296. Abbe’s Refractometer.—For practical use the small instrument invented by Abbe will be found sufficient. The one which has been in use for many years in this laboratory is shown in Fig. 86. The illustration represents the apparatus in the position preliminary to reading the index. In preparing the sample of oil for observation the instrument is turned on its axis until the prisms between which the oil is placed assume a horizontal position, as is seen in Fig. 87. The movable prism is unfastened and laid aside, the fixed prism covered with a rectangular shaped piece of tissue paper on which one or two drops of the oil are placed. The movable prism is replaced in such a manner as to secure a complete separation of the two prisms by the film of oiled tissue paper. A little practice will enable the analyst to secure this result.

After the paper disk holding the fat is secured by replacing the upper prism, the apparatus is placed in its normal position and the index moved until the light directed through the apparatus by the mirror shows the field of vision divided into dark and light portions. The dispersion apparatus is now turned until the rainbow colors on the part between the dark and light fields have disappeared. Before doing this, however, the telescope, the eyepiece of the apparatus, is so adjusted as to bring the cross lines of the field of vision distinctly into focus. The index of the apparatus is now moved back and forth until the line of the two fields of vision falls exactly at the intersection of the cross lines. The refractive index of the fat under examination is then read directly upon the scale by means of a small magnifying glass. To check the accuracy of the first reading, the dispersion apparatus should be turned through an angle of 180° until the colors have again disappeared, and, after adjustment, the scale of the instrument again read. These two readings should nearly coincide, and their mean is the true reading of the fat under examination.

For butter fats the apparatus should be kept in a warm place, the temperature of which does not fall below 30°. For reducing the results obtained to a standard temperature, say 25°, the factor 0.000176 may be used. As the temperature rises the refractive index falls.

Example.—Refractive index of a butter fat determined at 32°.4 = 1.4540, reduced to 25° as follows: 32.4 -25 = 7.4; 0.000176 × 7.4 = 0.0013; then 1.4540 + 0.0013 = 1.4553.

The instrument used should be set with distilled water at 18°, the theoretical refractive index of water at that temperature being 1.333. In the determination above given, the refractive index of pure water measured 1.3300; hence the above numbers should be corrected for theory by the addition of 0.0030, making the corrected index of the butter fat mentioned at the temperature given, 1.4583.

297. Pulfrich’s Refractometer.—For exact scientific measurements, Pulfrich’s apparatus has given here entire satisfaction. In this instrument a larger quantity of the oil is required than for the abbe, and this quantity is held in a cylindrical glass vessel luted to the top of the prism. The method of accomplishing this and also an illustration of the refraction of the rays of light are shown in Fig. 88.

The angle i is measured by a divided circle read with the aid of a small telescope. The index of the prism of highly refractive glass N is known. The oil is seen at n. The light used is the yellow sodium ray (D). From the observed angle the refractive index of n is calculated from the formula

For convenience the values of n for all usual values of i are computed once for all and arranged for use in tabular form. The latest model of Pulfrich’s apparatus, arranged both for liquid and solid bodies, and also for spectrometric observation is shown in Fig. 89.

When the sodium light is used it is placed behind the apparatus and the light is collected and reflected on the refractive prism by the lens N. Through H and G is secured the micrometric reading of the angle on the scale D by means of the telescopic arrangement F E. For regulating the temperature of the oil and adjacent parts, a stream of water at any desired temperature is made to circulate through L and S in the direction indicated by the arrows. The manner in which this is accomplished is shown in the cross section of that part of the apparatus as indicated in Fig. 90.

For further details of the construction and operation of the apparatus the original description may be consulted.[248]

In case a spectrometric observation is desired the H ray, for instance, is produced by the geissler tube Q, Fig. 91. The light is concentrated and thrown upon the refractive prism by the lens P, the lens N, Fig. 89, being removed for this purpose.

Tables, for correcting the dispersion and for calculating the indices for each angle and fraction thereof, and for corrections peculiar to the apparatus, accompany each instrument.

