Principles and Practice of Agricultural Analysis

PART FIRST. SAMPLING, DRYING, INCINERATION AND EXTRACTIONS. — DRYING ORGANIC BODIES. — INCINERATION. — TREATMENT WITH SOLVENTS. — AUTHORITIES CITED IN PART FIRST.

1. Introduction.—The analyst may approach the examination of agricultural products from various directions. In the first place he may desire to know their proximate and ultimate constitution irrespective of their relations to the soil or to the food of man and beast. Secondly, his study of these products may have reference solely to the determination of the more valuable plant foods which they have extracted from the soil and air. Lastly, he may approach his task from a hygienic or economic standpoint for the purposes of determining the wholesomeness or the nutritive and economic values of the products of the field, orchard, or garden. In each case the object of the investigation will have a considerable influence on the method of the examination.

It will be the purpose of the present volume to discuss fully the principles of all the standard processes of analysis and the best practice thereof, to the end that the investigator or analyst, whatever may be the design of his work, may find satisfactory directions for prosecuting it. As in the previous volumes, it should be understood that these pages are written largely for the teacher and the analyst already skilled in the principles of analytical chemistry. Much is therefore left to the individual judgment and experience of the worker, to whom it is hoped a judicious choice of approved processes may be made possible.

2. Scope Of the Work.—Under the term agricultural products is included a large number of classes of bodies of most different constitution. In general they are the products of vegetable and animal metabolism. First of all come the vegetable products, fruits, grains and grasses. These may be presented in their natural state, as cereals, green fruits and fodders, or after a certain preparation, as starches, sugars and flours. They may also be met with in even more advanced stages of change, as cooked foods, alcohols and secondary organic acids, such as vinegar. In general, by the term agricultural products is meant not only the direct products of the farm, orchard and forest, but also the modified products thereof and the results of manufacture applied to the raw materials. Thus, not only the grain and straw of wheat are proper materials for agricultural analysis, but also flour and bran, bread and cakes made therefrom. In the case of maize and barley, the manufactured products may extend much further, for not only do we find starch and malt, but also alcohol and beer falling within the scope of our work. In respect of animal products, the agricultural analyst may be called on to investigate the subject of leather and tanning; to determine the composition of meat, milk and butter; to pass upon the character of lard, oleomargarine, and, in general, to determine as fully as possible the course of animal food in all its changes between the field, the packing house and the kitchen.

3. Limitations of Work.—It is evident from the preceding paragraph, that in order to keep the magnitude of this volume within the limits fixed for a single volume the text must be rigidly confined to the fundamental principles and practice of agricultural analysis. The interesting region of pharmacy and allied branches, in respect of plant analysis, can find no description here, and in those branches of technical chemistry, where the materials of elaboration are the products of the field only a superficial view can be given. The main purpose and motive of this volume must relate closely to the more purely agricultural processes.

4. General Manipulations.—There are certain analytical operations which are more or less of a general nature, that is, they are of general application without reference to the character of the material at hand. Among these may be mentioned the determination of moisture and of ash, and the estimation of matters soluble in ether, alcohol and other solvents. These processes will be first described. Preliminary to these analytical steps it is of the utmost importance that the material be properly prepared for examination. In general, this is accomplished by drying the samples until they can be ground or crushed to a fine powder, the attrition being continued until all the particles are made to pass a sieve of a given fineness. The best sieve for this purpose is one having circular apertures half a millimeter in diameter. Some products, both vegetable and animal, require to be reduced to as fine a state as possible without drying. In such instances, passing the product through a sieve is obviously impracticable. Special grinding and disintegrating machines are made for these purposes and they will be described further on.

There are some agricultural products which have to be prepared for examination in special ways and these methods will be given in connection with the processes for analyzing the bodies referred to. Nearly all the bodies, however, with which the analyst will be concerned, can be prepared for examination by the general methods about to be described.

5. Preparation of the Sample. (a) Vegetable Substances.—For all processes of analysis not executed on the fresh sample, substances of a vegetable nature should, if in a fresh state, be dried as rapidly as possible to prevent fermentative changes. It is often of interest to determine the percentage of moisture in the fresh sample. For this purpose a representative portion of the sample should be rapidly reduced to as fine a condition as possible. To accomplish this it should be passed through a shredding machine, or cut by scissors or a knife into fine pieces. A few grams of the shredded material are dried in a flat-bottomed dish at progressively increasing temperatures, beginning at about 60° and ending at from 100° to 110°. The latter temperature should be continued for only a short time. The principle of this process is based upon the fact that if the temperature be raised too high at first, some of the moisture in the interior cells of the vegetable substance can be occluded by the too rapid desiccation of the exterior layers which would take place at a high temperature. The special processes for determining moisture will be given in another place.

The rest of the sample should be partly dried at a lower temperature or air-dried. In the case of fodders and most cattle foods the samples come to the analyst in a naturally air-dried state. When grasses are harvested at a time near their maturity they are sun-dried in the meadows before placing in the stack or barn. Such sun-dried samples are already in a state fit for grinding. Green grasses and fodders should be dried in the sun, or in a bath at a low temperature from 50° to 60° until all danger of fermentative action is over, and then air- or sun-dried in the usual way.

Seeds and cereals usually reach the analyst in a condition suited to grinding without further preliminary preparation. Fruits and vegetables present greater difficulties. Containing larger quantities of water, and often considerable amounts of sugar, they are dried with greater difficulty. The principles which should guide all processes of drying are those already mentioned, viz., to secure a sufficient degree of desiccation to permit of fine grinding and at a temperature high enough to prevent fermentative action, and yet not sufficiently high to cause any marked changes in the constituents of the vegetable organism.

(b) Animal Substances.—The difficulties connected with the preliminary treatment of animal substances are far greater than those just mentioned. Such samples are composed of widely differing tissues, blood, bone, tendon, muscle and adipose matters, and all the complex components of the animal organism are to be considered. The whole animal may be presented for analysis, in which case the different parts composing it should be separated and weighed as exactly as possible. Where only definite parts are to be examined it is best to separate the muscle, bone, and fat as well as may be, before attempting to reduce the whole to a fine powder. The soft portions of the sample are to be ground as finely as possible in a meat or sausage cutter. The bones are crushed in some appropriate manner, and thus prepared for further examination. Where the flesh and softer portions are to be dried and finely ground, the presence of fat often renders the process almost impossible. In such cases the fat must be at least partially removed by petroleum or other solvent. In practically fat-free samples the material, after grinding in a meat cutter, can be partially dried at low temperatures from 60° to 75°, and afterwards ground in much the same manner as is practiced with vegetable substances.

As is the case with the preliminary treatment of vegetable matters, it is impossible to give any general directions of universal applicability. The tact and experience of the analyst in all these cases are better than any dicta of the books. In some instances, as will appear further on, definite directions for given substances can be given, but in all cases the general principles of procedure are on the lines already indicated.

6. Preserving Samples.—In most cases, as is indicated in the foregoing paragraphs, the sample may be dried before grinding to such a degree as to prevent danger from fermentation or decay. The fine-ground samples are usually preserved in glass-stoppered bottles, carefully marked or numbered. In some cases it is advisable to sterilize the bottles after stoppering, by subjecting them to a temperature of 100° for some time. In the case of cereals assurance should be had that the samples do not contain the eggs of any of the pests that often destroy these products. As a rule, samples should be kept for a time after the completion of the analytical work, and this is especially true in all cases where there is any prospect of dispute or litigation. In general it may be said, that samples should be destroyed only when they are spoiled, or when storage room is exhausted.

7. Collecting Samples.—When possible, the analyst should be his own collector. There is often as much danger from data obtained on non-representative samples as from imperfect manipulation. When personal supervision is not possible, the sample when received, should be accompanied by an intelligible description of the method of taking it, and of what it represents. In all cases the object of the examination must be kept steadily in view. Where comparisons are to be made the methods of collecting must be rigidly the same.

The processes of analysis, as conducted with agricultural products, are tedious and difficult. The absolutely definite conditions that attend the analysis of mineral substances, are mostly lacking. The simple determinations of carbon, hydrogen, nitrogen and sulfur, which are required in the usual processes of organic analysis, are simplicity itself when compared with the operations which have to be performed on agricultural products to determine their character and their value as food and raiment. We have to do here with matters on which the sustenance, health and prosperity of the human race are more intimately concerned than with any other of the sciences. This fact also emphasizes the necessity for care in collecting the materials on which the work is to be performed.

8. Grinding Samples.—In order to properly conduct the processes of agricultural analysis it is important to have the sample finely ground. This arises both from the fact that such a sample is apt to contain an average content of the various complex substances of which the material under examination is composed, and because the analytical processes can be conducted with greater success upon the finely divided matter. In mineral analysis it is customary to grind the sample to an impalpable powder in an agate mortar. With agricultural products, however, such a degree of fineness is difficult to attain, and moreover, is not necessary. There is a great difference of opinion among analysts respecting the degree of fineness desirable. In some cases we must be content with a sample which will pass a sieve with a millimeter mesh; in fact it may be found impossible, on account of the stickiness of the material, to sift it at all. In such cases a thorough trituration, so as to form a homogeneous mass will have to be accepted as sufficient. Where bodies can be reduced to a powder however, it is best to pass them through a sieve with circular perforations half a millimeter in diameter. A finer degree of subdivision than this is rarely necessary.

9. The Grinding Apparatus.—The simplest form of apparatus for reducing samples for analysis to a condition suited to passing a fine sieve is a mortar. Where only a few samples are to be prepared and in small quantities, it will not be necessary to provide anything further. After the sample is well disintegrated it is poured on the sieve and all that can pass is shaken or brushed through. The sieve is provided with a receptacle, into which it fits closely, to avoid loss of any particles which may be reduced to a dust. The top of the sieve, when shaken, may also be covered if there be any tendency to loss from dust. Any residue failing to pass the sieve is returned to the mortar and the process thus repeated until all the material has been secured in the receiver. The particles more difficult of pulverization are often different in structure from the more easily pulverized portions, and the sifted matter must always be carefully mixed before the subsample is taken for examination. Often the materials, or portions thereof, will contain particles tough and resistant to the pestle, but the operator must have patience and persistence, for it is highly necessary to accurate work that the whole sample be reduced to proper size.

Figure 1. Mill for Grinding Dry Samples.

Where many samples are to be prepared, or in large quantities, mills should take the place of mortars. For properly air-dried vegetable substances, some form of mill used in grinding drugs may be employed. Grinding surfaces of chilled corrugated steel are to be preferred. The essential features of such a mill are that it be made of the best material, properly tempered, and that the parts be easily separated for convenience in cleaning. The grinding surfaces must also be so constructed and adjusted as to secure the proper degree of fineness. In fig. 1 is shown a mill of rather simple construction, which has long been in satisfactory use in this laboratory. Small mills may be operated by hand power, but when they are to be used constantly steam power should be provided. In addition to the removal of nearly all the moisture by air-drying there are many oleaginous seeds which cannot be finely ground until their oil has been removed. For this purpose the grinding surfaces of the mill are opened so that the seeds are only coarsely broken in passing through. The fragments are then digested with light petroleum in a large flask, furnished with a reflux condenser. After digestion the fragments are again passed through the mill adjusted to break them into finer particles.

The alternate grinding and digestion are thus continued until the pulverization is complete. On a small specially prepared sample the total content of oil is separately determined.

Fresh animal tissues are best prepared for preliminary treatment by passing through a sausage mill. The partially homogeneous mass thus secured should be dried at a low temperature and reground as finely as possible. Where much fat is present it may be necessary to extract it as mentioned above, in the case of oleaginous seeds. In such cases both the moisture and fat in the original material should be determined on small specially prepared samples with as great accuracy as possible. Bones, hoofs, horns, hair and hides present special difficulties in preparation, which the analyst will have to overcome with such skill and ingenuity as he may possess.

The analyst will find many specially prepared animal foods already in a fairly homogeneous form, such as potted and canned meats, infant and invalid foods, and the like. Even with these substances, however, a preliminary grinding and mixing will be found of advantage before undertaking the analytical work. Many cases will arise which are apparently entirely without the classification given above. But even in such instances the analyst should not be without resources. Frequently some dry inert substance may be mixed with the material in definite quantities, whereby it is rendered more easily prepared. Perhaps no case will be presented where persistent and judicious efforts to secure a fairly homogeneous sample for analysis will be wholly unavailing.

Figure 2. Comminutor for Green Samples.

In the case of green vegetable matters which require to be reduced rapidly to a fine state of subdivision in order to secure even a fairly good sample some special provision must be made. This is the case with stalks of maize and sugar cane, root crops, such as potatoes and beets, and green fodders, such as clover and grasses. The chopping of these bodies into fine fodders by hand is slow and often impracticable. The particles rapidly lose moisture and it is important to secure them promptly as in the preparation of beet pulp for polarization. For general use we have found the apparatus shown in fig. 2 quite satisfactory in this laboratory. It consists of a series of staggered circular saws carried on an axis and geared to be driven at a high velocity, in the case mentioned, 1,400 revolutions per minute. The green material is fed against the revolving saws by the toothed gear-work shown, and is thus reduced to a very fine pulp, which is received in the box below. Stalks of maize, green fodders, sugar canes, beets and other fresh vegetable matters are by this process reduced to a fine homogeneous pulp, suited for sampling and for analytical operations. Such pulped material can also be spread in a fine layer and dried rapidly at a low temperature, thus avoiding danger of fermentative changes when it is desired to secure the materials in a dry condition or to preserve them for future examination. Samples of sorghum cane, thus pulped and dried, have been preserved for many years with their sugar content unchanged.

Figure 3. Rasp for Sugar Beets.

Such a machine is also useful in preparing vegetable matter for the separation of its juices in presses. Samples of sugar cane, sugar beets, apples and other bodies of like nature can thus be prepared to secure their juices for chemical examination. Such an apparatus we have found is fully as useful and indispensable in an agricultural laboratory as a drug mill for air-dried materials.

It is often desirable in the preparation of roots for sugar analysis to secure them in a completely disintegrated state, that is with the cellular tissues practically all broken. Such a pulped material can be treated with water and the sugar juices it contains thus at once distributed to all parts of the liquid mass. The operation is known as instantaneous diffusion. The pulp of the vegetable matter is thus introduced into the measuring flask along with the juices and the content of sugar can be easily determined. Several forms of apparatus have been devised for this purpose, one of which is shown in fig. 3. This process, originally devised by Pellet, has come into quite general use in the determination of the sugar content of beets.[1] It is observed that it can be applied to other tubers, such as the turnip, potato, artichoke, etc. It is desirable, therefore, that an agricultural laboratory be equipped with at least three kinds of grinding machines; viz., first, the common drug mill used for grinding seeds, air-dried fodders, and the like; second, a pulping machine like the system of staggered saws above described for the purpose of reducing green vegetable matter to a fine state of subdivision, or one like the pellet rasp for tubers; third, a mill for general use such as is employed for making sausages from soft animal tissues.

Figure 4. Dreef Grinding Apparatus.

10. Grinding Apparatus at Halle Station.—The machine used at the Halle station for grinding samples for analysis is shown in Fig. 4.[2] It is so adjusted as to have both the upper and lower grinding surfaces in motion. The power is transmitted through the pulley D, which is fixed to an axis carrying also the inner grinding attachment B. Through C₂, C₃, C₄, and C₁, the reverse motion is transmitted to the outer grinder A. By means of the lever E the two grinding surfaces can be separated when the mill is to be cleaned. The dree mill above described is especially useful for grinding malt, dry brewers’ grains, cereals for starch determinations and similar dry bodies. It is not suited to grinding oily seeds and moist samples. These, according to the Halle methods, are rubbed up in a mortar until of a size suited to analysis, and samples such as moist residues, wet cereals, mashes, beet cuttings, silage, etc., are dried before grinding. If it be desired to avoid the loss of acids which may have been formed during fermentation, about ten grams of magnesia should be thoroughly incorporated with each kilogram of the material before drying.

11. Preliminary Treatment of Fish.—The method used by Atwater in preparing fish for analysis is given below.[3] The same process may also be found applicable in the preparation of other animal tissues. The specimens, when received at the laboratory, are at once weighed. The flesh is then separated from the refuse and both are weighed. There is always a slight loss in the separation, due to evaporation and to slimy and fatty matters and small fragments of the tissues which adhere to the hands and the utensils employed in preparing the sample. Perfect separation of the flesh from the other parts of the fish is difficult, but the loss resulting from imperfect separation is small. The skin of the fish, although it has considerable nutritive value, should be separated with the other refuse.

The partial drying of the flesh for securing samples for analytical work is accomplished by chopping it as finely as possible and subjecting from fifty to one hundred grams of it for a day to a temperature of 96° in an atmosphere of hydrogen. After cooling and allowing to stand in the open air for twelve hours, the sample is again weighed, and then ground to a fine powder and made to pass a sieve with a half millimeter mesh. If the samples be very fat they cannot be ground to pass so fine a sieve. In such a case a coarser sieve may be used or the sample reduced to as fine and homogeneous a state as possible, and bottled without sifting.

The reason for drying in hydrogen is to prevent oxidation of the fats. As will be seen further on, however, such bodies can be quickly and accurately dried at low temperatures in a vacuum, and thus all danger of oxidation be avoided. In fact, the preliminary drying of all animal and vegetable tissues, where oxidation is to be feared, can be safely accomplished in a partial vacuum by methods to be described in another place. In order to be able to calculate the data of the analysis to the original fresh state of the substance, a portion of the fresh material should have its water quantitively determined as accurately as possible.

12. Volatile Bodies.—In agricultural analysis it becomes necessary to determine the percentage of bodies present in any given sample which is volatile at any fixed temperature. The temperature reached by boiling water is the one which is usually selected. It is true that this temperature varies with the altitude and within somewhat narrow limits at the same altitude, due to variations in barometric pressure. As the air pressure to which any given body is subjected, however, is a factor in the determination of its volatile contents, it will be seen that within the altitudes at which chemical laboratories are found, the variations in volatile content will not be important. This arises from the fact that while water boils at a lower temperature, as the height above the sea level increases, the corresponding diminished air pressure permits a more ready escape of volatile matter. As a consequence, a body dried to constant weight at sea level, where the temperature of boiling water is 100°, will show the same percentage of volatile matter as if dried at an altitude where water boils at 99°. When, therefore, it is desirable to determine the volatile matter in a sample approximately at 100°, it is better to direct that it be done in a space surrounded by steam at the natural pressure rather than at exactly 100°, a temperature somewhat difficult to constantly maintain. However, where it is directed or desired to dry to constant weight exactly at 100°, it can be accomplished by means of an air-bath or by a water-jacketed-bath under pressure, or to which enough solid matter is added to raise the boiling-point to 100°. It is not often, however, that it is worth while to make any special efforts to secure a temperature of 100°. When bodies are to be dried at temperatures above 100°, such as 105°, 110°, and so on, an air-bath is the most convenient means of securing the desired end. The different kinds of apparatus to be employed will be described in succeeding paragraphs.

13. Drying at the Temperature of Boiling Water.—The best apparatus for this process is so constructed as to have an interior space entirely surrounded with boiling water or steam, with the exception of the door by which entrance is gained thereto. The metal parts of the apparatus are constructed of copper, and to keep a constant level of water and avoid the danger of evaporating all the liquid, it is advisable to have a reflux condenser attached to the apparatus. It is also well to secure entrance to the interior drying oven, not only by the door, but also by small circular openings, which serve both to hold a thermometer and to permit of the aspiration of a slow stream of dry air through the apparatus during the progress of desiccation. The gaseous bodies formed by the volatilization of the water and other matters are thus carried out of the drying box and the process thereby accelerated. The bath should be heated by a burner so arranged as to distribute the flame as evenly as possible over the base. A single lamp, while it will boil the water in the center, will not keep it at the boiling-point on the sides. The temperature of the interior of the bath will not therefore reach 100°. The interior of the oven should be coated with a non-detachable carbon paint to promote the radiation of the heat from its walls, as well as to protect the parts from oxidation where acid fumes are produced during desiccation. Instead of a reflux condenser a constant water level may be maintained in the bath by means of a mariotte bottle or other similar device.

Figure 5. Water-jacketed Drying Oven.

When a bath of this kind is arranged for use with a partial vacuum, it should be made cylindrical in shape, with conical ends, as shown in fig. 5, in order to bear well the pressure to which it is subjected. Among the many forms of steam-baths offered, the analyst will have but little difficulty in selecting one suited to his work. To avoid radiation the exterior of the apparatus should be covered with a non-conducting material.

Figure 6. Thermostat for Steam-Bath.

14. Drying In a Closed Water Oven.—When it is desired to keep the temperature of a drying oven exactly at 100° instead of at the heat of boiling water, a closed water oven with a thermostat is to be employed. The oven should be so constructed as to secure a free circulation of the water about the inner space. Since as a rule the water between the walls of the apparatus will be subjected to a slight pressure, these walls should be made strong, or the cylindrical form of apparatus should be used. The thermostat used by the Halle Station is shown in Fig. 6.[4] A ⋃ shaped tube, with a bulb on one arm and a lateral smaller tube sealed on the other, is partly filled with mercury and connected by rubber tubes on the right with the gas supply, and on the left with the burner. The end carrying the bulb is connected directly by a rubber and metal tube with the water space of the oven. This device is provided with a valve which is left open until the temperature of the drying space reaches about 95°. The tube conducting the gas is held in the long arm of the ⋃ by means of a cork through which it passes air-tight and yet is loose enough to permit of its being moved. Its lower end is provided with a long ▲ shaped slit. When the valve leading to the water space is closed and the water reaches the boiling point, the pressure of the vapor depresses the mercury in the bulb arm of the ⋃ and raises it in the other. As the mercury rises it closes the wider opening of the ▲ shaped slit, thus diminishing the flow of gas to the burner. By moving the gas entry tube up or down a position is easily found in which the temperature of the drying space, as shown by the thermometer, is kept accurately and constantly at 100°.

In a bath arranged in this way a steam condenser is not necessary. Since, however, in laboratories which are not at a higher altitude than 1,000 feet the boiling-point of water is nearly 100°, it does not seem necessary to go to so much trouble to secure the exact temperature named. There could be no practical difference in the percentage of moisture determined at 100°, and at the boiling-point of water at a temperature not more than 1° lower.

