 0·8859 | 0·8857 | 0·8948 |
| 70 | 0·8795 | 0·8793 | 0·8879 |
| 80 | 0·8731 | 0·8729 | 0·8810 |
| 90 | 0·8668 | 0·8665 | 0·8741 |
| 100 | 0·8605 | 0·8601 | 0·8672 |
+-------------+---------+------------+-------------+

The specific gravity of cocoanut oil is also greater than that of animal fats.

_Solubility._—Butter fat is considerably more soluble than animal body fats in various solvents, such as glacial acetic acid, alcohol, etc., and on this property have been based several rapid “sorting” tests for distinguishing between pure and adulterated butter.

One of these tests, known as the _Valenta test_, gives the temperature at which a solution of a definite quantity of the fat in a definite quantity of hot glacial acetic acid becomes turbid on cooling the liquid.

The following figures obtained by Chattaway, Pearmain, and Moor, show the difference between butter and margarine:—

+-----------+----------+----------+-------+
| | Maximum. | Minimum. | Mean. |
+-----------+----------+----------+-------+
|Butter fat | 39·0 | 29·0 | 36·0 |
|Margarine | 97·0 | 94·0 | 95·5 |
+-----------+----------+----------+-------+

The natural variation in this respect between different samples of genuine butter is thus too great to permit of the detection of small quantities of margarine in butter.

_Refractometric Examination._—The difference in the refractive power of butter fat and margarine is a valuable means of obtaining a preliminary idea as to the purity of a sample of butter, and several instruments have been specially constructed for the purpose. The most widely used of these is termed a _butyro-refractometer_, and it is readily possible by its means to distinguish in a few minutes between pure and grossly adulterated butter. Here again, however, the variations in the readings of samples of genuine butter of different origin are sometimes greater than between a sample of pure butter and one adulterated with a small percentage of margarine. Thus Crismer found that two pure samples of butter fat gave readings of 45·8 and 46, while two adulterated samples gave readings of 45·2 and 45·6 at 40° C.

=Chemical Characteristics.=—Owing to the large proportion of fatty acids of low molecular equivalent present in butter fat, the saponification value is naturally very high, ranging from about 219 to 240, though values both above and below these figures have been recorded for genuine butter.

Since cocoanut oil also has a high saponification value (254 to 260), mixtures of animal fat with that fat may readily be prepared which will give figures within the normal limits for butter fat.

The proportion of free fatty acids in freshly churned butter fat is very small (about 0·01 grm. per kilo. according to Duclaux), but it gradually increases as the butter is kept. A determination of the acid value may therefore afford some indication of the age of a sample of butter.

The iodine value of pure butter fat is usually between 25 and 40, but is liable to fluctuate considerably with the nature of the food given to the cows.

Of all the different so-called constants, the Hehner and Reichert values described below are the most important as tests of purity, and an official method of determining the latter value has been established in this country, so as to eliminate errors due to variations in the size of the apparatus, duration of distillation, etc.

_Hehner and Reichert Values._—The chief substances used in the adulteration of butter are beef fat and pig’s fat (in the form of margarine), vegetable oils (also mainly in margarine), and (especially of late) cocoanut oil.

The methods of distinguishing between butter fat and other animal fats are based upon the high proportion of soluble volatile fatty acids in the former, and their practical absence from the latter.

Originally the test was devised by Hehner, who measured the amounts of insoluble fatty acids yielded by butter fat, and showed that they were very much less than those given by other fats.

It is now more usual, however, to employ the more rapid process devised by Reichert (with modifications by other chemists), in which the fatty acids are distilled under definite conditions which must be rigidly followed, and the proportion of them thus obtained is estimated by neutralisation with a standard solution of alkali. The details of the method are described at length in Chapter VIII. The result is then expressed in the number of cubic centimetres of this standard alkali, and is termed the Reichert (or Reichert-Meissl) value of the fat.

The Reichert-Meissl value of pure butter fat has been found to vary under ordinary conditions from about 20 to 33, though figures far beyond either of these limits have been recorded in exceptional cases.