298. Refractive Indices of some Common Oils.—The following numbers show the refractive indices obtained by Long for some of the more common oils. The light used was the yellow ray of the sodium flame.[249]

In case of the use of Abbe’s apparatus, in which diffused sunlight is the source of the illumination, the numbers obtained cannot be compared directly with those just given unless the apparatus be first so adjusted as to read with distilled water at 18°, 1.333. In this case the reading of the scale gives the index as determined by the yellow ray. The numbers obtained with Abbe’s instrument for some common oils are given below.[250]

In the determinations the instrument was set with water at 18°, reading 1.3300, and they were corrected by adding 0.0030 in order to compensate for the error of the apparatus.

299. Oleorefractometer.—Instead of measuring the angular value of the refractive power of an oil it may be compared with some standard on a purely arbitrary scale. Such an apparatus is illustrated by the oleorefractometer of Amagat-Jean, or by Zeiss’s butyrorefractometer.

In the first named instrument, Fig. 92, the oil to be examined is compared directly with another typical oil and the shadow produced by the difference in refraction is located on a scale read by a telescope and graduated for two different temperatures.[251] The internal structure of the apparatus is shown in Fig. 93.

In the center of the apparatus a metal cylinder, A, is found carrying two plate glass pieces, C B, so placed as to form an angle of 107°. This cylinder is placed in a larger one, provided with two circular glass windows. To these two openings are fixed to the right and left, the telescopic attachments, G, V, S, E, and the apparatus M, H, Sʹ, Eʹ, for rendering the rays of light parallel. The field of vision is divided into two portions, light and dark, by a semicircular stop inserted in the collimator, and contains the double scale shown in the figure placed at H. The field of vision is illuminated by a gas or oil lamp placed at a convenient distance from the collimator. The inner metallic cylinder A is surrounded with an outer one, to which the optical parts are attached at D Dʹ by means of plane glass plates. This cylinder is in turn contained in the large water cylinder P P, carrying a thermometer in the opening shown at the top on the left. The manipulation of the apparatus is very simple. The outer cylinder is filled with water, at a temperature below 22°, the middle one with the typical oil furnished with the instrument, the cover of the apparatus carrying the thermometer placed in position and the cup-shaped funnel inserted in the cylinder A, which is at first also filled with the typical oil. The whole system is next brought slowly to the temperature of 22° by means of the lamp shown in Fig. 92. The telescope is adjusted to bring the scale of the field of vision into focus and the line dividing the light and shadow of the field should fall exactly on 0°a. If this be not the case the 0° is adjusted by screws provided for that purpose until it is in proper position. The typical oil is withdrawn from A by the cock R, the cylinder washed with a little of the oil to be examined and then filled therewith. On again observing the field of vision the line separating the shadow from the light will be found moved to the right or left, if the oil have an index different from that of the typical oil. The position of the dividing line is read on the scale.

For fats the temperature of the apparatus is brought exactly to 45° and the scale 0°b is used. In other respects the manipulation for the fats is exactly that described for oils. In the use of 0°a, in case the room be warmer than 22°, all the liquids employed should be cooled below 22° before being placed in the apparatus. It is then only necessary to wait until the room temperature warms the system to 22°. In the case of fats it is advisable to heat all the liquids to about 50° and allow them to cool to 45° instead of heating them to that temperature by means of the lamp.

One grave objection to this instrument is found in the absence of the proper scientific spirit controlling its manufacture and sale, as evidenced by the attempt to preserve the secret of the composition of the typical oil and the negligence in testing the scale of the instruments which will be pointed out further along.

According to Jean[252] the common oils, when purified, give the following readings at 22°:

In this instrument, therefore, vegetable and fish oils, as a rule, show a right hand, and animal fats a left hand deviation.

The oleorefractometer has been extensively used in this laboratory and the data obtained thereby have been found useful. We have not found, however, the values fixed by Jean to be constant. The numbers for lard have varied from -3.0 to -10.0, and other fats have shown almost as wide a variation from the values assigned by him.