15. Drying in an Air-Bath.—In drying a substance in a medium of hot air surrounded by steam, as has been described, the process is, in reality, one of drying in air. The apparatus usually meant by the term air-bath, however, has its drying space heated directly by a lamp, or indirectly by a stratum of hot air occupying the place of steam in the oven already described. The simplest form of the apparatus is a metal box, usually copper, heated from below by a lamp. In the jacketed forms the currents of hot air produced directly or indirectly by the lamp are conducted around the inner drying oven, thus securing a more even temperature. The bodies to be dried are held on perforated metal or asbestos shelves in appropriate dishes, and the temperature to which they are subjected is determined by a thermometer, the bulb of which is brought as near as possible to the contents of the dish. One advantage of the air-bath is in being able to secure almost any desired temperature from that of the room to one of 150° or even higher. Its chief disadvantage lies in the difficulty of securing and maintaining an even temperature throughout all parts of the apparatus. Radiation from the sides of the drying oven should be prevented by a covering of asbestos or other non-combustible and non-conducting substance. The burner employed should be a broad one and give as even a distribution of the heat as possible over the bottom of the apparatus.

Figure 7. Spencer’s Drying Oven.

16. Spencer’s Air-Drying Oven.—In order to secure an even distribution of the heat in the desiccating space of the oven, Spencer has devised an apparatus, shown in the figure, in which the temperature is maintained evenly throughout the apparatus by means of a fan.[5] The oven has a double bottom, the space between the two bottoms being filled with air. The sides are also double, the space between being filled with plaster. The fan is driven by a toy engine connected with the compressed air service or other convenient method. Thermometers placed in different parts of the apparatus, while in use, show a rigidly even heat at all points so long as the fan is kept in motion. The actual temperature desired can be controlled by a gas regulator. This form of apparatus is well suited to drying a large number of samples at once. Portions of liquids and viscous masses may also be dried by enclosing them in bulbs and connecting with a vacuum.

Spencer’s oven can also be used to advantage in drying viscous liquids in a partial vacuum. For this purpose the flask A, Fig. 7, containing the substance is made with a round bottom to resist the atmospheric pressure. Its capacity is conveniently from 150 to 200 cubic centimeters. It is closed with a rubber stopper carrying a trap, H Hʹ, to keep the evaporated water from falling back. The details of the construction of the trap H are shown at the right of the figure. The vapors enter at the lateral orifice, just above the bulb, while the condensed water falls back into the bulb instead of into the flask A. A series of flasks can be used at once connected through the stopcocks G with the circular tube E leading to the vacuum. A water pump easily exhausts the apparatus, maintaining a vacuum of about twenty-seven inches. The hot air in the oven is kept in motion by the fan B, thus ensuring an even temperature in every part. The flask A may be partly filled with sand or pumice stone before the addition of the samples to be dried, and the weight of water lost is determined by weighing A before and after desiccation. If it be desired to introduce a slow current of dry air or some inert gas into A, it is easily accomplished by passing a small tube, connected with the dry air or gas supply, through the rubber stopper and extending it into the flask as far as possible without coming into contact with the contents.

17. Drying Under Diminished Air Pressure.—The temperature at which any given body loses its volatile products is conditioned largely by the pressure to which it is subjected. At an air pressure of 760 millimeters of mercury, water boils at 100° but it is volatilized at all temperatures. As the pressure diminishes the temperature at which a body loses water at a given rate falls. This is a fact of importance to be considered in drying many agricultural products. This is especially true of those containing oils and sugars, nearly the whole number. Invert sugar especially is apt to suffer profound changes at a temperature of 100°, the levulose it contains undergoing partial decomposition. Oils are prone to oxidation and partial decomposition at high temperatures in the presence of oxygen.

In drying in a partial vacuum therefore a double advantage is secured, that of a lower temperature of desiccation and in presence of less oxygen. It is not necessary to have a complete vacuum. There are few organic products which cannot be completely deprived of their volatile matters at a temperature of from 70° to 80° in a partial vacuum in which the air pressure has been diminished to about one-quarter of its normal force.

Figure 8. Electric Vacuum Drying Oven.

18. Electric Drying-Bath.—The heat of an electric current can be conveniently used for drying in a partial vacuum by means of the simple device illustrated in Fig. 8. In ordering a heater of this kind the voltage of the current should be stated. The current in use in this laboratory has a voltage of about 120, and is installed on the three wire principle. It is well to use a rheostat with the heater in order to control the temperature within the bell jar. The ground rim of the bell jar rests on a rubber disk placed on a thick ground glass or a metal plate, making an air-tight connection. A disk of asbestos serves to separate the heater from the dish containing the sample, in order to avoid too high a temperature.

19. Steam Coil Apparatus.—For drying at the temperature of superheated steam, it is convenient to use an apparatus furnished with layers or coils of steam pipes. The drying may be accomplished either in the air or in a vacuum. In this laboratory a large drying oven, having three shelves of brass steam-tubes and sides of non-conducting material, is employed with great advantage. The series of heating pipes is so arranged as to be used one at a time or collectively. Each series is furnished with a separate steam valve, and is provided with a trap to control the escape of the condensed vapors. In the bottom of the apparatus are apertures through which air can enter, which after passing through the interior of the oven escapes through a ventilator at the top. With a pressure of forty pounds of steam to the square inch and a free circulation of air, the temperature on the first shelf of the apparatus is about 98°; on the second from 103° to 104°, and on the third about 100°. The vessels containing the bodies to be dried are not placed directly on the brass steam pipes, but the latter are first covered with thick perforated paper or asbestos. For drying large numbers of samples, or large quantities of one sample, such an apparatus is almost indispensable to an agricultural laboratory.

Figure 9. Steam Coil Drying Oven.

A smaller apparatus is shown in Fig. 9. The heating part G is made of a small brass tube arranged near the bottom in a horizontal coil and continued about the sides in a perpendicular coil. Bodies placed on the horizontal shelf are thus entirely surrounded by the heating surfaces except at the top.[6] The steam pipe S is connected with the supply by the usual method, and the escape of the condensation is controlled either by a valve or trap in the usual way. The whole apparatus is covered by a bell jar B, resting on a heavy cast-iron plate P, through which also the ends of the brass coil pass. The upper surface of the iron plate may be planed, or a planed groove may be cut into it, to secure the edge of the bell jar. When the air is to be exhausted from the apparatus, a rubber washer should be placed under the rim of the bell jar. The latter piece of apparatus may either be closed, as shown in the figure, by a rubber stopper, or it is better, though not shown, to have a stopper with three holes. One tube passes just through the stopper and is connected with the vacuum; the second passes to the bottom of the apparatus and serves to introduce a slow stream of dry air or of an inert gas during the desiccation. The third hole is for a thermometer. When no movement of the residual gas in the apparatus is secured, a dish containing strong sulfuric acid S’ is placed on the iron plate and under the horizontal coil, as is shown in the figure. The sulfuric acid so placed does not reach the boiling-point of water, and serves to absorb the aqueous vapors from the residual air in the bell jar. By controlling the steam supply the desiccation of a sample can be secured in the apparatus at any desired temperature within the limit of the temperature of steam at the pressure used. Where no steam service is at hand a strong glass flask may be used as a boiler, in which case the trap end of the coil must be left open. The vacuum may be supplied by an air or bunsen pump. When a vacuum is not used an atmosphere of dry hydrogen may be supplied through H.

Figure 10. Carr’s Vacuum Drying Oven.

Figure 10. (Bis) Vacuum Oven Open.

20. Carr’s Vacuum Oven.—A convenient drying oven has been devised in this laboratory by Carr.[7] It is made of a large tube, preferably of brass. The tube may be from six to nine inches in diameter and from twelve to fifteen inches long. One end is closed air-tight by a brass end-piece attached by a screw, or brazed. The other end is detachable and is made air-tight by ground surfaces and a soft washer. In the figure this movable end-piece is shown attached by screw-nuts, but experience has shown that these are not necessary. On the upper longitudinal surfaces are apertures for the insertion of a vacuum gauge and for attachment to a vacuum apparatus.

In the figure the thermometer and aperture for introducing dry air or an inert gas are shown in the movable end disk, but they would be more conveniently placed in the fixed end. The oven is heated below by a gas burner, which conveniently should be as long as the oven. The heat is not allowed to strike the brass cylinder directly, but the latter is protected by a piece of asbestos paper.

The temperature inside of the oven can be easily kept practically constant by means of a gas regulator, not shown in the figure, or by a little attention to the lamp. For a vacuum of twenty inches a temperature of about 80° should be maintained. When the vacuum is more complete a lower temperature can be employed. This apparatus is simple in construction, strong, cheap, and highly satisfactory in use.

21. Drying in Hydrogen.—In some of the processes of agricultural analysis it becomes important to dry the sample in hydrogen or other inert gas. This may be accomplished by introducing the dry gas desired into some form of the apparatus already described. The drying may either be accomplished in an atmosphere of hydrogen practically at rest or in a more limited quantity of the gas in motion. The latter method is to be preferred by reason of its greater rapidity. The analyst has at his command many forms of apparatus designed for the purpose mentioned above. It will be sufficient here to describe only two, devised particularly for agricultural purposes.

The first one of these, designed by the author, was intended especially for drying the samples of fodders for analysis according to the methods of the Association of Agricultural Chemists.[8]

Figure 11. Apparatus for Drying in a
Current of Hydrogen.

For the purpose of drying materials contained in flasks and tubes in a current of hydrogen the apparatus shown in Fig. 11 is used. This apparatus consists of a circular box, B, conveniently made of galvanized iron, having a movable cover, S, fitted for the introduction of steam into the interior of the apparatus. Condensed steam escapes at W. A stream of perfectly pure and dry hydrogen enters at H, passes up through the material to be dried, down through the bulb V, containing sulfuric acid, and follows the direction of the arrows through the rest of the apparatus. The stream of hydrogen is thus completely dried by passing through bulbs containing sulfuric acid, on the way from one piece of the apparatus to the other. A, represents a flask such as is used, with the extraction apparatus described. The apparatus which we have used will hold eight tubes or flasks at a time, and thus a single stream of hydrogen is made to do duty eight times in drying eight separate samples. The great advantage of the apparatus is in the fact that the stream of hydrogen must pass over and through the substance to be dried. In order to prevent any sulfuric acid from being carried forward into the next tube the bulb K, above the sulfuric acid, may be filled with solid pieces of soda or potash.

This apparatus has been in use for a long time and no accidents from sulfuric acid being carried forward have occurred, and there is no danger, provided the stream of hydrogen is kept running at a slow rate. If, however, by any accident the stream of hydrogen should be admitted with great rapidity, particles of the sulfuric acid might be carried forward and spoil the next sample. To avoid any such accident as this the proposal to introduce the potash bulb has been made. The apparatus works with perfect satisfaction, and it is believed that when properly adjusted check weighings can be made by weighing the bulbs, showing their increase in weight, which will give the volatile matter, and weighing the flasks or tubes, which will show the loss of weight. The only chance for error in weighing the bulbs is that some of the volatile matter may be material which is not dissolved in sulfuric acid, and is thus carried on and out of the apparatus. The blackening of the sulfuric acid in the bulbs, in the drying of all forms of organic matter, shows that the loss in weight of such bodies is not due to water alone, but also to organic volatile substances, which are capable of being decomposed by the sulfuric acid, thus blackening it.

22. Caldwell’s Hydrogen Drying-Bath.—An excellent device for drying in hydrogen has been described by Caldwell.[9] A vessel of copper or other suitable material serves to hold the tubes containing the samples to be dried. It should be about twenty-four centimeters long, fifteen high, and eight wide. This vessel is contained in another made of the same material and of the dimensions shown in the figure. On one side the edge of this containing vessel may not be more than one centimeter high and the bath should rest against it. The other side is made higher to form a support for the drying tubes as indicated.

Figure 12. Caldwell’s Hydrogen Drying Apparatus.

The tube containing the substance a d is made of glass and may be closed by the ground stoppers c b or the tube stoppers e f. At a it carries a perforated platinum disk for holding the filtering felt. The tube should be about thirteen centimeters long and have an internal diameter of about twenty millimeters. With its stoppers it should weigh only a little over thirty grams. The asbestos felt should not be thick enough to prevent the free passage of gas. Passing diagonally through the bath are metal tubes, preferably made of copper, and of such a size as just to receive the glass drying tubes. If these be a little loose they should be made tight by wrapping them with a narrow coil of paper at either end of the tubular receptacle. The entrance of cold air between the glass tube and its metal holder is thus prevented, and the glass tube is held firmly in position. The glass tube should be weighed with its two solid stoppers. Afterwards the sample, about two grams, is placed on the asbestos felt and the stoppers replaced and the whole reweighed. The exact weight of the sample is thus obtained. The solid stoppers are then removed and the tube stoppers inserted. The lower end of the tube is then connected with the supply of dry hydrogen. The upper tube stopper is connected by a rubber tube with a small bottle containing sulfuric acid through which the escaping hydrogen is made to bubble. A double purpose is thus secured; moisture is kept from entering the drying tube and the rate at which the hydrogen is passing is easily noted. After the drying is completed the solid stoppers are again inserted, the tube cooled in a desiccator and weighed. The loss of weight is entered as water. The tube containing the sample can afterwards be put into an extractor and treated with ether or petroleum in the manner hereafter described. This apparatus requires more hydrogen than the one previously described, but it is rather simple in construction, is easily controlled, and has given satisfactory results.

Figure 13. Liebig’s Ente.

23. Drying in Liebig’s Tubes.—In drying samples, especially of fodders, the method practiced at the Halle Station is to place them in drying tubes, the form of which is shown in Fig. 13. A stream of illuminating gas, previously dried by passing over sulfuric acid and calcium chlorid, is directed through the tubes.[10] Many of these tubes can be used at once, arranged as shown in Fig. 14. When the air is all driven out the stream of gas can be ignited so as to regulate the flow properly by the size of the flame. The tubes are held in drying ovens, as shown in the figure, the temperature of which should be kept at 105°-107°. The drying should be continued for eight or ten hours. At the end of this time the gas in the tube is to be expelled by a stream of dry air and the tubes cooled in a desiccator and weighed. There are few advantages in this method not possessed by the processes already described. The samples, moreover, are not left in a condition for further examination, either by incineration or extraction.

Figure 14. Drying Apparatus used at the Halle Station.

24. Wrampelmayer’s Drying Oven.—The apparatus used at the Wageningen Station, in Holland, for drying agricultural samples, was devised by Wrampelmayer and is shown in Fig. 15. The oven is so constructed as to permit of drying in a stream of inert gas. Illuminating gas is let into the drying space of the oven through the tube A B. At B the entering gas is heated by the same lamp which boils the liquid in the water space of the apparatus. The hot gas is dried in the calcium chlorid tube c and then passes into the oven at D. At E it leaves the apparatus and is thence conducted to the lamp F, used for heating the bath. The lamp should be closed by a wire gauze diaphragm to prevent any possible explosion by reason of any admixture with the air in the oven. The condensation of the aqueous vapors is effected by means of the condenser G. In the drying space is a small shelf holder, which, by means of the hook H, can be removed from the apparatus. The drying space is closed from the upper part of the apparatus, which contains no water by the cover J, resting on a support K. This rim is covered with a rubber gasket L, by means of which the cover J can be fastened with a bayonet latch air-tight. This fastening is shown at N. Being closed in this way the part of the cylindrical oven above the cover may be left entirely open. Instead of the rather elaborate method of closing the bath, some simple and equally effective device might be used. The cover J is best made with double metallic walls enclosing an asbestos packing.

Figure 15. Wrampelmayer’s Oven.

It is evident that this oven could be used with an atmosphere of carbon dioxid or of air, provided the gas for heating were derived from a separate source and the tube between E and F broken. In a drying oven designed by the author, the movable top is made with double walls and the space between is joined to the steam chamber by means of a flexible metallic tube, thus entirely surrounding the drying space with steam.

25. The Ulsch Drying Oven.—A convenient drying oven is described by Ulsch which varies from the ordinary form of a water-jacketed drying apparatus in having a series of drying tubes inserted in the water-steam space.

Figure 16. Ulsch Drying Oven.

The arrangement of the oven is shown in the accompanying figure. The water space is filled only to about one-third of its height. When the heat is applied the cock c is left open until the steam has driven out all the air. It is then closed and the temperature of the bath is then regulated by the manometer e, connected with the bath by d. The bottom of the manometer cylinder contains enough mercury to always keep sealed the end of the manometer tube. The rest of the space is filled with water. At the top the manometer tube is expanded into a small bulb which serves as a gas regulator, as shown in the figure. The gas is admitted also by a small hole above the mercury in the bulb, so that when the end of the gas inlet tube is sealed enough gas still passes through to keep the lamp burning. With a mercury pressure of thirty centimeters the temperature of the bath will be about 105°. The walls of the bath should be made strong enough to bear the pressure corresponding to this degree. The drying can be accomplished either in the cubical drying box a or in the drying tubes made of thin copper and disposed as shown in the figure. The natural draft is shown by the arrows. The substance is held in boats placed in the tube as indicated. The air in traversing the tube is brought almost to the temperature of the water-steam space in which the tube lies. The natural current of hot air can easily be replaced by a stream of dry illuminating or other inert gas.

26. Drying Viscous Liquids.—In the case of cane juices, milk, and similar substances, the paper coil method may be used.[11] The manipulation is conducted as follows: A strip of filtering paper from five to eight centimeters wide and forty centimeters in length, is rolled into a loose coil and dried at the temperature of boiling water for two hours, placed in a dry glass-stoppered weighing tube, cooled in a desiccator and weighed. The stoppered weighing tube prevents the absorption of hygroscopic moisture. About three cubic centimeters of the viscous or semi-viscous liquid are placed in a flat dish covered by a plate of thin glass and weighed. The coil is then placed on end in the dish, and the greater part of the liquid is at once absorbed. The proportions between the coil and the amount of liquid should be such that the coil will not be saturated more than two-thirds of its length. It is then removed and placed dry end down in a steam-bath and dried two hours. The dish, covered by the same plate of glass, is again weighed, the loss in weight representing the quantity of liquid absorbed by the coil. After drying for the time specified the coil is again placed in the hot weighing tube, cooled and its weight ascertained. The increase represents the solid matter in the sample taken. This method has been somewhat modified by Josse, who directs that it be conducted as follows:[12] Filter-paper is cut into strips from one to two centimeters wide and three meters long. The strips are crimped so they will not lie too closely together and then wrapped into coils. These coils can absorb about ten cubic centimeters of liquid. One of them is placed in a flat dish about two centimeters high and seven in diameter, and dried as described, covered, cooled and weighed. There are next placed in the dish and weighed one or two grams of the massecuite, molasses, etc., which are to be dried and the dish again weighed and the total weight of the matter added, determined by deducting the weight of the dish and cover. About eight cubic centimeters of water are added, the material dissolved with gentle warming, the coil placed in the dish, and the whole dried for two hours. The cover is then replaced and the whole cooled in a desiccator and weighed. The increase in weight represents the dry matter in the sample taken.

The above method of solution of a viscous sample in order to divide it evenly for desiccation is based on the principle of the method first proposed by the author and Broadbent for drying honeys and other viscous liquids.[13] In this process the sample of honey, molasses, or other viscous liquid is weighed in a flat dish, dissolved in eighty per cent alcohol, and then a weighed quantity of pure dry sand added, sufficient to fill the dish three-quarters full. The alcoholic solution of the viscous liquid is evenly distributed throughout the mass of sand by capillary attraction, and thus easily and rapidly dried when placed on the bath.

Pumice stone, on account of its great porosity, is also an excellent medium for the distribution of a viscous liquid in aiding the process of desiccation. The method has been worked out in great detail in this laboratory by Carr and Sanborn,[14] and most excellent results obtained. Round aluminum dishes two centimeters high and from eight to ten centimeters in diameter are conveniently used for this process. The pumice stone is dried and broken into fragments the size of a pea before use.

27. General Principles of Drying Samples.—It would be a needless waste of space to go into further details of devices for desiccation. A sufficient number has been given to fully illustrate all the principles involved. In general, it may be said that drying in the open air at a temperature not exceeding that of boiling water can be safely practiced with the majority of samples. For instance, we have found practically no change in this laboratory in the composition of cereals dried in the air and in an inert gas. The desiccation should in all cases be accomplished as speedily as possible. To this end the atmosphere in contact with the sample should be dry and kept in motion. An oven surrounded by boiling water and steam is to be preferred to one heated by air. Constancy of temperature is quite as important as its degree and this steadiness is most easily secured by steam at atmospheric pressure. Where higher temperatures than 100° are desired the steam must be under pressure, or the boiling-point of the water may be raised by adding salt or other soluble matters. A bath of paraffin or calcium chlorid may also be used or a sand or air-bath may be employed. The analyst must not forget, however, that inorganic matters are prone to change at temperatures above 100°, even in an inert atmosphere, and higher temperatures must be used with extreme caution.

Drying in partial vacuum and in a slowly changing atmosphere may be practiced with all bodies and must be employed with some. The simple form of apparatus already described will be found useful for this purpose. At a vacuum of twenty inches or more, even unstable organic agricultural products are in little danger of oxidation. In the introduction of a dry gas, therefore, air will be found as a rule entirely satisfactory. In the smaller form of vacuum apparatus described, however, there is no objection to the employment of hydrogen or of carbon dioxid. The gas entering the apparatus should be dried by passing over calcium chlorid or by bubbling through sulfuric acid. In this laboratory the vacuum is provided by an air-pump connected with a large exhaust cylinder. This cylinder is connected by a system of pipes to all the working desks. The chief objection to this system is the unsteadiness of the pressure. When only a few are using the vacuum apparatus for filtering or other purposes the vacuum will stand at about twenty inches. When no one is using it the vacuum will rise to twenty-eight or twenty-nine inches. At other times, when in general use, it may fall to fifteen inches. Where a constant vacuum is desired for drying, therefore, it is advisable to connect the apparatus with a special aspirator which will give a pressure practically constant.

The dishes containing the sample should be low and flat, exposing as large a surface as possible. For viscous liquids it will be found advisable to previously fill the dishes with pumice stone or other inert absorbent material to increase the surface exposed.

The special methods of drying milk, sirup, honeys, and like bodies, will be described in the paragraphs devoted to these substances.