In calculating the amount of foreign fat in a butter from this value it is assumed as an arbitrary figure that an average butter fat has a Reichert-Meissl value of 28·78, and the percentage of added fat may thus be found approximately.

The difficulty, however, is that the values normally range so widely above and below this average, that it is quite possible to add 10 (or more) per cent. of foreign fat to a butter with a high Reichert-Meissl value, and still have a product giving a normal value.

On the other hand, cows frequently produce butter which gives a value far below the standard figure, and without special knowledge of the conditions to which this abnormality is due the butter may be condemned as adulterated with foreign fat.

This was shown in a striking manner a few years ago, when large quantities of butter imported into this country from Holland were condemned as adulterated on account of their low Reichert-Meissl values.

The causes of this abnormality were investigated by van Rijn, who found that the mixed butter of a herd of cows might vary in its Reichert-Meissl value from 17·0 to 32·1.

The low figures were found to coincide with the end of the pasturage season, for after the cows had been stalled for a short time the butter became normal again. For example, the butter from a herd of seven cows had a Reichert-Meissl value of 24·4 on September 11, which had fallen to 19·0 on October 23. The animals were then taken from the fields, and their butter gave the following values: November 6, 21·5; November 20, 23·1; and December 11, 25·4.

This conclusion as to the effect of leaving the cows too late in the fields was borne out by the results of a Commission appointed to investigate the causes of abnormality of the butter produced in Belgium.

It was found that the butter giving such results was generally the product of small herds of cows; that its occurrence was most pronounced in the last four months of the year; and that with the return of spring the values became normal again.

It may be mentioned that in Belgium the sale of butter showing a lower Reichert-Meissl figure than 28 is prohibited by law, while in the United States the minimum value is fixed at 24.

Poor feeding has also an influence in lowering the Reichert-Meissl value, and to this cause must be attributed the low values frequently found in the case of Siberian butter.

Another factor influencing the degree of the Reichert-Meissl value is the time that has elapsed since calving. Thus it was shown by Kreit (_Analyst_, 1893, xviii. 134) that the values obtained at an early period of lactation were invariably higher, and that they then gradually fell. In some cases the butter from the milk of cows that had recently calved reached the high Reichert-Meissl value of 34·4.

This also affords the explanation of the occasionally low Reichert-Meissl values of Irish butter, which, as has been proved by Ball (_Analyst_, 1907, xxxii. 202), coincides with a period when the milk has been derived from cows at the very end of lactation. This period lasts for about six weeks, during which time the milk is richest in fat, while the fat shows the lowest Reichert-Meissl values. Thus the following results were obtained with butters churned on December 19, 1906:—

+---------------+---------+-------+-------+--------+----------+
| |Limerick.|Bruree.|Mallow.|Clonmel.|Tipperary.|
+---------------+---------+-------+-------+--------+----------+
|Reichert-Meissl| | | | | |
| values | 22·7 | 21·5 | 23·3 | 23·5 | 22·1 |
+---------------+---------+-------+-------+--------+----------+

The same abnormality does not occur in England, where the calving of the cows is distributed over the whole year, instead of, as in Ireland, taking place within six weeks of one another.

It is also possible that sufficient attention may not be given to the cows in Ireland, since the effect of feeding and good housing is to raise the Reichert-Meissl value of the butter, even in the case of cows at the end of their lactation period.

=Influence of the Food of the Cows.=—The results of numerous feeding experiments, in which quantities of different oils and fats were mixed with the daily fodder of the cows, have shown that the nature of the butter fat may be appreciably affected in this way. Thus Werenskiold (_Chem. Zentralbl._, 1900, ii. 215) proved that cotton-seed oil could be detected in the fat from the milk of cows which had taken a small amount of cotton-seed oil cake with their food, and this result was in agreement with the results of experiments carried out for the Board of Agriculture in this country (_Analyst_, 1898, xxiii. 255).

As a rule, however, the colour reaction indicating the presence of cotton-seed oil is very slight, and does not correspond to a proportion of more than 1 per cent.

In the corresponding experiments of the Agricultural Board with sesame oil cake the butter from the cows did not give the characteristic colour reaction for that oil, even after they had been fed upon it for several weeks.