Jean states that the number for lard, determined by the oleorefractometer, is -12, and he gives a definite number for each of the common oils and fats. On trying the pure lards of known origin in this instrument, I have never yet found one that showed a deviation of -12 divisions of the scale; but I have no doubt that there are many such lards in existence. The pure normal lards derived from the fat of a single animal would naturally show greater variations in their chemical and physical properties, than a typical lard derived from the mixed fats of a great many animals. In leaf lard, rendered in the laboratory, the reading of the oleorefractometer was found to be -10°, while with the intestinal lard it was -9°. On the other hand, a lard rendered from the fat from the back of the animal showed a reading of only -3°, and a typical cottonseed oil a reading of +12°. According to the statement of Jean, a lard which gives even as low a refractive number as -9, by his instrument, would be adjudged at least one-quarter cottonseed oil.

After a thorough trial of the instrument of Jean, I am convinced that it is of great diagnostic value, but if used in the arbitrary manner indicated by the author it would lead to endless error and confusion. In other words, this instrument is of greater value in analyses than Abbe’s ordinary refractometer, because it gives a wider expansion in the limits of the field of vision, and therefore can be more accurately read, but it is far from affording a certain means of discovering traces of adulteration with other fats.

300. Variations in the Instruments.—In the use of the oleorefractometer, attention should be called to the fact that, through some negligence in manufacture, the instruments do not give, in all instances, the same reading with the same substance. Allen obtained the following data with a sample of lard examined in three instruments, viz., 4°.5, 6°, and 11°. Such wide differences in the scales of the instruments cannot fail to disparage the value of comparative determinations.

The variations in samples of known origin, when read on the same instrument, however, will show the range of error to which the determinations made with the oleorefractometer are subject. Pearmain has tabulated a large number of observations of this kind, covering 240 samples of oils.[253]

Following are the data relating to the most important oils.

301. Butyrorefractometer.—Another instrument graduated on an arbitrary scale is the butyrorefractometer of Zeiss. This apparatus, which resembles in some respects the instrument of Abbe, differs therefrom essentially in dispensing with the revolving prisms of Amici, whereby the chromatic fringing due to dispersion is corrected, and on having the scale fixed for one substance, in this instance, pure butter fat. The form of the instrument is shown in Fig. 94. The achromatization for the butter fat is secured in the prisms between which a film of the fat is placed, as in the Abbe instrument. When a fat, differing from that for which the instrument is graduated is introduced, the fringes of the dark and light portions of the field will not only be colored (difference in dispersion), but the line of separation will also be displaced (difference in refractive power). The apparatus is therefore used in the differential determination of these two properties. It must not be forgotten, however, that butter fats differ so much in these properties among themselves as to make possible the condemnation of a pure as an adulterated sample.

302. Method of Charging the Apparatus.—The prism casing of the instrument is opened by turning the pin F to the right and pushing the half B of the prism casing aside. The prism and its appendages must be cleaned with the greatest care, the best means for this purpose being soft clean linen moistened with a little alcohol or ether.

Melt the sample of butter in a spoon and pour it upon a small paper filter held between the fingers and apply the first two or three drops of clear butter fat so obtained to the surface of the prism contained in prism casing B. For this purpose the apparatus should be raised with the left hand so as to place the prism surface in a horizontal position.

Press B against A and replace F by turning it in the opposite direction into its original position; thereby B is prevented from falling back and both prism surfaces are kept in close contact.

303. Method of Observation.—While looking into the telescope, give the mirror J such a position as to render the critical line which separates the bright left part of the field from the dark right part distinctly visible. It may also be necessary to move or turn the instrument about a little. First it will be necessary to ascertain whether the space between the prism surfaces be uniformly filled with butter, for, if not, the critical line will not be distinct.

By allowing a current of water of constant temperature to flow through the apparatus, some time previous to the taking of the reading, the at first somewhat hazy critical line approaches in a short time, generally after a minute, a fixed position and quickly attains its greatest distinctness. When this point has been reached note the appearance of the critical line (i. e., whether colorless or colored and in the latter case of what color); also note the position of the critical line on the centesimal scale, which admits of the tenth divisions being conveniently estimated, and at the same time read the thermometer. By making an extended series of successive readings and by employing an assistant for melting and preparing the small samples of butter, from twenty-five to thirty refractometric butter tests may, after a little practice, be made in an hour.