In drying agricultural products, not only water but all other matters volatile at the temperature employed are expelled. It is only necessary to conduct the products of volatilization through sulfuric acid to demonstrate the fact that organic bodies are given off. In the case mentioned the sulfuric acid will be speedily changed to a brown and even black color by these bodies. It is incontestable, however, that in most cases the essential oils and other volatile matters thus escaping are not large in quantity and could not appreciably affect the percentage composition of the sample. In such cases the whole of the loss on drying is entered in the note book as water. There are evidently many products, however, where a considerable percentage of the volatile products is not water. The percentage of essential oils, which have a lower boiling-point than water, can be determined in a separate sample and this deducted from the total loss on drying will give the water.

Simple as it seems, the determination of water in agricultural products often presents peculiar difficulties and taxes to the utmost the patience and skill of the analyst. Having set forth the substantial principles of the process and indicated its more important methods, there is left for the worker in the laboratory the choice of processes already described, or, in special cases, the device of new ones and adaption of old ones to meet the requirements of necessity.

28. Determination of Ash.—The principle to be kept in view in the preparation of the ash of agricultural products is to conduct the incineration at as low a temperature as possible to secure a complete combustion. The danger of too high a temperature is two-fold. In the first place some of the mineral constituents constantly present in the ash, notably, some of the salts of potassium and sodium are volatile at high temperatures and thus escape detection. In the second place, some parts of the ash are rather easily fusible and in the melted state occlude particles of unburned organic matter, and thus protect them from complete oxidation. Both of these dangers are avoided, and an ash practically free of carbon obtained, by conducting the combustion at the lowest possible temperature capable of securing the oxidation of the carbonaceous matter.

29. Products Of Combustion.—The most important product of combustion, from the present point of view, is the mineral residue obtained. The organic matter of the sample undergoes decomposition in various ways, depending chiefly on its nature. Complex volatile compounds are formed first largely of an acid nature. The residual carbon is oxidized to carbon dioxid and the hydrogen to water. The relative proportions of these bodies formed, in any given case, depend on the conditions of combustion. With a low temperature and a slow supply of oxygen, the proportion of volatile organic compounds is increased. At a high temperature, and in a surplus of oxygen, the proportions of water and carbon dioxid are greater. At the present time, however, our attention is to be directed exclusively to the mineral residue; the organic products of combustion belonging to the domain of organic chemistry. As has already been intimated, the ash of agricultural samples consists of the mineral matters derived from the tissues, together with any accidental mineral impurities which may be present, some unburned carbon, and the sulfur, phosphorus, chlorin, nitrogen, etc., existing previously in combination with the mineral bases. The organic sulfur and phosphorus may also undergo complete or partial oxidation during incineration and be found in the ash. Unless special precautions be taken, however, a portion of the organic sulfur and phosphorus may escape as volatile compounds during the combustion.[15] The organic nitrogen is probably completely lost, at most, only traces of it being oxidized during the combustion in such a way as to combine with a mineral base. The rare mineral elements that are taken up by plants will also be found in the ash. Here the analyst would look for copper, boron, zinc, manganese, and the other elements which, when existing in the soil, are apt to be found in the tissues of the plants, not, perhaps, as organic or essential compounds, but as concomitants of the other mineral foods absorbed by growing vegetation. This fact is often of importance in toxicological and hygienic examinations of foods. For instance, traces of copper or of boron in the ash of a prescribed food would not be evidence of the use of copper or borax salts as preservatives unless it could be shown that the soil on which the food in question was grown was free of these bodies.

This fact manifestly applies only to those cases where mere traces of these rare bodies are in question. The presence of considerable quantities of them, enough to be inimical to health, could only be attributed to artificial means.

30. Purpose and Conduct of Incineration.—In burning a sample of an agricultural product the analyst may desire to secure either a large sample of ash for analytical purposes as already described or to determine the actual percentage of ash. The first purpose is secured in many ways. In the preparation of ash for manurial purposes, for instance, little care is exercised either to prevent volatilization of mineral matters or to avoid the occurrence of a considerable quantity of carbon in the sample. With this operation we have, at present, nothing whatever to do. In preparing a sample of ash for chemical analysis it is important, where a sufficient quantity of the sample can be obtained, to use as large a quantity of it as convenient. While it is true that very good results may be secured on very small samples, it is always advisable to have a good supply of the material at hand. Since the materials burned have only from one to three per cent of ash, a kilogram of them will supply only from ten to thirty grams. To supply all needful quantities of material and replace the losses due to accident, whenever possible at least twenty grams of the ash should be prepared. The combustion can be carried on in platinum dishes with all bodies free of metallic oxids capable of injuring the platinum. Otherwise porcelain or clay dishes may be employed. As a rule the combustion is best conducted in a muffle at a low red heat. With substances very rich in fusible ash, as for instance the cereals, it is advisable to first char them, extract the greater part of the ash with water, and afterwards burn the residual carbon. The aqueous extract can then be added to the residue of combustion and evaporated to dryness at the temperature of boiling water. During the combustion the contents of the dish should not be disturbed until the carbon is as completely burned out as possible. The naturally porous condition in which the mass is left during the burning is best suited to the entire oxidation of the carbon. At the end however, it may become necessary to bring the superficial particles of unburned carbon into direct contact with the bottom of the dish by stirring its contents. In most instances very good results may be obtained by burning the ash in an open dish without the aid of a muffle. In this case a lamp should be used with diffuse flame covering as evenly as possible the bottom of the dish and thus securing a uniform temperature. The carbon, when once in active combustion, will as a rule be consumed, and an ash reasonably pure be obtained.

The second purpose held in view by the analyst is to determine the actual content of ash in a sample. For this purpose only a small quantity of the material should be used, generally from two to ten grams. The combustion should be conducted in flat-bottomed, shallow dishes, and at a low temperature. In many cases the residue, after determining the moisture, can be at once subjected to incineration, and thus an important saving of time be secured. A muffle, with gentle draft, will be found most useful for securing a white ash. The term, white ash, is sometimes a deceptive one. In samples containing iron or manganese, the ash may be practically free of carbon and yet be highly colored. The point at which the combustion is to be considered as finished therefore should be at the time the carbon has disappeared rather than when no coloration exists. In general the methods of incineration are the same for all substances, but some cases may arise in which special processes must be employed. Some analysts prefer to saturate the substance before incineration with sulfuric acid, securing thus a sulfated ash. This is practiced especially with molasses. In such cases the ash obtained is free of carbon dioxid and roughly the difference in weight is compensated for by deducting one-tenth of the weight of the ash when comparison is to be made with ordinary carbonated ash. Naturally this process could not be used when sulfuric acid is to be determined in the product.

Figure 17. Courtoune Muffle.

31. German Ash Method.—The method pursued at the Halle Station for securing the percentage of ash in a sample is as follows:[16] Five grams of the air-dried sample are incinerated in a platinum dish and the ash ignited until it has assumed a white, or at least a bright gray tint. As soon as combustible gases are emitted at the beginning of the incineration they are ignited and allowed to burn as long as possible. It is advisable to hasten the oxidation by stirring the mass with a piece of platinum wire. If the ash should become agglomerated, as sometimes happens with rich food materials, it must be separated by attrition. The ash, when cooled on a desiccator, is to be weighed. When great exactness is required, it is advised, as set forth in a former paragraph, to first carbonize the mass and then extract the soluble ash with hot water before completing the oxidation. When the latter is complete and the dish cooled the aqueous extract is added, evaporated to dryness and the incineration completed.

32. Courtonne’s Muffle.—The ordinary arrangement of a muffle, as in assaying, may be conveniently used in incineration. A special muffle arrangement has been prepared by Courtonne which not only permits of the burning of a large number of samples at once, but also effects a considerable saving in gas. The muffle as shown in Fig. 17, is made in two stages, and the floor projects in front of the furnace, forming a convenient hearth. The incineration is commenced on the upper stage, where the temperature is low, and finished on the lower one at a higher heat. The furnace is so arranged as to permit the flame of the burning gas to entirely surround the muffle. The draft and temperature within the muffle are controlled by the fire-clay door shown resting on the table.

33. Object Of Treatment.—The next step, in the analytical work, after sampling, drying, and incinerating, is the treatment of the sample with solvents. The object of this work is to separate the material under examination into distinct classes of bodies distinguished from each other by their solubilities. It is not the purpose of this section to describe the various bodies which may be separated in this way, especially from vegetable products. For this description the reader may consult the standard works on plant analysis.[17]

The chief object of a strictly agricultural examination of a field or garden product is to determine its food value. This purpose can be accomplished without entering into a minute separation of nearly allied bodies. For example, in the case of carbohydrates it will be sufficient as a rule, to separate them into four classes. In the first class will be found those soluble in water as the ordinary sugars. In the second group will be found those which, while not easily soluble in water, are readily rendered so by treatment with certain ferments or by hydrolysis with an acid. The starches are types of this class. In the third place are found those bodies which resist the usual processes of hydrolysis either with an acid or alkali, and therefore remain in the residue as fiber. Cellulose is a type of these bodies. In the fourth class are included those bodies which on hydrolysis with an acid yield furfurol on distillation, and therefore belong to the type containing five atoms of carbon or some multiple thereof in their molecule. For ordinary agricultural purpose the separation is not even as complete as is represented above.

What is true of the carbohydrates applies equally well to the fats and to other groups. Especially in the analysis of cereals and of cattle foods, the treatment with solvents is confined to the use in successive order of ether or petroleum, alcohol, dilute acids, and alkalies, the latter at a boiling temperature. The general method of treatment with these solvents will be the subject of the following paragraphs.

34. Extraction of the Fats and Oils.—Two solvents are in general use for the extraction of fats and oils; viz., ethylic ether and a light petroleum. The former is the more common reagent. Before use it should be made as pure as possible by washing first with water, afterwards removing the water by lime or calcium chlorid, and then completing the drying by treatment with metallic sodium. The petroleum spirit used should be purified by several fractional distillations until it has nearly a constant boiling-point of from 45° to 50°. The detailed methods of preparing these reagents will be given in another place. For rigid scientific determinations the petroleum is to be preferred to the ether. It is equally as good a solvent for fats and oils and is almost inert in respect to other vegetable constituents. Ether, on the other hand, dissolves chlorophyll and its partial oxidation products, resins, alkaloids and the like. The extract obtained by ether is therefore less likely to be a pure fat than that secured by petroleum. For purposes of comparison, however, the ether should be employed, inasmuch as it has been used almost exclusively in analytical operations in the past.

35. Methods Of Extraction.—The simplest method for accomplishing the extraction of fat from a sample consists in treating it with successive portions of the solvent in an open dish or a closed flask. This process is actually employed in some analytical operations, as, for instance, in the determination of fat in milk. Experience has shown, however, that a portion of the substance soluble, for instance, in ether, passes very slowly into solution, so that a treatment such as that just described would have to be long continued to secure maximum results. The quantity of solvent required would thus become very large and in the case of ether would entail a great expense. For the greater number of analytical operations, therefore, some device is employed for using the same solvent continually. The methods of extraction therefore fall into two general classes; viz., extraction by digestion and extraction by percolation. This classification holds good also for other solvents besides ether and petroleum. In general, the principles and practice of extraction described for ether may serve equally well for alcohol, acetone and other common solvents.

36. Extraction by Digestion.—In the use of ether or petroleum the sample is covered with an excess of the solvent and allowed to remain for some time in contact therewith. The soluble portions of the sample diffuse into the reagent. The speed of diffusion is promoted by stirring the mixtures. The operation may be conducted in an open dish or a flask. Inasmuch as the residue is, as a rule, to be dried and weighed, an open dish is to be preferred. To avoid loss of reagent and to prevent filling a working room with very dangerous gases, the temperature of digestion should be kept below the boiling-point of the solvent. The greater part of the soluble matter will be extracted with three or four successive applications of the reagent, but, as intimated above, the last portions of the soluble material are extracted with difficulty by this process. In pouring off the solvent care must be exercised to avoid loss of particles of the sample suspended therein. To this end it is best to pour the solvent through a filter. For the extraction of large quantities of material for the purpose of securing the extract for future examination, or simply to remove it, the digestion process is usually employed. This excess of solvent required is easily recovered by subsequent distillation and used again. The method is rarely used for the quantitive estimation of the extract, the process of continuous percolation being more convenient and more exact.

37. Extraction by Percolation.—In this method the solvent employed is poured on the top of the material to be extracted and allowed to pass through it usually by gravitation alone, sometimes with the help of a filter-pump. The principle of the process is essentially that of washing precipitates.

Two distinct forms of apparatus are in use for this process. In the first kind the solvent is poured over the material and after percolation is secured by distillation in another apparatus. In the second kind the solvent is secured after percolation in a flask where it is at once subjected to distillation. The vapors of the solvent are conducted by appropriate means to a condenser placed above the sample. After condensation the solvent is returned to the upper part of the sample. The percolation thus becomes continuous and a very small quantity of the solvent may thus be made to extract a comparatively large amount of material. This process is particularly applicable to the quantitive determination of the extract. After distillation and drying the latter may be weighed in the flask in which it was received or the sample may be dried and weighed in the vessel in which it is held both before and after extraction. One great advantage of the continuous extraction method lies in the fact that when it is once properly started it goes on without further attention from the analyst save an occasional examination of the flow of water through the condenser and of the rate of the distillation. For this reason the process may be continued for many hours without any notable loss of time. The vapor of the solvent in passing to the condenser may pass through a tube out of contact with the material to be extracted or it may pass directly around the tube holding the sample. In the former case the advantage is secured of conducting the extraction at a higher temperature, but there is danger of boiling the solvent in contact with the material and thus permitting the loss of a portion of the sample.

38. Apparatus Used for Extractions.—For extraction by digestion, as has already been said, an open dish may be used. When large quantities of material are under treatment, heavy flasks, holding from five to ten liters, will be found convenient. In these cases a condenser can be attached to the flask and the extraction conducted at the boiling temperature of the solvent. During the process of extraction it is advisable to shake the flask frequently. By proceeding in this way the greater part of the solvent matter will be removed after three or four successive treatments.

In extraction by percolation various forms of apparatus are employed. The ordinary percolators of the manufacturing pharmacist may be used for the larger operations, while the more elaborate forms of continuous extractors will be found most convenient for quantitive work. In each case the analyst must choose that process and form of apparatus best suited to the purpose in view. In the next paragraphs will be described some of the more common forms of apparatus in use.

39. Knorr’s Extraction Apparatus.—The apparatus which has been chiefly used in this laboratory for the past few years is shown in the accompanying figure.[18] The principle of the construction of the apparatus lies in the complete suppression of stoppers and in sealing the only joint of the device with mercury.

The construction and operation of the apparatus will be understood by a brief description of its parts.

A is the flask containing the solvent, W a steam bath made by cutting off the top of a bottle, inverting it and conducting the steam into one of the tubes shown in the stopper while the condensed water runs out of the other. The top of the bath is covered with a number of concentric copper rings, so that the opening may be made of any desirable size. B represents the condenser, which is a long glass tube on which a number of bulbs has been blown, and which is attached to the hood for holding the material to be extracted, as represented at Bʹ, making a solid glass union. Before joining the tube at Bʹ the rubber stopper which is to hold it into the outside condenser of B is slipped on, or the rubber stopper may be cut into its center and slipped over the tube after the union is made. In case alcohol is to be used for the solvent, requiring a higher temperature, the flask holding the solvent is placed entirely within the steam-bath, as represented at Aʹ.

Figure 18. Knorr’s Extraction Apparatus.

Figure 19. Extraction Flask.

Figure 20.
Extraction Tube.

Figure 21.
Extraction Siphon Tube.

A more detailed description of the different parts of the apparatus can be seen by consulting Figs. 19, 20, and 21. In A, Fig. 19, is represented a section of the flask which holds the solvent, showing how the sides of the hood containing the matters to be extracted pass over the neck of the flask, and showing at S a small siphon inserted in the space between the neck of the flask and the walls of the hood for the purpose of removing any solvent that may accumulate in this space. A view of the flask itself is shown at Aʹ. It is made by taking an ordinary flask, softening it about the neck and pressing the neck in so as to form a cup, as indicated at Aʹ, to hold the mercury which seals the union of the flask with the condenser. The flask is held in position by passing a rubber band below it, which is attached to two glass nipples, b, blown onto the containing vessel, as shown in Fig. 18. The material to be extracted may be contained in an ordinary tube, as shown in Fig. 20, which may be made from a test tube drawn out, as indicated in the figure, having a perforated platinum disk sealed in at D. The containing tube rests upon the edges of the flask containing the solvent by means of nipples shown at t. If a siphon tube is to be used, one of the most convenient forms is shown in Fig. 21, in which the siphon lies entirely within the extracting tube, thus being protected from breakage. By means of this apparatus the extractions can be carried on with a very small quantity of solvent, there being scarcely any leakage, even with the most volatile solvents, such as ether and petroleum. The apparatus is always ready for use, no corks are to be extracted, and no ground glass joints to be fitted.

40. Soxhlet’s Extraction Apparatus.—A form of continuous extraction apparatus has been proposed by Soxhlet which permits the passage of the vapors of the solvent into the condenser by a separate tube and the return of the condensed solvent after having stood in contact with the sample, to the evaporating flask by a siphon. The advantage of this process lies in freeing the sample entirely from the rise of temperature due to contact with the vapors of the solvent, and in the second place in the complete saturation of the sample with the solvent before siphoning. The sample is conveniently held in a cylinder of extracted filter-paper open above and closed below. This is placed in the large tube between the evaporating flask and the condenser. The sample should not fill the paper holder, and if disposed to float in the solvent, should be held down with a plug of asbestos fiber or of glass wool. The extract may be transferred, by dissolving in the solvent, from the flask to a drying dish, or it may be dried and weighed in the flask where first received.

Figure 22.
Soxhlet
Extraction
Apparatus.

There are many forms of apparatus of this kind, one of which is shown in Fig. 22, but a more extended description of them is not necessary. The disadvantages of this process as compared with Knorr’s, are quite apparent. The connections with the evaporating flask and condenser are made with cork stoppers, which must be previously thoroughly extracted with ether and alcohol. These corks soon become dry and hard and difficult to use. The joints are likely to leak, and grave dangers of explosion arise from the vapors of the solvents escaping into the working room. Moreover, it is an advantage to have the sample warmed by the vapors of the solvent during the progress of the extraction, provided the liquid in direct contact with the sample does not boil with sufficient vigor to cause loss.

The use of extraction apparatus with ground glass joints is also unsatisfactory. By reason of unequal expansion and contraction these joints often are not tight. They are also liable to break and thus bring danger and loss of time.

41. Compact Extraction Apparatus.—In order to bring the extraction apparatus into a more compact form, the following described device has been successfully used in this laboratory.[19] The condenser employed is made of metal and is found entirely within the tube holding the solvent.

This form of condenser is shown in Fig. 23, in which the tube E serves to introduce the cold water to the bottom of the condensing device. The tube D serves to carry away the waste water. The tube F serves for the introduction of the solvent by means of a small funnel. When the solvent is introduced and has boiled for a short time, the tube F should be closed. In each of the double conical sections of the condenser a circular disk B is found, which causes the water flowing from A upward to pass against the metallic surfaces of the condenser.

A section of the double conical condenser is shown in the upper right hand corner. It is provided with two small hooks hh, soldered on the lower surface, by means of which the crucible G can be hung with a platinum wire. The condenser is best made smooth and circular in form.

The crucible G, which holds the material to be extracted, can be made of platinum, but for sake of economy also of porcelain. The bottom of the porcelain crucible is left open excepting a small shelf, as indicated, which supports a perforated disk of platinum on which an asbestos film is placed.

Figure 23. Compact Condensing Apparatus.

The whole apparatus is of such size as to be easily contained in the large test-tube T.

The mouth of the test-tube is ground so as to fit as smoothly as possible to the ground-brass plate of the metallic condenser P.

In case it is desired to weigh the extract it may be done directly by weighing it in the test-tube T after drying in the usual way at the end of the extraction; or a glass flask H, made to fit freely into the test-tube, may be used, in which case a little mercury is poured into the bottom of the tube to seal the space between H and T. To prevent spirting of the substance in H, or projecting any of the extracted material without or against the bottom of the crucible G, the funnel represented by the dotted lines in the right hand section may be used.

Heat may be applied to the test-tube either by hot water, or steam, or by a bunsen, which permits of the flame being turned down to minimum proportions without danger of burning back. When the test-tube alone is used it is advisable to first put into it some fragments of pumice stone, particles of platinum foil, or a spoonful of shot, to prevent bumping of the liquid when the lamp is used as the source of heat.

Any air which the apparatus contains is pushed out through F when the boiling begins, the tube F not being closed until the vapor of the liquid has reached its maximum height. With cold water in the condenser the vapor of ether very rarely reaches above the lower compartment and the vapor of alcohol rarely above the second.

When the plate P is accurately turned so as to fit the ground surface of the mouth of T, it is found that ten cubic centimeters of anhydrous ether or alcohol are sufficient to make a complete extraction, and there is not much loss of solvent in six hours. The thickness of the asbestos film in G, or its fineness, is so adjusted as to prevent too rapid filtration so that the solvent may just cover the material to be extracted, or, after the material is placed in a crucible, a plug of extracted glass wool may be placed above it for the purpose of distributing the solvent evenly over the surface of the material to be extracted and of preventing the escape of fine particles.

Figure 24. Improved Compact Extraction Apparatus.

In very warm weather the apparatus may be arranged as shown in figure 24. The bath for holding the extraction tubes is made in two parts, K and Kʹ. The bath K has a false bottom shown in the dotted line O, perforated to receive the ends of the extraction tubes and which holds them in place and prevents them from touching the true bottom, where they might be unequally heated by the lamp. The upper bath Kʹ has a perforated bottom, partly closed with rubber-cloth diaphragms Gʹ Nʹ Hʹ. The extraction tubes passing through this bath, water-tight, permit broken ice or ice-water to be held about their tops, and thus secure a complete condensation of the vapors of the solvent which in warm weather might escape the metal condenser. In practice care must be taken to avoid enveloping too much of the upper part of the extraction tube with the ice-water, otherwise the vapors of the solvent will be chiefly condensed on the sides of the extraction tube and will not be returned through the sample. It is not often that the upper bath is needed, and then only with ether, never with alcohol. This apparatus has proved especially useful with alcohol, using, as suggested, glycerol in the bath. The details of its further construction and arrangement are shown in the figure. The extraction tubes are most conveniently arranged in a battery of four, one current of cold water passing in at A and out at B, serving for all. The bath is supported on legs long enough to allow the lamp plenty of room. The details of the condenser M are shown in Bʹ, Aʹ, T, Fʹ, and Lʹ. Instead of a gooch Lʹ for holding the sample a glass tube R, with a perforated platinum disk Q, may be used. The water line in the bath is shown by W. This apparatus may be made very cheaply and without greatly impairing its efficiency by using a plain concentric condenser and leaving off the upper bath Kʹ.