There is also evidence to show that the characteristic fatty acid of arachis oil does not pass into the milk of cows fed upon arachis-seed cake; but it has been shown by Paal and Amberger (_Zeit. Unters. Nahr. Genussm._, 1909, xvii. 1) that feeding the animals upon copra may have some influence upon the composition of their milk fat, and increase the proportion of insoluble volatile fatty acids.

Opinion is divided as to the advisability of having a fixed standard for a chemical property such as the Reichert value. Although such a limit as is fixed in Belgium excludes all doubt in condemning samples that fall below it, yet, as was found in the case of lard (p. 64), a fixed standard leads to an increase in the amount of petty adulteration, since it is not difficult for a skilful mixture to be made that will answer the requirements of a moderate standard.

In any case there is even now a systematic attempt to conform with what the adulterator presumes will be the standard by which the analyst will judge his product, and the present writer has frequently examined samples with a Reichert-Meissl value of 23-24 which in all probability contained a small quantity of foreign fat, but yet might conceivably have been genuine butter containing less than the average quantity of volatile fatty acids.

Nor does it follow that such butter is necessarily inferior as a food, for there is no proof that a slight deficiency in the amount of the glycerides of volatile fatty acids affects the nutritive value of the whole fat.

=Cocoanut Oil in Butter.=—The addition of cocoanut oil to butter, which has become increasingly prevalent of late years, has the effect of lowering the Reichert-Meissl value to a much smaller extent than the addition of animal fat.

This is due to the fact that cocoanut oil has itself a Reichert-Meissl value of 7-8, indicating the presence of a considerable amount of soluble volatile fatty acids, and it is therefore possible to prepare a mixture of animal fat, cocoanut oil, and fluid vegetable oil, which can be added to butter in a fairly large proportion without reducing the Reichert-Meissl value of the latter below 24 or 25.

The problem of detecting cocoanut oil in butter has therefore received much attention of late, and numerous methods have been devised for estimating the amount of such addition.

Speaking generally, these are based upon the fact that cocoanut oil contains a high percentage of lauric acid (up to 60 per cent.), whereas that acid is only present in very small proportion in butter fat.

Now, since lauric acid will volatilise in a current of steam, but, unlike the lower fatty acids (butyric, caproic, caprylic, and capric acids), is not soluble in water, it is possible to obtain a measure of its proportion by continuing the distillation as in the Reichert-Meissl process until the whole of the volatile fatty acids (soluble and insoluble) have passed over. The distillate, when filtered, is separated into a soluble and insoluble portion, and the latter may be dissolved in alcohol and its acidity determined by titration with standard alkali solution.

A method whereby the process of distillation is greatly accelerated is described in Chapter VIII.

Other methods of detecting cocoanut oil have been based upon the quantitative separation of the lauric acid in the form of various metallic salts, such as barium, cadmium, etc., so as to obtain “barium values,” “cadmium values,” etc., which will increase with the proportion of cocoanut oil in the butter.

These methods, however, are more complicated and no more effective than the method of extended distillation, although they are of value as affording confirmatory evidence of the adulteration (see p. 116).

=Artificial Colouring Matters.=—The nature of the food given to the cows, their breed, and the season of the year, all have an influence upon the colour of the butter; that produced in summer, for instance, being more yellow than that produced later in the year.

When exposed to the action of air and light the natural colour of butter gradually fades, and ultimately the fat becomes colourless.

The popular demand for a butter of pronounced yellow colour—a demand inspired by the belief that intensity of colour indicates purity—has led to the artificial colouring of pale butter.

The colouring matters employed for this purpose include that of the carrot, annatto, turmeric, saffron, marigold, and various aniline dye-stuffs.

Special azo dye-stuffs, soluble in oil, are frequently used, especially in the United States. They may be detected by mixing a little fuller’s earth with the butter-fat which has been separated from the curd, the earth assuming a pink or light red coloration in the presence of such dyes.

Since in the United States the sale of margarine containing artificial colouring matters is prohibited, the desired yellow colour is now frequently obtained by the use of palm oil, mustard-seed oil, and similar fats of intense colour. Special tests have therefore been devised for the detection of these fats in butter.