The readings of the refractive indices of a large number of butter samples made at 25° are, by means of a table which will be found below, directly reduced to scale divisions and yield the following equivalents:[254]

Whenever, in the refractometric examination of butter at a temperature of 25°, higher values than 54.0 are found for the critical lines these samples will, according to Wollny, by chemical analysis, always be found to be adulterated; but in all samples in which the value for the position of the critical line does not fall below 52.5, chemical analysis maybe dispensed with and the samples may be pronounced to be pure butter.

In calculating the position of the critical line for other temperatures than 25° allow for 1° variation of temperature a mean value of 0.55 scale division. The following table, which has been compiled in this manner, shows the values corresponding to various temperatures, each value being the upper limit of scale divisions admissible in pure butter:

If, therefore, at any temperature between 45° and 25° values be found for the critical line, which are less than the values corresponding to the same temperature according to the table, the sample of butter may safely be pronounced to be natural, i. e., unadulterated butter. If the reading show higher numbers for the critical line the sample should be reserved for chemical analysis. A special thermometer for use in the examination of butter will be described in the section devoted to dairy products.

304. Range of Application of the Butyrorefractometer.—The extended range of the ocular scale of the refractometer, n = 1.42 to 1.49, which embraces the refractive indices of the majority of oils and fats, renders the instrument applicable for testing oils and fats and also for examining glycerol.

By reference to the subjoined table the scale divisions may be transformed into terms of refractive indices. It gives the refractive indices for yellow light for every ten divisions of the scale. The differential column Δ gives the change of the refractive indices in terms of the fourth decimal per scale division. Owing to the accuracy with which the readings can be taken (0.1 scale division) the error of the value of n rarely exceeds one unit of the fourth decimal of n.

Table of Refractive Indices.

The process of observation is precisely the same as that already described. In cases, however, where the critical line presents very broad fringes (turpentine, linseed oil, etc.) it is advisable to repeat the reading with the aid of a sodium flame.

305. Viscosity.—An important property of an oil, especially when its lubricating qualities are considered, is the measure of the friction which the particles exert on other bodies and among themselves, in other words, its viscosity. In the measure of this property no definite element can be considered, but the analyst must be content with comparing the given sample with the properties of some other liquid regarded as a standard. The usual method of procedure consists in determining the time required for equal volumes of the two liquids to pass through an orifice of given dimensions, under identical conditions of temperature and pressure. In many instances the viscosity of oils is determined by comparing them with water or rape oil, while, in other cases, a solution of sugar is employed as the standard of measurement.

In case rape oil be taken as a standard and its viscosity represented by 100 the number representing the viscosity of any other oil may be found by multiplying the number of seconds required for the outflow of fifty cubic centimeters by 100 and dividing by 535. If the specific gravity vary from that of rape oil, viz., 0.915, at 15°, a correction must be made by multiplying the result obtained above by the specific gravity of the sample and dividing the product by 0.915. If n be the observed time of outflow in seconds and s the specific gravity the viscosity is expressed as follows:[255]

It is important that the height of the oil in the cylinders from which it is delivered be kept constant, and this is secured by supplying additional quantities, on the principle of the mariotte bottle.

306. The Torsion Viscosimeter.—In this laboratory the torsion viscosimeter, based on the principle described by Babcock is used. The instrument employed is the one described by Doolittle.[256] The construction of the apparatus is illustrated in Fig. 95.

A steel wire is suspended from a firm support and fastened to a stem which passes through a graduated horizontal disk, thus permitting the accurate measurement of the torsion of the wire. The disk is adjusted so that the index point reads exactly 0, thus showing that there is no torsion in the wire. A brass cylinder seven centimeters long by five in diameter, having a slender stem by which to suspend it, is immersed in the oil and fastened by a thumbscrew to the lower part of the stem of the disk. The oil cup is surrounded by a bath of water or high fire-test oil, according to the temperature at which it is desired to determine the viscosity. This temperature obtained, while the disk is resting on its supports, the wire is twisted 360° by rotating the milled head at the top. The disk being released, the cylinder rotates in the oil by virtue of the torsion of the wire.