42. Solvents Employed.—It has already been intimated that the chief solvents employed in the extraction of agricultural samples are ether or petroleum and aqueous alcohol. The ether used should be free of alcohol and water, the petroleum should be subjected to fractional distillation to free it of the parts of very high and very low boiling points, and the alcohol as a rule should contain about twenty per cent of water.

There are many instances, however, where other solvents should be used. The use of aqueous alcohol is sometimes preceded by that of alcohol of greater strength or practically free of water. For the extraction of soluble carbohydrates (sugars) cold or tepid water is employed, the temperature of which is not allowed to rise high enough to act upon starch granules. For the solution of the starch itself an acid solvent is used at a boiling temperature, whereby the starch molecules undergo hydrolysis and form dextrin or soluble sugars (maltose, dextrose). By this process also the carbohydrates, whose molecules contain five, or some multiple thereof, atoms of carbon form soluble sugars of which xylose and arabinose are types. The solvent action of acids followed by treatment with dilute alkalies at a boiling temperature, completes practically the solution of all the carbohydrate bodies, save cellulose and nearly related compounds. The starch carbohydrates are further dissolved by the action of certain ferments such as diastase.

Dilute solutions of mineral salts exert a specific solvent action on certain nitrogenous compounds and serve to help separate the albuminoid bodies into definite groups.

Under the proper headings the uses of these principal solvents will be described, but a complete discussion of their action, especially on samples of a vegetable origin, should be looked for in works on plant analysis.[20]

The application of acids and alkalies for the extraction of carbohydrates, insoluble in water and alcohol, will be described, in the paragraphs devoted to the analysis of fodders and cereals. The extraction of these matters, made soluble by ferments, will be discussed in the pages devoted to starch and artificial digestion. It is thus seen that the general preliminary treatment of a sample preparatory to specific methods of examination is confined to drying, extraction with ether and alcohol, and incineration.

43. Recovery of the Solvent.—In using such solvents as ether, chloroform, and others of high value, it is desirable often to recover the solvent. Various forms of apparatus are employed for this purpose, arranged in such a way as both to secure the solvent and to leave the residue in an accessible condition, or in a form suited to weighing in quantitive work. When the extractions are made according to the improved method of Knorr, the flask containing the extract may be at once connected with the apparatus shown in figure 25.[21] A represents the flask containing the solvent to be recovered, W the steam-bath, B the condenser sealed by mercury, M and R the flask receiving the products of condensation. It will be found economical to save ether, alcohol, and chloroform even when only a few cubic centimeters remain after the extraction is complete. In the figure the neck of the flask A is represented as narrower than it really is. The open end of the connecting tube, which is sealed on A by mercury, should be the same size as the tube connecting with the condenser in the extraction apparatus.

Figure 25.—Knorr’s Apparatus for
Receiving Solvents.

Figure 26. Apparatus for Recovering Solvents
from Open Dishes.

It often happens that materials which are dissolved by the ordinary solvents in use are to be collected in open dishes in order that their properties may be studied. At the same time large quantities of solvents must be used, and it is desirable to have some method of recovering them. The device shown in Fig. 26 has been found to work excellently well for this purpose.[22] It consists of a steam-bath, W, and a bottle, B, with the bottom cut off, resting on an iron dish, P, containing a small quantity of mercury, enough to seal the bottom of the bottle. The dish containing the solvent is placed on the mercury, and the bottle placed down over it, forming a tight joint. On the application of steam the solvent escapes into the condenser, C, and is collected as a liquid in the flask A. In very volatile solvents the flask A may be surrounded with ice, or ice-cold water passed through the condenser. When an additional quantity of the solvent is to be added to the dish for the purpose of evaporating it is poured into the funnel F, and the stopcock H opened, which allows the material to run into the dish in B without removing the bottle. In this way many liters of the solvent may be evaporated in any one dish, and the total amount of extract obtained together. At the last the bottle B is removed, and the extract which is found in the dish is ready for further operations.

PART SECOND. SUGARS AND STARCHES. — THE DETERMINATION OF THE PERCENTAGE OF SUGAR BY THE DENSITY OF ITS SOLUTION. — THE DETERMINATION OF SUGAR WITH POLARIZED LIGHT. — DESCRIPTION OF POLARIZING INSTRUMENTS. — THE ANALYTICAL PROCESS. — CHEMICAL METHODS OF ESTIMATING SUGARS. — VOLUMETRIC METHODS. — VOLUMETRIC METHODS BASED UPON THE USE OF AN AMMONIACAL COPPER SOLUTION. — GRAVIMETRIC COPPER METHODS. — DETERMINATION OF STARCH. — AUTHORITIES CITED IN PART SECOND.

44. Introduction.—Carbohydrates, of which sugars and starches are the chief representatives, form the great mass of the results of vegetable metabolism. The first functions of the chlorophyll cells of the young plant are the condensation of carbon dioxid and water. The simplest form of the condensation is formaldehyd, CH₂O. There is no convincing evidence, however, that this is the product resulting from the functional activity of the chlorophyll cells. The first evidence of the condensation is found in more complex molecules; viz., those having six atoms of carbon. It is not the purpose of this work to discuss the physiology of this process, but the interested student can easily find access to the literature of the subject.[23] When a sample of a vegetable nature reaches the analyst he finds by far the largest part of its substance composed of these products of condensation of the carbon dioxid and water. The sugars, starches, pentosans, lignoses, and celluloses all have this common origin. Of many air-dried plants these bodies form more than eighty per cent.

In green plants the sugars exist chiefly in the sap. In plants cut green and quickly dried by artificial means the sugars are found in a solid state. They also exist in the solid state naturally in certain sacchariferous seeds. Many sugar-bearing plants when allowed to dry spontaneously lose all or the greater part of their sugar by fermentation. This is true of sugar cane, sorghum, maize stalks, and the like. The starches are found deposited chiefly in tubers, roots or seeds. In the potato the starch is in the tuber, in cassava the tuber holding the starch is also a root, in maize, rice and other cereals the starch is in the seeds. The wood-fibers; viz., pentosans, lignose, cellulose, etc., form the framework and support of the plant structure. Of all these carbohydrate bodies the most important as foods are the sugars and starches, but a certain degree of digestibility cannot be denied to other carbohydrate bodies with the possible exception of pure cellulose. In the following paragraphs the general principles of determining the sugars and starches will be given and afterwards the special processes of extracting these bodies from vegetable substances preparatory to quantitive determination.

45. Nomenclature.—In speaking of sugars it has been thought best to retain for the present the old nomenclature in order to avoid confusion. The terms dextrose, levulose, sucrose, etc., will therefore be given their commonly accepted significations.

A more scientific nomenclature has recently been proposed by Fischer, in which glucose is used as the equivalent of dextrose and fructose as the proper name for levulose. All sugars are further classified by Fischer into groups according to the number of carbon atoms found in the molecule. We have thus trioses, tetroses, pentoses, hexoses, etc. Such a sugar as sucrose is called hexobiose by reason of the fact that it appears to be formed of two molecules of hexose sugars. For a similar reason raffinose would belong to the hexotriose group.[24]

Again, the two great classes of sugars as determined by the structure of the molecule are termed aldoses and ketoses according to their relationship to the aldehyd or ketone bodies.

Since sugars may be optically twinned, that is composed of equal molecules of right and left-handed polarizing matter it may happen that apparently the same body may deflect the plane of polarization to the right, to the left, or show perfect neutrality.

Natural sugars, as a rule, are optically active, but synthetic sugars being optically twinned are apt to be neutral to polarized light.

To designate the original optical properties of the body therefore the symbols d, l, and i, meaning dextrogyratory, levogyratory, and inactive, respectively, are prefixed to the name. Thus we may have d, l, or i glucose, d, l, or i fructose, and so on.

The sugars that are of interest here belong altogether to the pentose and hexose groups; viz., C₅H₁₀O₅ and C₆H₁₂O₆, respectively. Of the hexobioses, sucrose, maltose, and lactose are the most important, and of the hexotrioses, raffinose. In this manual, unless otherwise stated, the term dextrose corresponds to d glucose, and levulose to d fructose. In this connection, however, it should be noted that the levulose of nature, or that which is formed by the hydrolysis of inulin or sucrose is not identical in its optical properties with the l fructose of Fischer.

46. Preparation of Pure Sugar.—In using the polariscope or in testing solutions for the chemical analysis of samples, the analyst will be required to keep always on hand some pure sugar. Several methods of preparing pure sugar have been proposed. The finest granulated sugar of commerce is almost pure. In securing samples for examination those should be selected which have had a minimum treatment with bluing in manufacture. The best quality of granulated sugar when pulverized, washed with ninety-five per cent and then with absolute alcohol and dried over sulfuric acid at a temperature not exceeding 50° will be found nearly pure. Such a sugar will, as a rule, not contain more than one-tenth per cent of impurities, and can be safely used for all analytical purposes. It is assumed in the above that the granulated sugar is made from sugar cane.

Granulated beet sugars may contain raffinose and so may show a polarization in excess of 100. This sugar may be purified by dissolving seventy parts by weight in thirty parts of water. The sugar is precipitated by adding slowly an equal volume of ninety-six per cent alcohol with constant stirring, the temperature of the mixture being kept at 60°. While still warm the supernatant liquor is decanted and the precipitated sugar washed by decantation several times with strong warm alcohol. The sugar, on a filter, is finally washed with absolute alcohol and dried in a thin layer over sulfuric acid at from 35° to 40°. By this process any raffinose which the sugar may have contained is completely removed by the warm alcohol. Since beet sugar is gradually coming into use in this country it is safer to follow the above method with all samples.[25] In former times it was customary to prepare pure sugar from the whitest crystals of rock candy. These crystals are powdered, dissolved in water, filtered, precipitated with alcohol, washed and dried in the manner described above.

47. Classification of Methods.—In the quantitive determination of pure sugar the various processes employed may all be grouped into three classes. In the first class are included all those which deduce the percentage of sugar present from the specific gravity of its aqueous solution. The accuracy of this process depends on the purity of the material, the proper control of the temperature, and the reliability of the instruments employed. The results are obtained either directly from the scale of the instruments employed or are calculated from the arbitrary or specific gravity numbers observed. It is evident that any impurity in the solution would serve to introduce an error of a magnitude depending on the percentage of impurity and the deviation of the density from that of sugar. The different classes of sugars, having different densities in solution, give also different readings on the instruments employed. It is evident, therefore, that a series of tables of percentages corresponding to the specific gravities of the solutions of different sugars would be necessary for exact work. Practically, however, the sugar which is most abundant, viz., sucrose, may be taken as a representative of the others and for rapid control work the densimetric method is highly useful.

In the second class of methods are grouped all those processes which depend upon the property of sugar solutions to rotate the plane of polarized light. Natural sugars all have this property and if their solutions be found neutral to polarized light it is because they contain sugars of opposite polarizing powers of equal intensity. Some sugars turn the polarized plane to the right and others to the left, and the degree of rotation in each case depends, at equal temperatures and densities of the solutions, on the percentages of sugars present. In order that the optical examination of a sugar may give correct results the solution must be of a known density and free of other bodies capable of affecting the plane of polarized light. In the following paragraphs an attempt will be made to give in sufficient detail the methods of practice of these different processes in so far as they are of interest to the agricultural analyst. The number of variations, however, in these processes is so great as to make the attempt to fully discuss them here impracticable. The searcher for additional details should consult the standard works on sugar analysis.[26]

In the third class of methods are included those which are of a chemical nature based either on the reducing power which sugar solutions exercise on certain metallic salts, upon the formation of certain crystalline and insoluble compounds with other bodies or upon fermentation. Under proper conditions solutions of sugar reduce solutions of certain metallic salts, throwing out either the metal itself or a low oxid thereof. In alkaline solutions of mercury and copper, sugars exercise a reducing action, throwing out in the one case metallic mercury and in the other cuprous oxid. With phenylhydrazin, sugars form definite crystalline compounds, quite insoluble, which can be collected, dried and weighed. There is a large number of other chemical reactions with sugars such as their union with the earthy bases, color reactions with alkalies, oxidation products with acids, and so on, which are of great use qualitively and in technological processes, but these are of little value in quantitive determinations.

48. Principles of the Method.—This method of analysis is applied almost exclusively to the examination of one kind of sugar, viz., the common sugar of commerce. This sugar is derived chiefly from sugar cane and sugar beets and is known chemically as sucrose or saccharose. The method is accurate only when applied to solutions of pure sucrose which contain no other bodies. It is evident however, that other bodies in solution can be determined by the same process, so that the principle of the method is broadly applicable to the analyses of any body whatever in a liquid state or in solution. Gases, liquids and solids, in solution, can all be determined by densimetric methods.

Broadly stated the principle of the method consists in determining the specific gravity of the liquid or solution, and thereafter taking the percentage of the body in solution from the corresponding specific gravity in a table. These tables are carefully prepared by gravimetric determinations of the bodies in solution of known densities, varying by small amounts and calculation of the percentages for the intervening increments or decrements of density. This tabulation is accomplished at definite temperatures and the process of analysis secured thereby is rapid and accurate, with pure or nearly pure solutions.

49. Determination of Density.—While not strictly correct from a physical point of view, the terms density and specific gravity are here used synonymously and refer to a direct comparison of the weights of equal volumes of pure water and of the solution in question, at the temperature named. When not otherwise stated, the temperature of the solution is assumed to be 15°.5.

Figure 27. Common Forms of Pyknometers.

The simplest method of determining the density of a solution is to get the weight of a definite volume thereof. This is conveniently accomplished by the use of a pyknometer. A pyknometer is any vessel capable of holding a definite volume of a liquid in a form suited to weighing. It may be a simple flask with a narrow neck distinctly marked, or a flask with a ground perforated stopper, which, when inserted, secures always the same volume of liquid contents. A very common form of pyknometer is one in which the central stopper carries a thermometer and the constancy of volume is secured by a side tubulure of very small or even capillary dimensions, which is closed by a ground glass cap.

The apparatus may not even be of flask form, but assume a quite different shape as in Sprengel’s tube. Pyknometers are often made to hold an even number of cubic centimeters, but the only advantage of this is in the ease of calculation which it secures. As a rule, it will be found necessary to calibrate even these, and then the apparent advantage will be easily lost. A flask which is graduated to hold fifty cubic centimeters, may, in a few years, change its volume at least slightly, due to molecular changes in the glass. Some of the different forms of pyknometers are shown in the accompanying figures.

In use the pyknometer should be filled with pure water of the desired temperature and weighed. From the total weight the tare of the flask and stopper, weighed clean and dry, is to be deducted. The remainder is the weight of the volume of water of the temperature noted, which the pyknometer holds. The weight of the solution under examination is taken in the same way and at the same temperature, and thus a direct comparison between the two liquids is secured.

The specific gravity of the sugar solution is therefore, 13.0278 ÷ 11.6342 = 1.1198.

For strictly accurate results the weight must be corrected for the volume of air displaced, or in other words, be reduced to weights in vacuo. This however is unnecessary for the ordinary operations of agricultural analysis.

If the volume of the pyknometer be desired, it can be calculated from the weight of pure water which it holds, one cubic centimeter of pure water weighing one gram at 4°.

The weights of one cubic centimeter of water at each degree of temperature from 1° to 40°, are given in the following table:

Table Showing Weights of One
Cubic Centimeter of Pure Water
at Temperatures Varying from
1° To 40°.

From the table and the weight of water found, the volume of the pyknometer is easily calculated.

Example.—Let the weight of water found be 11.72892 grams, and the temperature 20°. Then the volume of the flask is equal to 11.72892 ÷ 0.998259, viz., 11.95 cubic centimeters.

50. Use of Pyknometer at High Temperatures.—It is often found desirable to determine the density of a liquid at temperatures above that of the laboratory, e. g., at the boiling-point of water. This is easily accomplished by following the directions given below:

Weight of Flask.—Use a small pyknometer of from twenty-five to thirty cubic centimeters capacity. The stopper should be beveled to a fine edge on top and the lower end should be slightly concave to avoid any trapping of air. The flask is to be thoroughly washed with hot water, alcohol and ether, and then dried for some time at 100°. After cooling in a desiccator the weight of the flask and stopper is accurately determined.[27]

Figure 28. Bath for Pyknometers.

Weight of Water.—The flask in an appropriate holder, Fig. 28, conveniently made of galvanized iron, is filled with freshly boiled and hot distilled water and placed in a bath of pure, very hot distilled water, in such a way that it is entirely surrounded by the liquid with the exception of the top.

The water of the bath is kept in brisk ebullition for thirty minutes, any evaporation from the flask being replaced by the addition of boiling distilled water. The stopper should be kept for a few minutes before use in hot distilled water and is then inserted, the flask removed, wiped dry, and, after it is nearly cooled to room temperature, placed in the balance and weighed when balance temperature is reached. A convenient size of holder will enable the analyst to use eight or ten flasks at once. The temperature at which water boils in each locality may also be determined; but unless at very high altitudes, or on days of unusual barometric disturbance the variations will not be great, and will not appreciably affect the results.

51. Alternate Method of Estimating the Weight of Water in Flasks.—Formulas for calculating the volume V, in cubic centimeters, of a glass vessel from the weight P of water at the temperature t contained therein, and the volume Vʹ at any other temperature t’ are given by Landolt and Börnstein.[28] They are as follows:

in which p = weight (in brass weights) of one cubic centimeter H₂O in vacuo. This is so nearly one gram that it will not affect the result in the fifth place of decimals and may therefore be disregarded. Hence the formula stands:

d = density of water at temperature t.

γ = 0.000025, the cubical expansion coefficient of glass.

From this volume the weight of the water may be readily obtained by referring to tables 13, 14 and 15a in Landolt and Börnstein’s book.

52. Example Showing Determination of Specific Gravity of a Fat.—The flask is emptied of its water, rinsed with alcohol and ether, and dried again for a few minutes at 100°. It is then filled with the dry, hot, fresh-filtered fat, which should be entirely free from air bubbles.

The stoppered flask is then replaced in the water-bath, kept for thirty minutes at the temperature of boiling water, removed, and treated as above. The weight of fat having been determined, the specific gravity is obtained by dividing it by the weight of water previously found.

The weight of the flask dry and empty and the weight of water at 99° to 100° contained therein may be used constantly if great care be taken in handling and cleaning the apparatus.

Figure 29. Aereometers, Pyknometers, and Hydrostatic Balance.

53. Determination of Density by the Hydrostatic Balance.—While the pyknometer is useful in control work and in fixing standards of comparison, it is not used extensively in practical work. Quicker methods of determination are desired in such work, and these are found in the use of other forms of apparatus. A convenient method of operation consists in determining the weight of a sinker, whose exact weights in air and in pure water of a definite temperature, have been previously determined. The instrument devised by Mohr and modified by Westphal, is based upon that principle, and is extensively used in practical work. The construction of this apparatus and also that of the pyknometers and areometers is shown in the illustrations, figures 29 and 30.

Figure 30. Hydrostatic Balance.

The weight of the sinker is so adjusted that the index of the balance arm marks zero when the sinker is wholly immersed in pure water at the standard temperature. The density of a solution of sugar at the same temperature, is then determined by placing the rider-weights on the divided arm of the balance, until the index again marks zero. The density can then be read directly from the position of the weights in the arm of the balance or calculated therefrom.

54. The Areometric Method.—The most rapid method of determining the density of a solution and the one in most common use, is based on the distance to which a heavy bulb with a slender graduated stem will sink therein. An instrument of this kind is called an areometer. Many forms of this instrument are employed but they all depend on the same principle and differ only in the manner of graduation. The one of widest application has the stem graduated in such a manner as to give directly the specific gravity of the solution in which it is placed.

Others are made with a special graduation giving directly the percentage of solid matter in the solution. These instruments can be used only for the special purposes for which they are constructed. Other forms are provided with an arbitrary graduation, the numbers of which by appropriate tables can be converted into expressions of specific gravity or of per cents of dissolved matters. It is not practicable to give here, a discussion of the principles of the construction of areometers.[29] The two which are commonly used, are the baumé hydrometer and the balling or brix spindle.

In the baumé instrument the zero of the scale is fixed at the point marked by the surface of distilled water at 15°, and the point to which it sinks in pure monohydrated sulfuric acid at the same temperature is marked 66, corresponding to a specific gravity of 1.8427.

The specific gravity corresponding to any degree of the scale, may be calculated in the absence of a table giving it, by the following formula

In this formula P is the density and d the degree of the scale.[30] In former times the baumé instruments were graduated with a solution of common salt and a different formula was employed for calculating specific gravity, but these older instruments are no longer in common use.

The following table shows the specific gravities of solutions corresponding to baumé degrees from 1° to 75° consecutively[31]:

55. Correction for Temperature.—The baumé hydrometer should be used at the temperature for which it is graduated, usually 15°. In this country the mean temperature of our working rooms is above 15°. The liquid in the hydrometer flask should therefore be cooled to a trifle below 15°, or kept in a bath exactly at 15° while the observation is made. When this is not convenient, the observation may be made at any temperature, and the reading corrected as follows: When the temperature is above 15° multiply the difference between the observed temperature and fifteen, by 0.0471 and add the product to the observed reading of the baumé hydrometer; when the temperature on the other hand, is below fifteen, the corresponding product is subtracted.[32]

56. The Balling or Brix Hydrometer.—The object of the balling or brix instrument is to give in direct percentages the solid matter in solution. It is evident that for this purpose the instrument must be graduated for a particular kind of material, since ten per cent of sugar in solution, might have a very different specific gravity from a similar quantity of another body. Instruments of this kind graduated for pure sugar, find a large use in technical sugar analysis. To attain a greater accuracy and avoid an instrument with too long a stem, the brix hydrometers are made in sets. A convenient arrangement is to have a set of three graduated as follows; one from 0° to 30°, one from 25° to 50°, and one from 45° to 85°. When the percentage of solid matter dissolved is over seventy the readings of the scale are not very reliable.

57. Correction for Temperature.—The brix as the baumé scale is graduated at a fixed temperature. This temperature is usually 17°.5. The following table shows the corrections to be applied to the scale reading when made at any other temperature:[33]

Per Cent of Sugar in Solution.

According to observations of Gerlach, the correction for temperature varies with the concentration of the solution and the range of temperature as shown in the table.

58. Comparison of Brix and Baumé Degrees.—The following table shows the degree baumé and the specific gravity of a sugar solution for each degree brix (per cent of sugar in solution) from zero to ninety-five:[34]

59. Error Due to Impurities.—The fact that equal per cents of solid bodies in solution affect the specific gravity in different degrees has already been noted. The specific gravities of the solutions of the common sugars, however, are so nearly the same for equal per cents of solid matter in solution as to render the use of a brix hydrometer quite general for technical purpose. For the mineral salts which often occur in sugar solutions the case is quite different. A twenty per cent solution of cane sugar at 17°.5 has a specific gravity 1.08329 and of dextrose 1.08310, practically identical. But a solution of calcium acetate of similar strength has a specific gravity of 1.0874; of sodium sulfate 1.0807, and of potassium nitrate 1.1359. This latter number would correspond to a sugar content of nearly twenty-seven per cent. The brix scale can, therefore, be regarded as giving only approximately the percentage of solid matter in sugar solutions and, while useful in technical work, should never be relied upon for exact analytical data.

60. Optical Properties of Natural Sugars.—The solutions of all natural sugars have the property of deflecting the plane of polarized light and the degree of deflection corresponds to the quantity of sugar in solution. By measuring the amplitude of the rotation produced the percentage of sugar in the solution can be determined. In order to secure accuracy in the determinations it is necessary that only one kind of sugar be present, or, if more than one, that the quantities of all but one be determined by other means, and the disturbances produced thereby in the total rotation be properly arranged. In point of fact the process in practice is applied chiefly to cane and milk sugars, both of which occur in nature in an approximately pure state. The process is also useful in determining cane sugar when mixed with other kinds, by reason of the fact that this sugar after hydrolysis by treatment with a weak acid for a long or a strong acid for a short time, definitely changes its rotating power. Since, by the same treatment, the rotating power of other sugars which may be present is only slightly altered, the total disturbance produced is approximately due to the inversion of the cane sugar.

Dextrose and maltose arising from the hydrolysis of starch may also be determined with a fair degree of accuracy by their deportment with polarized light. When a solution of natural sugars shows negative results when examined with polarized light, it is due to an admixture of two or more sugars of opposite polarizing powers in such proportions as to produce neutrality. This condition often occurs in the examination of honeys or in submitting artificial sugars to polarimetric observations. In the latter case the neutrality is caused by the tendency manifested by artificially produced sugars to form twin compounds of optically opposite qualities.

The instrument used for measuring the degree of deflection produced in a plane of polarized light is called a polariscope, polarimeter, or optical saccharimeter. For a theoretical discussion of the principles of polarization and the application of these principles in the construction of polariscopes, the reader is referred to the standard works on optics and the construction of optical instruments.[35] For the purposes of this work a description of the instruments commonly employed and the methods of using them will be sufficient.

61. Polarized Light.—When a ray of light has been repeatedly reflected from bright surfaces or when it passes through certain crystalline bodies it acquires peculiar properties and is said to be polarized.

Polarization is therefore a term applied to a phenomenon of light, in which the vibrations of the ether are supposed to be restricted to a particular form of an ellipse whose axes remain fixed in direction. If the ellipse become a straight line it is called plane polarization. This well-known phenomenon is most easily produced by a nicol prism, consisting of a cut crystal of calcium carbonate (Iceland spar). This rhombohedral crystal, the natural ends of which form angles of 71° and 109°, respectively, with the opposite edges of its principal section, is prepared as follows:

The ends of the crystals are ground until the angles just mentioned become 68° and 112°. The crystal is then divided diagonally at right angles with the planes of the ends and with the principal section, and after the new surfaces are polished they are joined again by canada balsam. The principal section of this prism passes through the shorter diagonal of the two rhombic ends. If now a ray of light fall on one of the ends of this prism, parallel with the edge of its longer side, it suffers double refraction, and each ray is plane polarized, the one at right angles with the other. That part of the entering ray of light which is most refracted is called the ordinary and the other the extraordinary ray. The refractive index of the film of balsam being intermediate between those of the rays, permits the total reflection of the ordinary ray, which, passing to the blackened sides of the prism, is absorbed. The extraordinary ray passes the film of balsam without deviation and emerges from the prism in a direction parallel with the incident ray, having, however, only half of its luminous intensity.

Two such prisms, properly mounted, furnish the essential parts of a polarizing apparatus. They are called the polarizer and the analyzer, respectively.

If now the plane of vibration in each prism be regarded as coincident with its principal section, the following phenomena are observed: If the prisms are so placed that the principal sections lie in the prolongation of the same plane, then the extraordinary polarized ray from the polarizer passes into the analyzer, which practically may be regarded in this position as a continuation of the same prism. It happens, therefore, that the extraordinary polarized ray passes through the analyzer exactly as it did through the polarizer, and is not reflected by the film of balsam, but emerges from the analyzer in seemingly the same condition as from the polarizer. If now the analyzer be rotated 180°, bringing the principal section again in the same plane, the same phenomenon is observed. But if the rotation be in either direction only 90°, then the polarized ray from the first prism, incident on the second, deports itself exactly as the ordinary ray, and on meeting the film of balsam is totally reflected. The field of vision, therefore, is perfectly dark.

In all other inclinations of the planes of the principal sections of the two prisms the ray incident in the analyzer is separated into two, an ordinary and extraordinary, varying in luminous intensity in proportion to the square of the cosine of the angle of the two planes.

Figure 31. Course
of Rays of Light
In a Nicol.

Thus, by gradually turning the analyzer, the field of vision passes slowly from maximum luminosity to complete obscurity. The expression crossed nicols refers to the latter condition of the field of vision.

62. Description of the Prism.—In a nicol made as described above, Fig. 31, suppose a ray of light parallel with the longer side of the prism be incident to the end a b at m. By the double refracting power of the spar the ray is divided into two, which traverse the first half of the prism. The two rays are polarized at right angles to one another. The less refracted ray when it strikes the film of Canada balsam passes through it without interference. The more refracted ray strikes the balsam at o at such an angle as to be totally reflected and made to pass out of the prism in the direction o r. If the prism be blackened at the surface the ray will be entirely absorbed. The other ray passes on through the other half of the prism and emerges in the direction of qs. It is evident that the emergent light from a nicol has only half the illuminating power possessed by the immergent rays.

The polarized plane of light from the nicol just described may be regarded as passing also into a second nicol of essentially the same construction as the first.

This second nicol, called the analyzer, is so constructed as to revolve freely about its longitudinal axis, and is attached to a graduated circle in such a way that the degree of rotation can be accurately read. If the planes of polarization of the two nicols are coincident when prolonged, the ray of light passing from the first nicol will pass through the second practically unchanged in character or intensity. If, however, the analyzing nicol be turned until the plane of polarization is at right angles to that of the polarizer the immergent ray will suffer refraction in such a manner as to be totally reflected when reaching the film of balsam and will be thus entirely lost. In making a complete revolution of the analyzer, therefore, two positions of maximum intensity of light and two of darkness will be observed. In intermediate positions the ray immergent to the analyzer will be separated as in the first instance into two rays g p varying intensities, one of which will be always totally reflected.

Figure 32. Theory of
the Nicol.

In Fig. 32 is given a more detailed illustration of the action of the rays of light. The film of balsam is represented as enlarged and of the thickness bb. Draw the perpendiculars represented by the dotted lines n₁ nʹ₁, n₂ nʹ₂, n₃ nʹ₃ and n₄ nʹ₄. In passing into the prism at m both refracted rays are bent towards the normal m nʹ₁. The degree of deflection depends on the refractive index of the two rays 1.52 and 1.66 respectively. The refractive index of the extraordinary ray in calcspar being 1.52, and in Canada balsam 1.54, it suffers but little disturbance in passing from one to the other. On the other hand the balsam, being considerably less refractive for the ordinary ray than the calcspar, causes that ray to diverge outwards from the normal o nʹ₂, and to such a degree as to suffer total reflection. The critical angle, that is the angle at which a ray issuing from a more refractive into a less refractive medium, emerges just parallel to the bounding surfaces, depends on the relative index of refraction. In the case under consideration the ratio for balsam and spar is 1.54/1.66 = 0.928 = sin 68°. Therefore the limiting value of m o n₃ so that m o may just emerge in the direction od is 68°. If now mo were parallel to o d the angle m o n, would be just 68°, being opposite b a d. which has been ground to 68° in the construction of the prism. But in passing into the prism, m o is refracted so that the angle m o n₃ is greater than b a d. It is therefore always certain that by grinding b a d to 68° the ordinary ray m o will be with certainty entirely thrown out in every case. In respect of the analyzing nicol the following additional observations will be found useful. In all uniaxial crystals there are two directions at right angles to each other, one of greatest and one of least resistance to the propagation of luminous vibrations. These planes are in the direction of the principal axis and at right angles thereto. Only light vibrating in these two directions can be transmitted through calcspar; and all incident light propagated by vibrations in a plane at any other angle to the principal section is resolved into two such component rays. But the velocities of transmission in the two directions are unequal, that is, the refractive index of the spar for the two rays is different. If the analyzing nicol be so adjusted as to receive the emergent light from the polarizer when the corresponding planes of the two prisms are coincident when extended, the emergent extraordinary ray falling into a plane of the same resistance as that it had just left is propagated through the second nicol with the same velocity that it passed the first one. It is therefore similarly refracted. If, however, the two prisms be so arranged that corresponding planes cross then the extraordinary ray falls into a plane which it traverses with greater velocity than it had before and is accordingly refracted and takes the course which ends in total reflection at the film of balsam. No light therefore can pass through the prism in that position. If any other substance, as for instance a solution of sugar, capable of rotating a plane of polarized light, be interposed between the two nicols the effect produced is the same as if the analyzer had been turned to a corresponding degree. When the analyzer is turned to that degree the corresponding planes again coincide and the light passes. This is the principle on which the construction of all polarizing instruments is based.[36]

63. The Polariscope.—A polariscope for the examination of solutions of sugar consists essentially of a prism for polarizing the light, called a nicol, a tube of definite length for holding the sugar solution, a second nicol made movable on its axis for adjustment to the degree of rotation and a graduated arc for measuring it. Instead of having the second nicol movable, many instruments have an adjusting wedge of quartz of opposite polarizing power to the sugar, by means of which the displacement produced on the polarized plane is corrected. A graduated scale and vernier serve to measure the movement of the wedges and give in certain conditions the desired reading of the percentage of sugar present. Among the multitude of instruments which have been devised for analytical purposes, only three will be found in common use, and the scope of this volume will not allow space for a description of a greater number. For a practical discussion of the principles of polarization and their application to optical saccharimetry, the reader may conveniently refer to the excellent manuals of Sidersky, Tucker, Landolt, and Wiechmann.[37]

64. Kinds of Polariscopes.—The simplest form of a polarizing apparatus consists of two nicol prisms, one of which, viz., the analyzer, is capable of rotation about its long axis. The prolongation of this axis is continuous with that of the other prism, viz., the polarizer. The two prisms are sufficiently removed from each other to allow of the interposition of the polarizing body whose rotatory power is to be measured.

For purposes of description three kinds of polarimeters may be mentioned.

1. Instruments in which the deviation of the plane of polarization is measured by turning the analyzer about its axis.

Instruments of this kind conform to the simple type first mentioned, and are coeteris paribus the best. The Laurent, Wild, Landolt-Lippich, etc., belong to this class.

2. Instruments in which both nicols are fixed and the direction of the plane of polarized light corrected by the interposition of a wedge of a solid polarizing body.

Belonging to this class are the apparatus of Soleil, Duboscq, Scheibler, and the compensating apparatus of Schmidt and Haensch.

3. Apparatus in which the analyzer is set at a constant angle with the polarizer, and the compensation secured by varying the length or concentration of the interposed polarizing liquid.

The apparatus of Trannin belongs to this class.

65. Appearance of Field of Vision.—Polarimeters are also classified in respect of the appearance of the field of vision.

1. Tint Instruments.—The field of vision in these instruments in every position of the nicols, except that on which the plane of vibration of the polarized light is coincident with the three principal sections, is composed of two semi-disks of different colors.

2. Shadow Instruments.—The field of vision in this class of polarimeters in all except neutral positions, is composed of two semi-disks, one dark and one yellow. As the neutral position is approximated the two disks gradually assume a light yellow color, and when neutrality is reached they appear to be equally colored.

The Laurent, Schmidt and Haensch shadow and Landolt-Lippich instruments, are of this class.

3. Striated Instruments.—In this class the field of vision is striated. The lines may be tinted as in Wild’s polaristrobometer or black, as in the Duboscq and Trannin instruments. The neutral position is indicated either by the disappearance of the striae (Wild) or by the phenomenon of their becoming continuous. (Duboscq, Trannin.)

66. Character of Light Used.—Polariscopes may be further divided into two classes, based on the kind of light employed.

1. Instruments which Use Ordinary White Light.—(Oil lamp, etc.) Scheibler, Schmidt and Haensch.

2. Instruments Employing Monochromatic Light.—(Sodium flame, etc.) Laurent, Landolt-Lippich, etc.

67. Interchangeable Instruments.—Some of the instruments in common use are arranged to be used either with ordinary lamp or gas light, or with a monochromatic flame. Laurent’s polarimeter is one of this kind. The compensating instruments also may have the field of vision arranged for tints or shadows. Theoretically the best instrument would be one in which the light is purely monochromatic, the field of vision a shadow, and the compensation secured by the rotation of the second nicol.

The accuracy of an instrument depends, however, on the skill and care with which it is constructed and used. With quartz wedges properly ground and mounted, and with ordinary white light, polariscopes may be obtained which give readings as accurate as can be desired.

Since many persons are more or less affected with color-blindness, the shadow are to be preferred to the tint fields of vision.

For practical use in sugar analysis the white light is much more convenient than the monochromatic light.

For purposes of general investigation the polarimeters built on the model of the laurent are to be preferred to all others. Such instruments are not only provided with a scale which shows the percentage of sucrose in a solution, but also with a scale and vernier by means of which the angular rotation which the plane of vibration has suffered, can be accurately measured in more than one-quarter of the circle.

68. Rotation Instruments.—This instrument has already been described as one in which the extent of deviation in the plane of polarized light caused by the intervention of an optically active substance is measured by rotating one of the nicols about its axis and measuring the degree of this rotation by a vernier on a graduated arc.

In Germany these instruments are called polaristrobometers, and in France polarimètrés. In England and this country the term polariscope or polarimeter is applied without discrimination to all kinds of optical saccharimeters.

The polariscope of Mitscherlich was one of the earliest in use. It has now been entirely superseded by more modern and accurate instruments.

69. The Laurent Instrument.—A polariscope adapted by Laurent to the use of monochromatic yellow light is almost exclusively used in France and to a considerable extent in this country. In case a worker is confined to the use of a single instrument, the one just mentioned is to be recommended as the best suited to general work. It has the second nicol, called the analyzer, movable and the degree of rotation produced is secured in angular terms directly on a divided circle. The scale is graduated both in angular measurements and in per cents of sugar for a definite degree of concentration of the solution and length of observation tube. The normal solution in the laurent instrument contains 16.19 grams of pure sugar in 100 true cubic centimeters, and the length of the observation tube is 200 millimeters. Both the angular rotation and the direct percentage of sugar can be read at the same time. Great accuracy can be secured by making the readings in each of the four quadrants. The light is rendered yellow monochromatic by bringing into the flames of a double bunsen, spoons made of platinum wire, which carry fragments of fused sodium chlorid.

Figure 33. Laurent Lamp.

70. The Laurent Burner.—The theory of the illumination of the laurent burner is illustrated by the accompanying Fig. 33. The lamp consists essentially of two bunsens, surmounted by a chimney.[38] A curved spoon made of platinum gauze serves to hold the fused particles of sodium chlorid which are used to produce the yellow light. The spoon is shown at G, held by the arm F, fastened by the key P. The interior intense flame B B is surrounded by an exterior less highly colored flame A A. It is important that the optical axis of the polariscope be directed accurately upon the disk B, which is the most intense part of the illumination. The point of the spoon carrying the salt should be coincident with the prolongation of the lamp TT, so that it just strikes the edge of the blue flame. Care should be taken not to press the spoons into the interior of the flame as by so doing the intensity of the illumination is very much diminished. Great care must be observed in the position of the spoon G, and the platinum arm F being flexible, the operator with a little patience, will be enabled to properly place the spoon by bending it. Moreover, if the spoon be pressed too far into the flame, the melted particles of salt collecting in the bottom of G may drop into the lamp and occlude the orifices through which the gas enters. The light of the yellow flame produced by the lamp may be further purified by passing through a prism filled with a solution of potassium dichromate, or better, a homogeneous disk cut from a crystal of that salt.

Since the flame produced by the above device is not perfectly constant, being more intense at the moment of introducing a fresh portion of the fused salt, the author has used a lamp designed to furnish an absolutely constant flame.[39] This device which is shown in Fig. 34, is based on the principle of adding constantly a fresh portion of the salt to the flame. The flame is thus kept perfectly uniform in its intensity.

The lamp consists essentially of two wheels with platinum gauze perimeters and platinum wire spokes, driven by a clock-work D, and mounted by the supports AAʹ as shown in the figure. The sodium salt, chlorid or bromid, in dilute solution, is placed in the porcelain crucibles F, supported by BBʹ as indicated in the figure, to such a depth that the rims of the platinum wheels dip beneath the surface as they revolve. The salt is volatilized by the lamp E. By means of the crossed bands the wheels are made to revolve in opposite directions as indicated by the arrows. The solution of the salt which is taken up by the platinum net-work of the rim of the wheel, thus has time to become perfectly dry before it enters the flame and the sputtering which a moist salt would produce is avoided. At every instant, by this arrangement, a minute fresh portion of salt is introduced into the flame with the result of making a perfectly uniform light which can be used for hours without any perceptible variation. The mechanism of the apparatus is so simple that no further description is necessary. The polariscope should be so directed toward the flame as to bring into the field of vision its most luminous part. The platinum wheels are adjustable and should be so arranged as to produce between them an unbroken yellow flame. The wheels are eight centimeters in diameter and are driven at a rate to make one revolution in from six to ten minutes.

Figure 34. Lamp for Producing Constant
Monochromatic Flame.

71. Construction of Laurent’s Apparatus.—The shadow polariscope invented by Laurent is constructed as follows: The polarizer is a special nicol which is not fixed in its position, but is so arranged as to be turned through a small arc about its axis. By this device, the quantity of light passing through it can be regulated, and the apparatus is thus useful with colored solutions which are not easily cleared by any of the common bleaching agents. The greater the quantity of light admitted, however, the less delicate is the reading of the shadow produced. The plane of polarized light emergent from this prism, falls on a disk of glass half covered by a thin lamina of quartz which thus divides the field of vision into halves. It is this semi-disk of quartz which is the distinguishing feature of the apparatus.[40] The polarized light thus passes without hindrance the half field of vision which is covered by the glass only, but can not pass the quartz plate unless its axis is set in a certain way. The field of vision may be thus half dark, or both halves may be equally illuminated or equally dark according to the position of the nicol analyzer which is freely movable about its axis and carries a vernier and reading glass over a graduated circle. The field of vision in the laurent may have any of the following forms.[41] Let the polarizer be first so adjusted that the plane of polarization of the transmitted pencil of light is parallel to the axis of the plate lying in the direction A B. The two halves of the field of vision will then appear equally illuminated in every position of the analyzer. But if the polarizing nicol be inclined to AB at an angle a, the plane of polarization of the rays passing through the quartz plate will undergo deviation through an equal angle in the opposite direction.

Figure 35. Field of Vision of a Laurent Polariscope.

It happens from this, that when in the uncovered half of the field, the plane of polarization has the direction AC, in the other half it will have the direction ACʹ. When the analyzer is rotated, if its plane of polarization lie in the direction cc, the rays polarized parallel to AC will be completely extinguished and the corresponding half of the field will be dark. The opposite happens when the plane of polarization lies in the direction of cʹcʹ. When one-half of the field is thus obscured, the other suffers only a partial diminution in the intensity of its illumination. When the middle position bb is reached in the rotation of the analyzer, the illumination of the two halves is uniform, and this is the point at which the zero of the scale is reached. The slightest rotation of the analyzer to the right or left of this neutral point will cause a shadow to appear on one of the halves of the field, which by an oscillatory movement of the analyzer, seems to leap from side to side. The smaller the angle a or BAC, the more delicate will be the shading and the more accurate the observation. This angle being adjustable by the mechanism already described, should be made as small as will permit the admission of the quantity of light requisite for accurate observation.

The various pieces composing the polariscope are arranged in the following positions, beginning on the right of Fig. 36, and passing to the left, where the observer is seated.[42]

1. The lamp VV, TT, AA, or the wheel burner:

2. The lens B for condensing the rays and rendering them parallel:

3. The tube I, blackened inside to carry the lens:

4. A thin lamina E, cut from a crystal of potassium bichromate, serving to render the sodium light more monochromatic: When the saccharine liquids under examination are colored the crystal of bichromate is removed before the observation is made.

5. The polarizer R, which is rotatable through a small angle by the lever K:

6. The lever JK for rotating the tube containing the polarizer: This is operated by the rod X extending to the left.

7. Diaphragm D, half covered with a lamina of quartz.

8. Trough L for holding the observation tube: In the large instrument shown in the figure, it is more than half a meter in length and arranged to hold an observation tube 500 millimeters long.

9. Disk C, carrying divided circle and arbitrary sugar scale:

10. Mirror M, to throw the light of the lamp on the vernier of the scale:

Figure 36. Laurent Polariscope.

11. Reading glass N, carried on the same radius as the mirror and used to magnify and read the scale:

12. Device F, to regulate the zero of the instrument:

13. Tube H, carrying a nicol analyzer and ocular O for defining the field of vision: This tube is rotated by the radial arm G, carrying the mirror and reading glass.

72 Manipulation.—The lamp having been adjusted, the instrument, in a dark room, is so directed that the most luminous spot of the flame is in the line of vision. An observation tube filled with water is placed in the trough and the zero of the vernier is placed accurately on the zero of the scale. The even tint of the field of vision is then secured by adjusting the apparatus by the device number 12.