The action now observed is identical with that of the simple pendulum.

If there were no resistance to be overcome, the disk would return to 0, and the momentum thus acquired would carry it 360° in the opposite direction. But the resistance of the oil to the rotation of the cylinder causes the revolution to fall short of 360°, and the greater the viscosity of the oil the greater will be the resistance, and also the retardation. This retardation is found to be a very delicate measure of the viscosity of the oil.

This retardation may be read in a number of ways, but the simplest is to read directly the number of degrees of retardation between the first and second complete arcs covered by the rotating pendulum. For example, suppose the wire be twisted 360° and the disk released so that rotation begins. In order to obtain an absolute reading to start from, which shall be independent of any slight error in adjustment, ignore the starting point and make the first reading of the index at the end of the first swing. The disk is allowed to complete a vibration and the needle is read again at its nearest approach to the first point read. The difference in the two readings will measure the retardation due to the viscosity of the liquid. In order to eliminate errors duplicate determinations are made, the milled head being rotated in an opposite direction in the second one. The mean of the two readings will represent the true retardation. Each instrument is standardized in a solution of pure cane sugar, as proposed by Babcock, and the viscosity, in each case, is a number representing the number of grams of sugar in 100 cubic centimeters, which, at 22°, would produce the retardation noted.

Each instrument is accompanied by a table which contains the necessary corrections for it and the number expressing the viscosity, corresponding to the different degrees of retardation, as read on the index. The following numbers, representing the viscosity of some oils as determined by the method of Doolittle, were obtained by Krug.[257]

307. Microscopic Appearance.—When fats are allowed to slowly crystallize from an ethereal solution they may afford crystalline forms, which, when examined with a magnifying glass, yield valuable indications of the nature and origin of the substance under examination.[258]

The method of securing fat crystals for microscopic examination, which has been used in this laboratory, is as follows: From two to five grams of the fat are placed in a test tube and dissolved in from ten to twenty cubic centimeters of ether. The tube is loosely stoppered with cotton and allowed to stand, for fifteen hours or longer, in a moderately warm room where no sudden changes of temperature are likely to take place. It is advisable to prepare several solutions of the same substance with varying properties of solvent, for it is not possible to secure in a given instance those conditions which produce the most characteristic crystals. The rate and time of the crystallization should be such that the microscopic examination can take place when only a small portion of the fat has separated in a crystalline condition. A drop of the mass containing the crystals is removed by means of a pipette, placed on a slide, a drop of cotton or olive oil added, a cover glass gently pressed down on the mixture and the preparation subjected to microscopic examination. Several slides should be prepared from the same or different crystallizations. Sometimes the results of an examination made in this way are very definite, but the analyst must be warned not to expect definite data in all cases. Often the microscopic investigations result in the production of negative or misleading observations, and, at best, this method of procedure must be regarded only as helpful and confirmatory.

A modification of the method of preparation described above has been suggested by Gladding.[259] About five grams of the melted fat are placed in a small erlenmeyer, dissolved in a mixture of ten cubic centimeters of absolute alcohol mixed with half that quantity of ether. The flask is stoppered with a plug of cotton and allowed to stand in a cool place for about half an hour. By this treatment the more easily crystallizable portions of the fat separate in a crystalline form, while the triolein and its nearly related glycerids remain in solution. The crystalline product is separated by filtration through paper wet with alcohol and washed once with the solvent mentioned above. After drying in the air for some time the crystals are removed from the paper and dissolved in twenty-five cubic centimeters of ether, the cotton plug inserted, and the erlenmeyer placed, in a standing position, in a large beaker containing water. The water jacket prevents any sudden changes of temperature and affords an opportunity for the uniform evaporation of the ether which should continue for fifteen hours or longer in a cool place.

Other solvents, viz., alcohol, chloroform, carbon disulfid, carbon tetrachlorid, petroleum and petroleum ether have been extensively used in the preparation of fat crystals for microscopic examination, but in our experience none of these is equal to ether when used as already described.