73 The Soleil-Ventzke Polariscope.—A form of polariscope giving a colored field of vision was in use in this country almost exclusively until within ten years, and is still largely employed. There are many forms of tint instruments, but the one almost exclusively used here is that mentioned. A full description of their construction and manipulation is given by Tucker.[43] By the introduction of a third rotating nicol in front of the lens next to the lamp, the sensitive tint at which the reading is made can be kept at a maximum delicacy. These instruments are capable of rendering very reliable service, especially in the hands of those who have a delicate perception of color. They are inferior, however, to the shadow instruments in delicacy, and are more trying to the eye of the observer. The shadow instruments therefore, especially those making use of an ordinary kerosene lamp, have practically driven the tint polariscopes out of use.

The general arrangement of a tint instrument as modified by Scheibler is shown in Fig. 37.

Figure 37. Tint Polariscope.

Beginning on the right of the figures, its optical parts are as follows: A is a nicol which, with the quartz plate B, forms the apparatus for producing the light rose neutral tint. The proper degree of rotation of these two parts is secured by means of the button L attached to the rod carrying the ratchet wheel as shown. The polarizing nicol is at C, and D is a quartz disk, one-half of which is right-handed and the other left-handed. At G is another quartz plate belonging to the analyzing part of the apparatus. This, together with the fixed quartz wedge F, and the movable quartz wedge E, constitute the compensating apparatus of the instrument whereby the deviation produced in the plane of polarized light by the solution in the tube is restored.

Next to the compensating apparatus is the analyzing nicol which in this instrument is fixed in a certain place, viz., the zero of the scale. The analyzer and the telescope for observing the field of vision are carried in the tube HJ. The movable quartz wedge has a scale which is read with a telescope K, provided with a mirror inclined at an angle of 45°, just over the scale and serving to illuminate it. The quartz wedges are also provided with a movement by which the zero point of the scale can be adjusted. A kerosene lamp with two flat wicks is the best source of illumination and the instrument should be used in a dark room and the light of the lamp, save that which passes through the polariscope, be suppressed by a shade. The sensitive or transition tint is produced by that position of the regulating apparatus which gives a field of view of such a nature that a given small movement of the quartz compensating wedge gives the greatest contrast in color between the halves of the field of vision. For most eyes a faint rose-purple tint, as nearly colorless as possible, possesses this quality. A slight movement of the quartz wedge by means of the screw head M will, with this tint, produce on one side a faint green and on the other a pink color, which are in strong contrast. The neutral point is reached by so adjusting the quartz wedge as to give to both halves of the field the same faint rose-purple tint.

74. The Shadow Polariscope for Lamp Light.—This form of instrument is now in general use for saccharimetric purposes. It possesses on the one hand, the advantages of those instruments using monochromatic light, and on the other, the ease of manipulation possessed by the tint polariscopes. It differs from the tint instrument in dispensing with the nicol and quartz plate used to regulate the sensitive tint, and in having its polarizing nicol peculiarly constructed in harmony with the optical principles of the jellet-corny prism. The more improved forms of the apparatus have a double quartz wedge compensation. The two wedges are of opposite optical properties, and serve to make the observations more accurate by mutual correction. The optical arrangement of the different parts of such a polariscope is shown in the following figure.

The lenses for concentrating the rays of light and rendering them parallel are contained in the tube N. At O is placed the modified polarizing nicol. The two compensating quartz wedges are moved by the milled screw-heads EG. The rest of the optical apparatus is arranged as described under the tint polariscope. For practical purposes, only one of the wedges is employed, but for all accurate work the readings should be made with both wedges and thus every possible source of error eliminated.

Figure 38. Double Compensating Shadow Polariscope.

75. The Triple Shadow Instrument.—When properly made, all the instruments which have been mentioned, are capable of giving accurate results if used in harmony with the directions given. In the use of polariscopes having colored fields of vision a delicate sense of distinguishing between related tints is necessary to good work. Color-blind observers could not successfully use such apparatus. In the shadow instruments it is only necessary to distinguish between the halves of a field of vision unequally illuminated and to reduce this inequality to zero. A neutral field is thus secured of one intensity of illumination and of only one color, usually yellow. Such a field of vision permits of the easy discrimination between the intensity of the coloration of its two halves, and is consequently not trying to the eye of the observer, and allows of great accuracy of discrimination. This field of vision has lately been still further improved by dividing it into three parts instead of two. An instrument of this kind, Fig. 39, in use in this laboratory, permits a delicacy of reading not possessed by any other instrument used for sugar analysis, and approaching that of the standard Landolt-Lippich apparatus, used by us for research work and for determining the rotation of quartz plates and testing the accuracy of other polariscopes.

Figure 39. Triple Shadow Polariscope.

The triple shadow is secured by interposing in front of the polarizing nicol two small nicols as indicated in Fig. 40. The end views in different positions of the polarizer are shown in the lower part of the diagrams.

Figure 40. Apparatus for Producing a Triple Shadow.

Instead of the comparison of the intensity of the illumination being made on the halves of the field of vision it is made by comparing the segments of the halves with a central band, which also changes in intensity synchronously with the two segments, but in an opposite direction.

76. General Principles.—Having described the instruments chiefly employed in the optical examination of sugar solutions, the next step is to apply them to the analytical work. A common set of directions for use will be found applicable to all instruments with such modifications only as are required by peculiarities of construction. With the best made instruments it is always advisable to have some method of controlling the accuracy of the observation. The simplest way of doing this is to test the apparatus by standard quartz plates. These plates are made from right-handed polarizing quartz crystal ground into plates of definite thickness and accurately tested by standard instruments. Theoretically such quartz plates deflect the plane of polarized light in a degree proportionate to their thickness, but practically some small deviations from the rule are found. With a source of light of the same tint, and at a constant temperature, such plates become a safe test for the accuracy of the graduation of polariscopes. They are more convenient for use than pure sugar solutions of known strength which are the final standards in all disputed cases. These quartz plates are conveniently mounted in tubes of the same size as those holding the sugar solution, and thus fit accurately into the trough of the polariscope, the optical axis of which passes through their center. The quartz plate when used for setting the scale of a polariscope should be placed always in the same position. In some plates slight differences of reading may be noticed on rotating the tubes holding them. Theoretically, such differences should not exist, but in practice they are sometimes found. The temperature of observation should also be noted, and if not that at which the value of the plate was fixed a proper correction should be made.

77. Setting the Polariscope.—While mention has been made of several forms of apparatus in the preceding paragraphs, those in common use are limited to a very small number. In this country quite a number of color instruments may still be found, together with a few laurents, and a constantly increasing number of shadow instruments for use with lamp light. The following description of setting the polariscope is especially adapted to the last named instrument, but the principles of adjustment are equally applicable to all.

The scale of the instrument is first so adjusted by means of the adjusting screws provided with each instrument, as to bring the zero of the vernier and that of the scale exactly together. The telescope or ocular is then adjusted until the sharp line separating the halves of the field of vision is brought into focus. This being accomplished an observation tube filled with pure water is placed in the apparatus and the telescope again adjusted to bring the dividing line of the field into focus. The beginner especially, should repeatedly study this adjustment and be impressed with the fact that only in a sharply defined field are practical observations of any worth. The importance of having all the lenses perfect and all the cover glasses without a flaw may be fully appreciated when it is remembered that the polarized ray, already deprived of half its original luminous power, must pass through several centimeters of crystallized calcium carbonate, and half a dozen disks of glass and quartz, and as many lenses before reaching the eye of the observer. Only with the greatest care and neatness is it possible to secure the required degree of illumination. The zero point having been well studied and accurately adjusted, the scale of the instrument may be tried with a series of quartz plates of known polarizing power at the temperature of the observation. In the apparatus with double quartz wedge compensation, it will be noticed that the marks on one scale are black and on the other red. The black is the working and the red the control scale. To operate this instrument, the red scale is placed exactly at the zero point. The black scale is also placed at zero, and if the field of vision is not neutral, it is made so by the micrometer screw with which the black scale is provided. In a right-handed solution, the red scale is left at zero and the black one moved to the right until neutrality in the field of vision is reached and the reading is taken. The observation tube containing the sugar solution is taken out and the red scale moved until the field of vision is again neutral and the reading of the red scale taken. The two readings should agree. Any failure in the agreement shows some fault either in adjusting the apparatus or in its construction, or some error in manipulation.

The double compensating shadow instruments are more readily tested for accuracy in all parts of the scale than those of any other construction. The two compensating wedges are cut with the greatest care, one from a left-handed and the other from a right-handed perfectly homogeneous quartz crystal. Since faults in these wedges are due either to lack of parallelism of surface, or of perpendicularity to the optical axis of the crystal, and since these faults of crystallization or construction must be in a very limited degree common they would not coincide once in many thousand times in the two wedges. This is easily shown by the theory of probabilities. If, therefore, the two readings made at any point, should not agree, it must be due either to a fault in one of the wedges, or to a fault in reading or a lack of adjustment, as has been mentioned. In such cases the readings should be retaken and the errors are usually easily discovered.

78. Control Observation Tube.—Instead of using quartz plates of known values for testing the accuracy of the scale, an observation tube may be used, the length of which can be varied at the pleasure of the observer.

The construction of a tube of this kind is shown in Fig. 40. The tube B is movable telescopically in A by means of the ratchet wheel shown. It is closed at D water-tight by a glass disk. The tube B fits as accurately into A as is possible to permit of free movement, and any liquid which may infilter between its outer surface and the inner surface of A is prevented from gaining exit by the washer C, which fits both tubes water-tight. The ratchet which moves B in A carries a millimeter scale and vernier N whereby the exact thickness of the liquid solution between the surfaces of the glass disks D and E can be always determined.

Figure 41. Control Observation Tube.

By this device the length of liquid under observation can be accurately read to a tenth of a millimeter. The cover glass E is held in position by any one of the devices in common use for this purpose in the case in question, by a bayonet fastening. The funnel T, communicating directly with the interior of A, serves to hold the solution, there being always enough of it to fill the tube when D is removed to the maximum distance from C, which is usually a little more than 200 millimeters.

Let the control tube be adjusted to 200 millimeters and filled with a solution of pure sugar, which reads 100 per cent or degrees in a 200 millimeter tube. Since the degree of rotation is, other things being equal, proportional to the length of the column of polarizing solution, it follows that if the tube B be moved inward until the distance between D and C is 100 millimeters, the scale should read 50° or per cent. By adjusting the length of the distance between B and C it is easily seen that every part of the scale can be accurately tested.

The tube should be filled by removing the funnel and closing the orifice with a screw cap which comes with the apparatus. The cap E is then removed and the tube filled in the ordinary manner. This precaution is practiced to avoid carrying air bubbles into the tube when filled directly through the funnel. With a little care, however, this danger may be avoided, or should air bubbles enter they can be easily removed by inclining the tube.

In case the solution used be not strictly pure it may still be employed for testing the scale. Suppose, for instance, that a solution made up in the usual way, has been made from a sample containing only 99.4 per cent of sugar. Then in order to have this solution read 100° on the scale the tube should be set at 201.2 millimeters, according to the formula

By a similar calculation the position of the tube for reading any desired degree on the scale can be determined. The importance of controlling all parts of the scale in compensating instruments is emphasized by the fact that a variation of only 0.016 millimeter in the thickness of the compensating wedge will cause a change of one degree in the reading of the instrument.

79. Setting the Polariscope with Quartz Plates.—Pure sugar is not always at the command of the analyst, and it is more convenient practically to adjust the instrument by means of quartz plates, the sugar values of which have been previously tested for the character of the light used. Assuming the homogeneity of a plate of quartz, the degree of deflection which it imparts to a plane of polarized light depends on the quality of the light, the thickness of the plate, and the temperature.

In respect of the quality of light, red polarized rays are least, and violet most deflected. The degree of rotation produced with any ray, at a given temperature, is directly proportional to the thickness of the plate. Temperature affects the rotating power of a quartz plate in a degree highly significant from a scientific point of view and not wholly negligible for practical purposes. The rotating power of a quartz plate increases with the temperature and the variation may be determined by the formula given below:[44]

The formula is applicable for temperatures between 0° and 100°. Its values are expressed in degrees of angular measure which can be converted into degrees of the sugar scale by appropriate factors:

Formula.— aᵗ = a°(1 + 0.000146t);

in which a° = polarization in angular degrees at 0°, t the temperature of observation and aᵗ the rotation desired.

Example.—A quartz plate which has an angular rotation of 33° at 0° will have a rotation at 20° of 33°.09834.

aᵗ = 33(1 + 0.000146 × 20) = 33.09834.

Since in instruments using the ventzke scale one degree of the sugar scale is equal to 0.3467 degree angular measure, the sugar value of the quartz plate mentioned is equal to 95.47 percent; 33.09834 ÷ 0.3467 = 95.47.

The sugar value of this plate at 0° is 95.18 per cent; 33 ÷ 0.3467 = 95.18.

80. Tables for Correcting Quartz Plates.—Instead of calculating the variation in quartz plates for each temperature of observation, it is recommended by the Bureau of Internal Revenue of the Treasury, to use control quartz plates the values of which at any given temperature, are found on a card which accompanies each one.[45] The variations given, are from temperatures between 10° and 35°. Three control plates are provided with each instrument used by the Bureau, for polarimetric work in the custom houses, or in ascertaining bounties to be paid on the production of domestic sugars. For example, the case of a sugar which polarizes 80°.5 may be cited. One of the control plates nearest to this number, is found to have at the temperature of observation, a polarization of 91°.4, the reading being made in each case at 25°. On consulting the card which accompanies the control plate, it is seen that its value at the temperature mentioned, is 91°.7. The reading of the instrument is therefore too low by three-tenths of a degree, and this quantity should be added to the observed polarization, making it 80°.8. In this method of correcting the reading for temperature, it is assumed that the compensating wedges of the instrument, are free of error at the points of observation. The plates used for the purpose above, are all standardized in the office of weights and measures of the Coast and Geodetic Survey, before delivery to the analysts.

81. Applicability of Quartz Plates.—Quartz plates which are correctly set for one instrument or kind of light, should be equally accurate for the sugar scales of all instruments, using the same sugar factor. In other words a quartz plate which reads 99° on a scheibler color polariscope, should give the same reading on the sugar scale of a shadow compensating or a monochromatic direct reading apparatus using 26.048 grams of sugar.

The most useful quartz plates for sugar analysis, are those which give the readings at points between 80° and 96°, which cover the limits of ordinary commercial sugars. For molasses the plates should read from 45° to 55°. For sugar juices of the cane and beet, the most convenient graduation would be from 10° to 20°, but plates of this value would be too thin for practical work and are not in use. When quartz plates are to be used for control purposes, they should be purchased from reliable manufacturers, or better, tested directly against pure sugar solutions by the observer.

In practice we have found quartz plates as a rule, true to their markings.

82. The Sugar Flask.—Sugar solutions are prepared for polarization in flasks graduated to hold fifty or one hundred cubic centimeters. For scientific work a flask is marked to hold 100 grams of distilled water at 4°. The weights are all to be reduced to a vacuum standard. One flask having been marked in this way, others may be compared directly therewith by means of pure mercury. For this purpose the flasks must be perfectly dry and the mercury pure, leaving no stain on the sides of the flask. The glass must also be strong enough to undergo no change in shape from the weight of mercury used.

For sugar work the true 100 gram flask is not usually employed, but one graduated by weighing at 17°.5. These flasks are graduated by first weighing them perfectly dry, filling with distilled water and again weighing fifty and fifty-five, or 100 and 110 grams of water at the temperature named. Since the volume of water at 17°.5 is greater than at 4° the sugar flask in ordinary use has a greater volume by about 0.25 cubic centimeter than the true flask. The observer should always secure a statement from the dealer in respect of the volume of the flask used in testing the scale of the polariscope purchased. In the graduation of a flask in true cubic centimeters, when brass weights are used it will be necessary to correct the weight of each gram of water by adding to it one milligram, which is almost exactly the weight of the volume of air displaced by one gram of water in the circumstances named. If the flask be first counterbalanced and it be desired to mark it at 100 cubic centimeters the sum of the weights placed in the opposite pan should be 100 - 0.100 = 99.900 grams. While this is not a rigidly exact correction it will be sufficient for all practical purposes. A liter of dry air weighs 1.29366 grams; and 100 cubic centimeters of water would therefore displace 0.129 gram of air. But the brass weights also displace a volume of air which when deducted reduces the correction to be made for the water to nearly the one named. For convenience in inverting sugar solutions the flasks used in practical work are graduated at fifty and fifty-five and 100 and 110 cubic centimeters respectively.

83. Preparing Sugar Solutions for Polarization.—If sugar samples were always pure the percentage of sugar in a given solution could be directly determined by immediate polarization. Such cases, however, are rarely met in practice. In the majority of cases the sample is not only to be brought into solution but is also to be decolorized and rendered limpid by some one of the methods to be described. A perfectly limpid liquid is of the highest importance to secure correct observations. With a cloudy solution the field of vision is obscured, the dividing line of the two halves, or the double line in the triple field, becomes blurred or invisible and the intensity of illumination is diminished. A colored liquid which is bright is far more easy to polarize than a colorless liquid which is turbid. In fact, it is only rarely in sugar work that samples will be found which require any special decolorizing treatment other than that which is received in applying the reagents which serve to make the solutions limpid. In the following paragraphs the approved methods of clarifying sugar solutions preparatory to observation in the polariscope will be described.

84. Alumina Cream.—The hydrate of alumina, commonly known as alumina cream, is always to be preferred as a clarifying agent in all cases where it can be successfully applied.[46] It is a substance that acts wholly in a mechanical way and therefore leaves the sugars in solution unchanged, carrying out only suspended matters. In the preparation of this reagent a solution of alum is treated with ammonia in slight excess, the aluminum hydroxid produced washed on a filter or by decantation until neutral in reaction. The hydroxid is suspended in pure water in proportions to produce a creamy liquid. Although apparently very bulky, the actual space occupied by the amount of dry hydroxid added in a few cubic centimeters is so small as to produce no disturbing effect of importance on the volume of the sugar solution. The cream thus prepared is shaken just before using and from one to five cubic centimeters of it, according to the degree of turbidity of the saccharine solution, are added before the volume in the flask is completed to the mark. After filling the flask to the mark the ball of the thumb is placed over the mouth and the contents well shaken and allowed to stand for a few moments before filtering.

The alumina cream is well suited to use with solutions of commercial sugars of not too low a grade and of most honeys and high grade sirups. It is usually not powerful enough to clarify beet and cane juices, molasses and massecuites.

85. Basic Lead Acetate.—A solution of basic lead acetate is an invaluable aid to the sugar analyst in the preparation of samples for polarimetric observation. It acts as a clarifying agent by throwing out of solution certain organic compounds and by uniting with the organic acids in solution forms an additional quantity of precipitate, and these precipitates act also mechanically in removing suspended matters from solution. The action of this reagent is therefore much more vigorous than that of alumina cream. Coloring matters are often precipitated and removed by treatment with lead acetate. It happens therefore that there are few samples of saccharine bodies whose solutions cannot be sufficiently clarified by lead acetate to permit of polarimetric observation.

The reagent is most frequently employed of the following strength:[47] Boil for half an hour in one and a half liters of water 464 grams of lead acetate and 264 grams of litharge with frequent stirring. When cool, dilute with water to two liters, allow to stand until clear, and decant the solution. The specific gravity of this solution is about 1.267.

In a solution of basic lead acetate of unknown strength the percentage of lead acetate may be determined from its specific gravity by the following table:[48]

Percentage of Lead Acetate Corresponding
to Different Specific Gravities at 15°.

86. Errors Due to use of Lead Solutions.—In the use of lead solutions there is danger of errors intruding into the results of the work. These errors are due to various sources. Lead subacetate solution, when used with low grade products, or sugar juices, or sirups from beets and canes, precipitates albuminous matters and also the organic acids present. The bulk occupied by these combined precipitates is often of considerable magnitude, so that on completing the volume in the flask the actual sugar solution present is less than indicated. The resulting condensation tends to give too high a polarimetric reading. With purer samples this error is of no consequence, but especially with low grade sirups and molasses it is a disturbing factor, which must be considered.

One of the best methods of correcting it has been proposed by Scheibler.[49] To 100 cubic centimeters of a solution of the sample, ten of lead solution are added, and after shaking and filtering the polarimetric reading is taken. Another quantity of 100 cubic centimeters of the solution with ten of lead is diluted to 220 cubic centimeters, shaken, filtered, and polarized. Double the second reading, subtract it from the first, multiply the difference by 2.2, and deduct the product from the first reading. The remainder is the correct polarization.

The process just described is for the usual work with beet juices and sirups. For cane juices measured by the graduated pipette, hereafter to be described, and for weighed samples of molasses and massecuites, the following method of calculation is pursued.[50] To the sample dissolved in water, add a measured portion of the lead subacetate solution, make its volume 100 cubic centimeters and observe the polarimetric reading. Prepare a second solution in the same way and make the volume double that of the first and again take the polarimetric reading. Multiply the second reading by two, subtract the product from the first reading and multiply the remainder by two, and subtract the product from the first reading.

The corrected reading therefore shows that the sample contained 29.6 per cent of sugar.

87. Error Due to Action of Lead Subacetate on Levulose.—In the use of lead subacetate solution not only is there danger of error due to the causes just described, but also to a more serious one, arising from the chemical interaction of the clarifying agent and levulose.[51]

Lead subacetate forms a chemical union with levulose and the resulting compound has a different rotatory power from the left-handed sugar in an uncombined state. By adding a sufficient quantity of subacetate solution, the left-handed rotation of levulose may be greatly diminished if not entirely destroyed. In this case the dextrose, which with levulose forms inverted sugar, serves to increase the apparent right rotation due to the sucrose in solution. The reading of the scale is therefore higher than would be given by the sucrose alone. If the lead subacetate could be added in just the proportion to make the invert sugar neutral to polarized light, its use would render the analysis more accurate; but such a case could only arise accidentally. To correct the error, after clarification, the compound of levulose and lead may be decomposed by the addition of acetic acid according to the method of Spencer. In this case the true content of sucrose can only be obtained by the method of inversion proposed by Clerget, which will be described in another paragraph.

88. Clarification with Mercuric Compounds.—Where the disturbing bodies in a solution are chiefly of an albuminoid nature, one of the best methods of securing clarification is by the use of a solution containing an acid mercuric compound.[52] In the case of milk this method is to be preferred to all others. Albuminoid bodies themselves, have the property of deflecting the plane of polarization, as a rule, to the left, and therefore, should be completely removed from solutions containing right-handed sugars such as lactose. For this purpose the mercuric compound is more efficient than any other. It is prepared and used as follows.[53] Dissolve mercury in double its weight of strong nitric acid and dilute the solution with an equal volume of water. One cubic centimeter of this solution is sufficient to clarify fifty times its volume of milk.

89. Decolorization by Means of Bone-Black.—Where the means already described fail to make a solution sufficiently colorless to permit of the passage of a ray of polarized light, recourse should be had to a decolorizing agent. The most efficient of these is bone-black. For laboratory work it is finely ground and should be dry if added to an already measured solution. When moist it should be added to the flask before the volume is completed, and a correction made for the volume of the dry char employed. Bone-black has the power of absorbing a certain quantity of sugar, and for this reason as little of it should be employed as is sufficient to secure the end in view. If not more than one gram of the char be used for 100 cubic centimeters of solution, the error is not important commercially. The error may be avoided by placing the char on the filter and rejecting the first half of the filtered solution. The char becomes saturated with the first portion of the solution, and does not absorb any sugar from the second. This method, however, does not secure so complete a decolorization as is effected by adding the black directly to the solution and allowing to stand for some time with frequent shaking.

90. Remarks on Analytical Process.—Since large weights of sugar are taken for polarization, a balance which will weigh accurately to one milligram may be used. In commercial work the weighing is made in a counterpoised dish with a prominent lip, by means of which the sample can be directed into the mouth of the flask after partial solution. Where the air in the working room is still, an uncovered balance is most convenient. With a little practice the analyst will be able to dissolve and transfer the sample from the dish to the flask without danger of loss. The source of light used in polarizing should be in another room, and admitted by a circular opening in the partition. In a close polarizing room, which results from the darkening of the windows, the temperature will rapidly rise if a lamp be present, endangering notably the accuracy of the work, and also interfering with the comfort of the observer. The greatest neatness must be practiced in all stages of the work, and especially the trough of the polariscope must be kept from injury which may arise from the leaking of the observation tubes. Dust and dirt of all kinds must be carefully excluded from the lenses, prisms, wedges and plates of the instrument.

91. Determination of Sucrose by Inversion.—In the foregoing paragraphs directions have been given for the estimation of sugar (sucrose) by its optical properties. It has been assumed so far, that no other disturbing bodies have been present, save those which could be removed by the clarifying agents described. The case is different when two or more sugars are present, each of which has a specific relation to polarized light. In such cases some method must be used for the optical determination of sucrose, which is independent of the influence of the other polarizing bodies, or else recourse must be had to other methods of analysis. The conversion of the sucrose present into invert sugar by the action of an acid or a ferment, affords an opportunity for the estimation of sucrose in mixed sugars, by purely optical methods. This process rests upon the principle that by the action of a dilute acid for a short time, or of a ferment for a long time, the sucrose is completely changed, while other sugars present are not sensibly affected. Neither of these assumptions is rigidly correct but each is practically applicable.

The sucrose by this process of hydrolysis is converted into an equal mixture of levulose and dextrose. The former, at room temperatures, has the higher specific rotating power, and the deflection of the plane of polarization in a solution of inverted sugar is therefore to the left. The levorotatory power of invert sugar varies with the temperature, and this arises from the optical properties of the levulose. The influence of temperature on the rotating power of other sugars, is not imperceptible in all cases, but in practice is negligible.

This method of analysis is invaluable in control work in factories, in the customs and in agricultural laboratories. Since the rotating power of levulose diminishes as the temperature rises, an accurate thermometric observation must accompany each polarimetric reading. At about 88° the rotatory powers of dextrose and levulose are equal, and a solution of pure invert sugar examined at that degree, is found to be neutral to polarized light.

92. Clerget’s Method of Inversion.—The classical method of Clerget for the determination of cane sugar by double polarization before and after inversion, was first described in a memoir presented to the Society of Encouragement for National Industry on the 14th of October, 1846. The following description of the original method is taken from a reprint of the proceedings of that Society, dated Nov. 1846:

Clerget points out first the observation of Mitscherlich regarding the influence of temperature on the rotatory power of invert sugar, and calls attention to the detailed experiments he has made which resulted in the determination of the laws of the variation. From these studies he was able to construct a table of corrections, applicable in the analysis of all saccharine substances in which the cane sugar is polarized before and after inversion. The basis of the law rests upon the observation that a solution of pure sugar, polarizing 100° on the sugar scale, before inversion, will polarize 44° to the left after inversion at a temperature of zero. The quantity of sugar operated upon by Clerget amounted to 16.471 grams in 100 cubic centimeters of liquid. On the instrument employed by him this quantity of sugar in 100 cubic centimeters gave a reading of 100° to the right on the sugar scale when contained in a tube twenty centimeters in length. The process of inversion carried on by Clerget is as follows:

The sugar solution is placed in a flask, marked on the neck at 100 and 110 cubic centimeters; or if smaller quantities are used, in a flask marked on the neck at fifty and fifty-five cubic centimeters. The flask is filled with the sugar solution to the first mark and then a sufficient quantity of strong hydrochloric acid added to bring the volume of the liquid to the second mark. The mouth of the flask is then closed with the thumb and its contents thoroughly mixed by shaking. A thermometer is placed in the flask which is set in a water-bath in such a way that the water comes just above the level of the liquid in the neck of the flask. The water is heated in such a manner as to bring the temperature of the contents of the flask, as determined by the thermometer, exactly to 68° and at such a rate as to require fifteen minutes to reach this result. At the end of fifteen minutes the temperature having reached 68° the flask is removed and placed at once in another water-bath at the temperature of the room, to which temperature the contents of the flask are cooled as rapidly as possible. To make the polarimetric observation a tube twenty-two centimeters in length is filled with the inverted sugar solution by means of a tubulure in its center, which serves not only the purpose of filling the tube but also afterwards to carry the thermometer, by means of which the temperature of observation can be taken. If the sugar solution be turbid, or contain any lead chlorid due to the previous use of basic lead acetate in clarification, it should be filtered before being introduced into the observation tube. This tube being one-tenth longer than the original compensates for the dilution caused by the addition of the hydrochloric acid in inversion.

When reading, the bulb of the thermometer should be withdrawn far enough to permit the free passage of the ray of light and the exact temperature of the solution noted.

The above outline of Clerget’s method of inversion is given in order that the analyst may compare it with any of the variations which he may find in other works. The chief points to which attention is called, are, first, the fact that only a little over sixteen grams of sugar are used for ten cubic centimeters of strong hydrochloric acid, and second, that the time of heating is exactly fifteen minutes, during which time the contents of the flask should be raised from room temperature to exactly 68°.

From the above it is seen that the process of Clerget, as originally described, can be applied directly to all instruments, using approximately sixteen grams of sugar in 100 cubic centimeters. Experience has also shown that even when larger quantities of sugar are employed, as for instance, approximately twenty-six grams, the inversion is effected with practical completeness in the same circumstances. It is advised, therefore, that in all analytical processes, in which cane sugar is to be determined by the process of inversion with an acid, the original directions of Clerget be followed as strictly as possible. Experience has shown that no one of the variations proposed for Clerget’s original method has any practical advantage and the analyst is especially cautioned against those methods of inversion in which the temperature is continued at 68° for fifteen minutes or in which it is allowed to go above that degree.

93. Influence of Strength of Solution and Time of Heating on the Inversion of Sucrose.—As has been intimated, the strength of a sugar solution and the time of heating with hydrochloric acid are factors that must be considered in determining a formula for the calculation of sucrose by inversion. The Clerget formula holds good only for the conditions specified and these conditions must be rigidly adhered to in order to secure the proper results. This matter has been thoroughly studied by Bornträger, who also gives a nearly complete bibliography of the subject.[54] As a result of his investigations it seems well established that the original Clerget formula is practically correct for the conditions indicated, Bornträger modifying it only by substituting in the formula 143.66 for 144. This is so nearly the same as the Clerget factor that it is not advisable to substitute it therefor. If, however, the inverted sugar solution be diluted to double its volume before polarization the factor proposed by Landolt, viz., 142.4, gives more nearly accurate results. If the hydrochloric acid be neutralized before polarization by an alkaline body, the character of the salt which is formed also influences, to a greater or less extent, the specific rotatory power of the solution. Hydrochloric acid itself also influences the rotation to a certain degree.[55]

94. Calculation of Results.—The percentage of sucrose in a solution which has been polarized before and after inversion is calculated by an appropriate formula from the data obtained or is taken directly from tables. These tables are too long to insert here, and in point of fact the calculation can be made from the formula almost as quickly as the result can be taken from a table.

Two factors are commonly used in the calculations, one based on the supposition that a sugar solution polarizing 100° to the right will, after inversion, give a reading of 44° to the left, at zero temperature. In the second formula in common use the polarization to the left in the circumstances mentioned above is assumed to be 42.4, a number reached by Landolt after a long series of experiments.[56] The principle of the calculation of the percentage of sucrose is based upon the original observation of Clerget to the effect that the algebraic difference of the two readings, divided by 144, less half of the temperature, will give the percentage of sucrose desired. The formula by which this is obtained is

In this formula a is the polarization on the sugar scale before inversion, b the polarization after inversion, K the constant representing the algebraic difference of the two polarizations of pure sugar at 0° and t the temperature of the observation. To K may be assigned the values 144 or 142.4, the one in more common use. In case the polarization, after inversion, is to the left, which is more commonly the case, the sum of the two readings is taken for a - (-b) = a + b; when both polarizations are to the right or left the difference is taken. S is the percentage of sucrose desired.

Substituting the value 142.4 for K, the result of the calculation is 91.4.

In high grade sugars, therefore, the difference in the results secured by taking the two values of K amounts to about 1 per cent of sucrose.

For a further discussion of the theory and practice of inversion the reader is referred to the articles of Herles, Herzfeld, and Wohl.[57]

95. Method Of Lindet.—Courtonne recommends the method of Lindet for securing the inversion instead of the method of Clerget.[58] Modified by Courtonne, the method is as follows:

Make two or three times the normal weight of sugar dissolved in water to a volume of 200 or 300 cubic centimeters, as the case may be. After thoroughly mixing proceed as follows:

First, to Obtain the Polarization Direct.—Place fifty or 100 cubic centimeters of the prepared solution in a flask marked at fifty and fifty-five or at 100 and 110 cubic centimeters, add a sufficient quantity of lead acetate to secure a complete clarification, make the volume to fifty-five or 110 cubic centimeters, shake thoroughly, filter, and polarize in a 220 millimeter tube.

Second, to Obtain the Rotation after Inversion.—Place twenty cubic centimeters of the original solution, in a flask marked at fifty cubic centimeters, containing five grams of powdered zinc. The flask should be placed in boiling water. Add, little by little so as to avoid a too rapid evolution of hydrogen, ten cubic centimeters of hydrochloric acid made of equal parts of the strongest acid and water. After the operation is terminated, cool to the temperature of the room, make the volume to fifty cubic centimeters, polarize, and determine the rotation. The volume occupied by the zinc which is not dissolved, will be about one-half cubic centimeter, hence the deviation should be multiplied by the factor 2.475 in order to get the true deviation which would have been produced by the pure liquor. We have then:

• A = the deviation direct.
• B = the deviation after inversion.
• C = the algebraic difference of the deviations.

The amount of sucrose, therefore, would be calculated by the formula of Creydt,[59]

for raffinose the formula would be

in which S is the deviation due to the sucrose present. The solutions inverted in the manner described are absolutely colorless. There is no need of employing bone-black to secure the saccharimetric reading nor does it present any uncertainty. It is thought by Courtonne that this method will soon take the place of the method of Clerget on account of the advantages above mentioned. The method will be somewhat improved by adopting the following suggestions:

1. Instead of allowing any arbitrary number for the volume of the undissolved zinc, decant the liquid, after inversion, into another flask and wash repeatedly with hot water until all trace of sugar is removed from the flask in which the inversion took place.

2. Instead of polarizing in a 200 millimeter tube make the observation in a 500 millimeter tube, which will permit of the reading being made without any correction whatever.

96. Inversion by Means of Invertase.—Instead of using acids for the inversion of cane sugar the hydrolysis can be easily effected by means of a ferment derived from yeast. A complete history of the literature and characteristics of this ferment, together with a study of its properties and the various methods of preparing it, has been given by O’Sullivan and Tompson.[60] In the preparation of invertase, the method found most effective is the following:

The yeast is allowed to liquify for at least a month in a fairly warm room without stirring. At the end of this time the surface is removed and any supernatant liquid poured away. The lower sedimentary part is thrown on a quick-acting filter and allowed to drain for two days. To the filtrate, alcohol of specific gravity 0.87 is gradually added to the extent of one and a half times its volume, with continued and vigorous stirring. The process of adding the alcohol and stirring should require about half an hour, after which the mixture is allowed to stand for twenty-four hours to allow the precipitated invertase to settle. The supernatant liquid is poured away and the precipitate washed several times on successive days by decantation with alcohol of 0.92 specific gravity. When the washings become nearly colorless the precipitate is thrown on a filter, allowed to drain, and immediately removed and mixed with a large bulk of alcohol of 0.92 specific gravity. The precipitate is again collected, mixed thoroughly with its own bulk of water, and some alcohol of 0.97 specific gravity, allowed to stand for a few hours and thrown on a filter. The filtrate contains the invertase.

97. Determination of Activity of Invertase.—The activity of a solution of invertase, prepared as above, is measured by the number of minutes required for it to reduce to zero the optical power of a solution of 100 times its weight of cane sugar at a temperature of 15°.5. In order to facilitate the action of the invertase, a trace of sulfuric acid is added to the solution. The manipulation is as follows:

Fifty grams of sucrose are dissolved in water and made to a volume of nearly a quarter of a liter and placed in a bath maintained at 15°.5. Half a gram of the invertase is added, the time noted, the solution immediately made up to a quarter of a liter and well shaken. The contents of the flask are poured rapidly into five beakers; the actual quantity in each beaker is not necessarily the same. To each of these beakers, in succession, are added the following amounts of decinormal sulfuric acid, viz., one-tenth, three-tenths, six-tenths, one, and one and four-tenths cubic centimeters. After an hour a small quantity of the solution is taken from beaker No. 3 and the reaction of the invertase stopped by adding a few drops of strong potassium hydroxid and the time of adding this reagent noted. This solution is then read in the polariscope and the percentage of sugar inverted is calculated from the formula C₁₂H₂₂O₁₁ + H₂O = C₆H₁₂O₆ + C₆H₁₂O₆.

The calculation of the amount of cane sugar inverted is based on the formula,

(38.4 - d) ÷ 0.518 = p.

In this formula d equals the divisions of the sugar scale read on the polariscope; p the percentage of cane sugar inverted; 38.4 the reading on the sugar scale of the original sugar solution and 51.8 the total number of divisions of the cane sugar scale that the polariscope reading would fall through if all the sugar were inverted. The observation tubes used in the polarization are only 100 millimeters in length. After stopping the action of the invertase with potassium hydroxid the solution is allowed to stand for some time before polarization inasmuch as the dextrose formed appears to assume the state of birotation and some time is required for it to reach its normal rotatory power. If the invertase be used in the alcoholic solution a sufficient quantity should be added to be equivalent to 0.01 of the sucrose present. The time which the contents of beaker No. 3 will take to reach optical activity is calculated in a manner described by O’Sullivan and Tompson, but too long to be inserted here.[61] The five beakers mentioned above are examined in succession and the amount of sulfuric acid best suited to the maximum inversion thus determined. This quantity is then used in subsequent hydrolyses with the given sample of invertase.

The action of invertase on sucrose is very rapid at the first and becomes very much slower towards the end. At a temperature of 15°.5 it is advisable to let the solution stand for forty-eight hours in order to be sure that complete inversion has taken place. For this reason the method by inversion by means of invertase is one of no great practical importance, but it may often be useful to the analyst when the employment of an acid is inadmissible.

98. Inversion by Yeast.—Owing to the difficulty of preparing invertase, O’Sullivan and Thompson[62] propose to use yeast as the hydrolytic agent, as first suggested by Kjeldahl. It is shown that in the use of yeast it is not necessary to employ thymol or any other antiseptic. The method of procedure is as follows: The cane sugar solution of usual strength should not be alkaline, but, if possible, should be exactly neutral. If there be any ferment suspected, the temperature should be momentarily raised to 80° to destroy its activity. The polariscopic reading of the solution is then taken at 15°.5 and the amount of copper reduced by the solution should also be determined.

Fifty cubic centimeters of the solution are poured into a beaker and raised to a temperature of 55° in a constant temperature bath. Some brewers yeast amounting to about one-tenth of the total amount of sugar to be inverted, pressed in a towel, is thrown into the hot solution and the whole stirred until mixture is complete. The solution is left for four hours in the water-bath, at the end of this time it is cooled to 15°.5, a little freshly precipitated aluminum hydroxid added, and the volume made to 100 cubic centimeters. A portion of this solution is filtered and its polariscopic reading observed. The solution is then left till the next day, when another polariscopic reading is taken in order to prove that inversion is complete. The copper reducing power is also determined. The method of calculating the results is the same as when invertase is used. The following formulas are employed.

a = the number of divisions indicated by the polariscopic reading for a 200 millimeter tube:

aʹ = the same number after inversion:

m = the number of the divisions of the polariscopic scale which 200 millimeters of the sugar solution containing one gram of cane sugar per 100, alter at 15°.5 on being inverted: In the case of the ventzke polarimeter scale, one gram of cane sugar in 100 cubic centimeters, indicates +3.84 divisions and after inversion it gives -1.34 div. In experiments of this kind, therefore, m = 5.18.

P = the weight of cane sugar present in 100 cubic centimeters of the original solution:

The formula employed then is

For the copper reduction data the following are used:

G = the weight of 100 cubic centimeters of the original solution:

Gʹ; = the same for the inverted solution: Allowance must be made here both for the dilution and for the 5 per cent increase of the inverted sugar, but the latter number is so small that it need not be calculated accurately.

w = the weight of the original solution used for the estimation:

wʹ = the same factor for the inverted solution:

k = the weight of cupric oxid reduced by w:

kʹ = the same factor for wʹ:

p = the weight of cane sugar present in 100 cubic centimeters of the original solution: The formula to be employed then is

p = 0.4308 2	Gʹ kʹ	-	G k	.
	wʹ		w

This method has been applied to the estimation of cane sugar in molasses, apple juices and other substances. It is recommended by the authors as a simple and accurate means of estimating sucrose in all solutions containing it. The methods of making the copper reductions will be given hereafter.

99. Application of the Process.—In practice the process of inversion is used chiefly in the analysis of molasses and low grade massecuites. In approximately pure sugars the direct polarization is sufficiently accurate for all practical purposes. In molasses resulting from the manufacture of beet sugar are often found considerable quantities of raffinose, and the inversion process has been adapted to that character of samples. In molasses, in sugar cane factories, the disturbing factors are chiefly invert sugars and gums. The processes used for molasses will be given in another paragraph. In certain determinations of lactose the process of inversion is also practiced, but in this case the lactose is converted into dextrose and galactose, and the factors of calculation are altogether different. The process has also been adapted by McElroy and Bigelow to the determination of sucrose in presence of lactose, and this method will be described further on. In general the process of inversion is applicable to the determination of sucrose in all mixtures of other optically active bodies, which are not affected by the methods of inversion employed.

100. Determination of Sucrose and Raffinose.—Raffinose is a sugar which often occurs in beets, and is found chiefly in the molasses after the chief part of the sucrose has been removed by crystallization. It is also found in many seeds, notably in those of the cotton plant. In a pure solution of sucrose and raffinose, both sugars may be determined by the inversion method of Creydt.[63] The inversion is effected by means of hydrochloric acid in the manner described by Clerget. The following formulas are calculated for a temperature of observation of 20°, and the readings should be made as near that temperature as possible.

In these formulas S and R are the respective per cents of sucrose and raffinose desired, A the polarization in sugar degrees before inversion, B the polarization after inversion read at 20°, and C is the algebraic difference between A and B. It must be understood that these formulas are applicable only to a solution containing no other optically active substances, save sucrose and raffinose.

101. Specific Rotatory Power.—In order to compare among themselves the rotations produced on a plane of polarized light by different optically active bodies in solution, it is convenient to refer them all to an assumed standard. The degree of rotation which the body would show in this condition, is found by calculation, since, in reality, the conditions assumed are never found in practice. In the case of sugars and other optically active bodies, the standard of comparison is called the specific rotatory power. This factor in any given case, is the angular rotation which would be produced by any given substance in a pure anhydrous state if it were one decimeter in length and of a specific gravity equal to water. These are conditions which evidently do not exist in the case of sugars, since crystalline sugar particles have no polarizing power, and it would be impossible to pass a ray of light through an amorphous sugar column of the length specified. The specific rotatory power is therefore to be regarded as a purely theoretical factor, calculated from the actual data obtained by the examination of the solution of any given substance. If the length of the observation tube in decimeters be represented by l, the percentage of the polarizing body in 100 grams by p, and the specific gravity of the solution by d, and the observed angle of rotation by a, then the factor is calculated from the formula:

The symbols Dj refer to the character of light employed, D indicating the monochromatic sodium flame, and j the transition tint from white light.

If the weight of the polarizing body c be given or known for 100 cubic centimeters of the solution the formula becomes

The latter formula is the one easier of application since it is only necessary in applying it to dissolve a given weight of the active body in an appropriate solvent and to complete the volume of the solution exactly to 100 cubic centimeters. It is therefore unnecessary in this case to determine the specific gravity.