308. Microscopic Appearance of Crystals of Fats.—For an extended study and illustration of the characteristics of fat crystals the bulletin of the Division of Chemistry, already cited, may be consulted. In the case of lard, there is a tendency, more or less pronounced, to form prismatic crystals with rhombic ends. Beef fat on the other hand shows a tendency to form fan-shaped crystals in which the radii are often curved.

Typical crystals of swine and beef fat are shown in the accompanying figures, 96 and 97.[260] In mixtures of swine and beef fats the typical crystals are not always developed, but in most cases the fan-shaped crystals of the beef fat will appear more or less modified when that fat forms twenty per cent or more of the mixture. When only five or ten per cent of the beef fat on the one hand or a like amount of swine fat on the other are present the expectation of developing any characteristic crystals of the minimum constituent is not likely to be realized.

The typical crystals of lard are thought by some experts to be palmitin and those of beef fat stearin, but no direct evidence has been adduced in support of these a priori theories.

In the experience of this laboratory, as described by Crampton,[261] the differences between the typical crystallization of beef and swine fats are plainly shown. In mixed fats, on the contrary, confusing observations are often made. In a mixture of ten per cent of beef and ninety per cent of swine fats a uniform kind of crystallization is observed, not distinctly typical, but the characteristics of the lard crystals predominate. In many cases a positive identification of the crystals is only made possible by repeated crystallizations. In the examination of so-called refined lards, which are mixtures of lard and beef fat, the form of aggregation of the crystals is found to resemble the fan-shaped typical forms of beef fat. When the single crystals, however, are examined with a higher magnifying power, they are not found to be pointed but blunt, and some present the appearance of plates with oblique terminations, but not so characteristic as those obtained from pure lard. In other cases in compound lards no beef fat crystals are observed and these lards may have been made partly of cotton oil stearin. When a lard crystal presents its edge to observation it may readily escape identification, or may even be mistaken for a crystal of beef fat. In order to insure a side view the cover glass should be pressed down with a slight rotatory movement, whereby some of the lard crystals at least may be made to present a side view.

309. Observation of Fat Crystals with Polarized Light.—The appearance of fat crystals, when observed by means of polarized light alone or with the adjunct of a selenite plate, is often of value in distinguishing the nature and origin of the sample.[262]

Every fat and oil which is amorphous will present the same set of phenomena when observed with polarized light through a selenite plate, but when a fat has been melted and allowed to cool slowly the field of vision will appear mottled and particolored when thus examined. This method has been largely used in the technical examination of butter for adulterants, and the microscope is extensively employed by the chemists of the Bureau of Internal Revenue for this purpose. In the examination of the crystals of butter fat by polarized light a cross is usually observed when the nicols are turned at the proper angle, but the cross, while almost uniformly seen with butter, is not distinctive, since other fats often show it. These forms of crystals are best obtained by heating the butter fat to the boiling-point of water for about a minute and then allowing it to slowly solidify, and stand for twenty-four hours.

Pure butter, properly made, is never subjected to fusion, and hence, when examined through a selenite plate, presents a uniform field of vision similarly illuminated and tinted throughout. In oleomargarin, the fats are sometimes, during their preparation, in a fused condition. The field of vision is therefore filled to a greater or less extent with crystals more or less perfect in form. Some of these crystals, being doubly refracting, will impart to a selenite field a mottled appearance. Such a phenomenon is therefore indicative of a fraudulent butter or of one which has been at some time subjected to a temperature at or above its fusing point.

310. Spectroscopic Examination of Oils.—The presence of chlorophyll or of its alteration products is a characteristic of crude oils of vegetable origin. In refined oils, even when of a vegetable origin, all traces of the chlorophyll products may disappear. The absorption bands given by oils are not all alike and in doubtful cases a suspected sample should be compared with one of known origin.

In conducting the examination, the oil in a glass vessel with parallel sides, is placed in front of the slit of the spectroscope and any absorption band is located by means of the common divided scale and by the color of the spectrum on which it falls. Olive and linseed oils give three sharply defined absorption bands, a very dark one in the red, a faint one on the orange and a well marked one in the green.

Sesame, arachis, poppyseed and cottonseed oils also show absorption bands. Castor and almond oils do not affect the spectrum.