102. Formulas for Calculating Specific Rotatory Power.—In order to determine the specific rotatory power (gyrodynat[64]) of a given substance it is necessary to know the specific gravity and percentage composition or concentration of its solution, and to examine it with monochromatic polarized light in an instrument by which the angular rotation can be measured. The gyrodynat of any body changes with its degree of concentration, in some cases with the temperature, and always with the color of the light. With the red rays the gyrodynat is least and itprogressively increases as the violet end of the spectrum is approached. In practice the yellow ray of the spectrum has been found most convenient for use, and in the case of sugars the gyrodynat is always expressed either in terms of this ray or if made with color compensating instruments in terms of the sensitive or transition tint. In the one case the symbol used is (a)_D and in the other (a)_j. From this statement it follows that (a)_D is always numerically less than (a)_j. Unless otherwise specified the gyrodynat of a body is to be considered as determined by yellow monochromatic light, and therefore corresponds to a_D.[65]

103. Variations in Specific Rotatory Power.—The gyrodynat of any optically active body varies with the nature of the solvent, the strength of the solution, and the temperature.[66]

Since water is the only solvent of importance in determining the gyrodynat of sugars it will not be necessary here to discuss the influence of the nature of the solvent. In respect of the strength of the solution it has been established that in the case of cane sugar the gyrodynat decreases while with dextrose it increases with the degree of concentration. The influence of temperature on the gyrodynat of common sugars is not of great importance save in the case of levulose, where it is the most important factor, the gyrodynat rapidly increasing as the temperature falls. It is of course understood that the above remarks do not apply to the increase or decrease in the volume of a solution at changed temperatures. This influence of temperature is universally proportional to the change of volume in all cases, and this volumetric change is completely eliminated when the polarizations are made at the temperatures at which the solutions are completed to standard volumes.

104. Gyrodynatic Data for Common Sugars.—In the case of cane sugar the gyrodynat for twenty-five grams of sugar in 100 grams of solution at 20° is [a]_D = 66°.37. This is about the degree of concentration of the solutions employed in the shadow lamplight polariscopes. For seventeen grams of sugar in 100 grams of solution the number is [a]_D = 66°.49. This is approximately the degree of concentration for the laurent instrument.

For any degree of concentration according to Tollens the gyrodynat may be computed by the following formula: [a]_D = 66°.386 + 0.015035p - 0.0003986p², in which p is the number of grams of sugar in 100 grams of the solution.[67] In the table constructed by Schmitt the data obtained are as follows:

105. Bi-Rotation.—Some sugars in fresh solution show a gyrodynat much higher than the normal, sometimes lower. The former phenomenon is called bi- the latter semi-rotation. Dextrose shows birotation in a marked degree, also maltose and lactose. After standing for a few hours, or immediately on boiling, solutions of these sugars assume their normal state of rotation. The addition of a small quantity of ammonia also causes the birotation to disappear.[68] This phenomenon is doubtless due to a certain molecular taxis, which remains after solution is apparently complete. The groups of molecules thus held in place have a certain rotatory power of their own and this is superadded to that of the normal solution. After a time, under the stress of the action of the solvent, these groups are broken up and the solution then assumes its normal condition.

106. Gyrodynat of Dextrose.—The gyrodynat of dextrose, as has already been mentioned, increases with the degree of concentration, thus showing a property directly opposite that of sucrose.

The general formula for the anhydrous sugar is [a]_D = 52.°718 + 0.017087p + 0.0004271p². In this formula p represents the grams of dextrose in 100 grams of the solution. In a ten per cent solution the gyrodynat of dextrose is therefore nearly exactly [a]_D20° = 53°. As calculated by Tollens the gyrodynats corresponding to several degrees of concentration are shown in the following table:

107. Gyrodynats of Other Sugars.—Of the other sugars it will be sufficient to mention only levulose, maltose, lactose, and raffinose. For complete tables of gyrodynatic powers the standard books on carbohydrates may be consulted.[69]

The gyrodynat of levulose is not definitely established. At 14° the number is nearly expressed by [a]_D14° = -93°.7.

Invert sugar, which should consist of exactly equal molecules of dextrose and levulose, has a gyrodynat expressed by the formula [a]_D0° = -27°.9, with a concentration equivalent to 17.21 grams of sugar in 100 cubic centimeters. The gyrodynat decreases with increase of temperature, according to the formula [a]_Dt° = - (27°.9 - 0.32t°). According to this formula the solution is neutral to polarized light at 87°.2, and this corresponds closely to the data of experiment.

Maltose, in a ten per cent solution at 20°, shows a gyrodynat of [a]_D20° = 138°.3.

The general formula for other degrees of concentration is [a]_D = 140°.375 - 0.01837p - 0.095t, in which p represents the number of grams in 100 grams of the solution and t the temperature of observation.

In the case of lactose [a]_D = 52°.53, and this number does not appear to be greatly influenced by the degree of concentration; but is somewhat diminished by a rising temperature.

The gyrodynat of raffinose in a ten per cent solution is [a]_D = 104°.5.

108. General Principles.—The methods for the chemical estimation of sugars in common use depend on the reducing actions exerted on certain metallic salts, whereby the metal itself or some oxid thereof, is obtained. The reaction is either volumetric or the resulting oxid or metal may be weighed. The common method is, therefore, resolved into two distinct processes, and each of these is carried out in several ways. Not all sugars have the faculty of exerting a reducing action on highly oxidized metallic salts and the most common of them all, viz., sucrose is practically without action. This sugar, however, by simple hydrolysis, becomes reducing, but the two components into which it is resolved by hydrolytic action do not reduce metallic salts in the same proportion. Moreover, in all cases the reducing power of a sugar solution is largely dependent on its degree of concentration, and this factor must always be taken into consideration. Salts of copper and mercury are most usually selected to measure the reducing power of a sugar and in point of fact copper salts are almost universally used. Copper sulfate and carbonate are the salts usually employed, and of these the sulfate far more frequently, but after conversion into tartrate. Practically, therefore, the study of the reducing action of sugar as an analytical method will be confined almost exclusively to the determination of its action on copper tartrate.

Direct gravimetric methods are also practiced to a limited extent in the determination of sugars as in the use of the formation of sucrates of the alkaline earths and of the combinations which certain sugars form with phenylhydrazin. Within a few years this last named reaction has assumed a marked degree of importance as an analytical method. The most practical treatment of this section, therefore, for the limited space which can be given it, will be the study of the reducing action of sugars, both from a volumetric and gravimetric point of view, followed by a description of the best approved methods of the direct precipitation of sugars by such reagents as barium hydroxid and phenylhydrazin.

109. Classification.—Among the volumetric methods will be given those which are in common use or such as have been approved by the practice of analysts. Since the use of mercuric salts is now practiced to a limited extent, only a brief study of that process will be attempted. With the copper methods a somewhat extended description will be given of those depending on the use of copper sulfate, and a briefer account of the copper carbonate process.

In the copper sulfate method two distinct divisions must be noted, viz., first an indirect process depending first upon the reduction of the copper to a suboxid, the subsequent action of this body on iron salts, measured finally by titration with potassium permanganate; and second, a direct process determined either by the disappearance of the blue color from the copper solution, or by the absence of copper from a drop of the solution withdrawn and tested with potassium ferrocyanid. This last mentioned reaction is one which is found in common use. The volumetric methods are not, as a rule, as accurate as the gravimetric, depending on weighing the resultant metal, but they are far more rapid and well suited to technical control determinations.

110. Reduction of Mercuric Salts.—The method of determining sugar by its action on mercuric salts, is due to Knapp.[70] The method is based on the observation that dextrose and other allied sugars, will reduce an alkaline solution of mercuric cyanid, and that the mercury will appear in a metallic state.

The mercuric liquor is prepared by adding to a solution of ten grams of mercuric cyanid, 100 cubic centimeters of a solution of caustic soda of 1.145 specific gravity, and making the volume to one liter with water. The solution of sugar to be titrated, should be as nearly as possible of one per cent strength.

To 100 cubic centimeters of the boiling solution, the sugar solution is added in small portions from a burette and in such a way as to keep the whole mass in gentle ebullition.

To determine when all the mercuric salt has been decomposed, a drop of the clear boiling liquid is removed and brought into contact with a drop of stannous chlorid solution on a white surface. A brownish black coloration or precipitate will indicate that the mercury is not all precipitated. Fresh portions of the sugar must then be added, until no further indication of the presence of mercury is noted. The approximate quantity of sugar solution required to precipitate the mercury having thus been determined, the process is repeated by adding rapidly, nearly the quantity of sugar solution required, and then only a few drops at a time, until the reduction is complete.

One hundred cubic centimeters of the mercuric cyanid solution prepared as directed above, will be completely reduced by

By reason of the unpleasant odor of the boiling mercuric cyanid when in presence of a reducing agent, the process should be conducted in a well ventilated fume chamber. With a little practice the process is capable of rapid execution, and gives reasonably accurate results.

111. Sachsse’s Solution.—The solution of mercuric salts proposed by Sachsse, is made by dissolving eighteen grams of mercuric iodid in twenty-five cubic centimeters of an aqueous solution of potassium iodid. To this solution are added 200 cubic centimeters of potash lye, containing eighty grams of caustic potash. After mixing the solution, the volume is completed to one liter. The sugar solutions used to reduce this mixture, should be more dilute than those employed with the mercuric cyanid, and should not be over one-half per cent in strength. The end of the reduction is determined as already described. After a preliminary trial, nearly all the sugar necessary to complete reduction, should be added at once, and the end of the reduction then determined by the addition of successive small quantities. One hundred cubic centimeters of the mercuric iodid solution prepared as directed above, require the following quantities of sugar to effect a complete reduction:

By reason of the great difference between the reducing power of dextrose and levulose in this solution, it has been used in combination with the copper reduction method, to be described, to determine the relative proportion of dextrose and levulose in a mixture.[71]

It is now known that copper solutions require slightly different quantities of dextrose, levulose, or invert sugar to effect complete reduction, but the variations are not great and in the calculation above mentioned, it may be assumed that these differences do not exist.

Instead of using stannous chlorid as an indicator, the end of the reaction may be determined as follows: A disk of filtering paper is placed over a small beaker containing some ammonium sulfid. A drop of the clear hot solution is placed on this disk, and if salts of mercury be still present a dark stain will be produced; or a drop of the ammonium sulfid may be brought near the moist spot formed by the drop of mercury salt. An alkaline solution of zinc oxid may also be used.

The methods depending on the use of mercuric salts have, of late, been supplanted by better processes, and space will not be given here to their further discussion.

112. The Volumetric Copper Methods.—The general principle on which these methods depend, is found in the fact that certain sugars, notably, dextrose, (glucose), levulose, (fructose), maltose and lactose, have the property of reducing an alkaline solution of copper to a lower state of combination, in which the copper is separated as cuprous oxid. The end of the reaction is either determined by the disappearance of the blue color of the solution, or by the reaction produced by a drop of the hot filtered solution, when placed in contact with a drop of potassium ferrocyanid acidified with acetic.

The copper salt which is found to give the most delicate and reliable reaction, is the tartrate. The number of volumetric processes proposed and which are in use, is very great, and an attempt even to enumerate all of these can not be made in this volume. A few of the most reliable and best attested methods will be given, representing if possible, the best practice in this and other countries. The rate of reduction of the copper salt to suboxid, is influenced by the rate of mixing with the sugar solutions, the temperature, the composition of the copper solution and the strength of the sugar solution.

The degree of reduction is also modified by the rate at which the sugar solution is added, and by the degree and duration of heating, and all these variables together, make the volumetric methods somewhat difficult and their data, to a certain extent, discordant. By reason, however, of the ease with which they are applied and the speed of their execution, they are invaluable for approximately correct work and for use in technical control.

113. Historical.—It is not the purpose in this paragraph to trace the development of the copper reduction method for the determination of reducing sugars, but only to refer to the beginning of the exact analytical application of it.

Peligot, as early as 1844, made a report to the Society for the Encouragement of National Industry on methods proposed by Barreswil and Fromherz for the quantitive estimation of sugar by means of copper solution.[72] These methods were based on the property of certain sugars to reduce alkaline copper solution to a state of cuprous oxid first announced by Trommer.[73] This was followed by a paper by Falck on the quantitive determination of sugar in urine.[74]

In 1848 the methods, which have been proposed, were critically examined by Fehling, and from the date of his paper the determination of sugar by the copper method may be regarded as resting on a scientific basis.[75]

Since the date mentioned the principal improvements in the process have been in changing the composition of the copper solution in order to render it more stable, which has been accomplished by varying the proportions of copper sulfate, alkali and tartaric acid. For the better keeping of the solution the method of preserving the copper sulfate and the alkaline tartrates in separate flasks and only mixing them at the time of use has been found very efficacious.[76] For testing for the end of the reaction by means of an acetic acid solution of potassium ferrocyanid the filtering tube suggested by the author, the use of which will be described further on, has proved quite useful. Pavy has suggested that by the addition of ammonia to the copper solution the precipitated suboxid may be kept in solution and the end of the reaction thus easily distinguished by the disappearance of the blue color.[77] Allen has improved on this method by covering the hot mixture with a layer of paraffin oil whereby any oxidation of the suboxid is prevented.[78]

The introduction and development of the gravimetric process depending on securing the reduced copper oxid in a metallic state as developed by Allihn, Soxhlet, and others, completes the resumé of this brief sketch of the rise and development of the process.

114. Action of Alkaline Copper Solution on Dextrose.—The action to which dextrose and other reducing sugars are subjected in the presence of a hot alkaline copper solution is two-fold in its nature. In the first place there is an oxidation of the sugar which is transformed into tartronic, formic and oxalic acids, the two latter in very small quantities. At the same time another part of the sugar is attacked directly by the alkali and changed to complex products among which have been detected lactic, oxyphenic and oxalic acids, also two bodies isomeric with dioxyphenolpropionic acid. When the sugar is in large excess melassic and glucic acids have also been detected. The glucic acid may be regarded as being formed by simple dehydration but becomes at once resolved into pyrocatechin and gluconic acid according to the reaction C₁₂H₁₈O₉ = C₆H₆O₂ + C₅H₁₂O₇. The gluconic acid also is decomposed and gives birth to lactic and glyceric acids according to the formula C₆H₁₂O₇ = C₃H₆O₃ + C₃H₆O₄. The glyceric acid also in the presence of a base is changed into lactic and oxalic acids. Between lactic acid and pyrocatechin, existing in a free state, there is produced a double reciprocal etherification in virtue of which there arise two ethers isomeric with hydrocaffeic acid, C₉H₁₀O₄. One of these bodies is an acid and corresponds to the constitution

• CH₃
•   /
•   O ── CH
• /  \
• C₆H₄ CO₂H (2)
• \
• OH (2)

and the other is of an alcoholic nature corresponding to the formula

•  CO₂ ── CHOH ── CH₃ (1)
• /
• C₆H₄
• \
• OH₂ (2)

Of all these products only oxyphenic and lactic acids and their ethers and oxalic acid remain unchanged and they can be isolated. All the others are transformed in an acid state and they can only be detected by operating in the presence of metallic oxids capable of precipitating them at the time of their formation.[79]

115. Fehling’s Solution.—The copper solution which has been most used in the determination of reducing sugars is the one proposed by Fehling as a working modification of the original reagent used by Trommer.[80]

Following is the formula for the preparation of the fehling solution:

The copper sulfate is dissolved in water and the potassium tartrate in the aqueous solution of the sodium hydroxid which should have a volume of about 700 cubic centimeters. The two solutions are mixed and the volume completed to a liter. Each cubic centimeter of this solution will be reduced by five milligrams of dextrose, equivalent to four and a half milligrams of sucrose.

The reaction which takes place is represented by the following molecular proportions:

Fehling’s solution is delicate in its reactions but does not keep well, depositing cuprous oxid on standing especially in a warm place exposed to light. The fehling liquor was soon modified in its constitution by substituting 173 grams of the double sodium and potassium tartrate for the neutral potassium tartrate first used, and, in fact, the original fehling reagent contained forty grams of copper sulfate instead of the quantity mentioned above. Other proportions of the ingredients are also given by many authors as fehling solution.

116. Comparison of Copper Solutions for Oxidizing Sugars.—For the convenience of analysts there is given below a tabular comparison of the different forms of fehling liquor which have been proposed for oxidizing sugars. The table is based on a similar one prepared by Tollens and Rodewald, amended and completed by Horton.[81] The solutions are arranged alphabetically according to authors’ names:

117. Volumetric Method used in this Laboratory.—The alkaline copper solution preferred in this laboratory has the composition proposed by Violette. The copper sulfate and alkaline tartrate solutions are kept in separate vessels and mixed in proper proportions immediately before use, and diluted with about three volumes of water. The reduction is accomplished in a long test tube at least twenty-five centimeters in length, and from thirty-five to forty millimeters in diameter.

The sugar solution employed should contain approximately one per cent of reducing sugar. If it should have a greater content it should be reduced with water to approximately the one named. If it have a less content, it should be evaporated in a vacuum at a low temperature until it reaches the strength mentioned above. A preliminary test will indicate almost the exact quantity of the sugar solution to be added to secure a complete reduction of the copper. This having been determined the whole quantity should be added at once to the boiling copper solution, the test tube held in the open flame of a lamp giving a large circular flame and the contents of the tube kept in brisk ebullition for just two minutes. The lamp is withdrawn and the precipitated suboxid allowed to settle. If a distinct blue color remain an additional quantity of the sugar solution is added and again boiled for two minutes. When the blue coloration is no longer distinct, the presence or absence of copper is determined by aspirating a drop or two of the hot solution with the apparatus described below. This clear filtered liquor is then brought into contact with a drop of potassium ferrocyanid solution acidulated with acetic. The production of a brown precipitate or color indicates that some copper is still present, in which case an additional quantity of the sugar solution is added and the operation continued as described above until after the last addition of sugar solution no coloration is produced.

118. The Filtering Tube.—The filtering tube used in the above operation is made of a long piece of narrow glass tubing with thick walls. The length of the tube should be from forty to forty-five centimeters. One end of the tube being softened in the flame is pressed against a block of wood so as to form a flange. Over this flange is tied a piece of fine linen.[82]

Instead of using a linen diaphragm the tube is greatly improved, as suggested by Knorr, by sealing into the end of the tube while hot a perforated platinum disk. Before using, the tube is dipped into a vessel containing some suspended asbestos felt and by aspiration a thin felt of asbestos is formed over the outer surface of the platinum disk. By inverting the tube the water which has entered during aspiration is removed. The tube thus prepared is dipped into the boiling solution in the test tube above described and aspiration continued until a drop of the liquor has entered the tube. It is then removed from the boiling solution, the asbestos felt wiped off with a clean towel, and the drop of liquor in the tube blown through the openings in the platinum disk and brought into contact with a drop of potassium ferrocyanid in the usual way. In this way a drop of the liquor is secured without any danger of a reoxidation of the copper which may sometimes take place on cooling.

Figure 42. Apparatus for the Volumetric Estimation
of Reducing Sugars.

The careful analyst by working in this way with the volumetric method is able to secure highly accurate results. The apparatus used is shown in the accompanying illustration.

119. Suppression of the Error Caused by the Action of the Alkali on Reducing Sugars.—Three methods are proposed by Gaud for correcting or suppressing the error due to the action of the alkali upon reducing sugars. In the first place, the common method followed may be employed, depending upon the use of an alkaline copper solution of known composition and the employment of a reducing sugar solution of a strength varying between one-half and one per cent. The error which is introduced into such a reaction is a constant one and the solution having been tested once for all against pure sugar is capable of giving fairly accurate results.

In the second place, a table may be constructed in which the error is determined for sugar solutions for varying strengths, viz., from one-tenth of one per cent to ten per cent. If y represent the error and x the exact percentage of reducing sugar present then the correction may be made by the following formula;

y = - 0.00004801x + 0.02876359x².

In order to use this formula in practice the percentage of reducing sugar obtained by the actual analysis must be introduced and may be represented by θ. The formula for correction then becomes 0.02876x² - 1.000048x + θ = 0; whence the value of x is easily computed.

In the third place, the error may be eliminated by substituting for an alkali which acts upon the glucose one which does not, viz., ammonia. At the temperature of boiling water ammonia does not have any decomposing effect upon reducing sugars. It is important, however, that the reduction take place in an inert atmosphere in order to avoid the oxidation of the dissolved cuprous oxid and the temperature need not be carried beyond 80°. The end of the reaction can be easily distinguished in this case by the disappearance of the blue color. When one reaction is finished the copper may be completely reoxidized by conducting through it a current of air or oxygen for half an hour, when an additional quantity of ammonia may be added to supply any that may have evaporated, and a new reduction accomplished with exactly the same quantity of copper as was used in the first. The solution used by Gaud contains 36.65 grams of crystallized copper sulfate dissolved in water and the volume completed to one liter with ordinary aqueous ammonia.[83]

120. Permanganate Process for the Estimation of Reducing Sugars.—Dextrose, invert sugar, and other reducing sugars can also be determined with a fair degree of accuracy by an indirect volumetric process, in which a standard solution of potassium permanganate is used as the final reagent.[84] The principle of the process is based upon the observation that two molecules of dextrose reduce from an alkaline cupric tartrate solution five molecules of cuprous oxid. The five molecules of cuprous oxid thus precipitated when added to an acid solution of ferric sulfate, will change five molecules of the ferric sulfate to ten molecules of ferrous sulfate. The reaction is illustrated by the following equation:

5Cu₂O 715 parts

+

5Fe₂(SO₄)₃ 2000 parts

+

5H₂SO₄ 490 parts

=

10CuSO₄ 1595 parts

+

10FeSO₄ 1520 parts

+

5H₂O 90 parts

The ten molecules of ferrous sulfate formed as indicated in the above reaction, are reoxidized to ferric sulfate by a set solution of potassium permanganate. This reaction is illustrated by the equation given below:

10FeSO₄ 1520 parts

+

K₂Mn₂O₈ 316.2 parts

+

5H₂SO₄ 784 parts

=

5Fe₂(SO₄)₂ 2000 parts

+

2MnSO₄ 302 parts

+

K₂SO₄ 174.2 parts

+

8H₂O 144 parts

By the study of the above equations it is seen that two molecules of dextrose or other similar reducing sugar, are equivalent to one molecule of potassium permanganate, as is shown by the following equations:

2C₆H₁₂O₆ 360 parts

=

5Cu₂O 715 parts

=

10FeSO₄ 1520 parts

=

K₂Mn₂O₈ 316.2 parts

It is thus seen that 316.2 parts by weight of potassium permanganate are equivalent to 360 parts by weight of dextrose; or one part of permanganate corresponds to 1.1385 parts by weight of dextrose. If, therefore, the amount of permanganate required in the above reaction to restore the iron to the ferric condition, be multiplied by the factor mentioned above, the quotient will represent in weight the amount of dextrose which enters into the reaction. The standard solution of potassium permanganate should contain 4.392 grams of the salt in a liter. One cubic centimeter of this solution is equivalent to five milligrams of dextrose.

121. Manipulation.—The saccharine solution whose strength is to be determined should contain approximately about one per cent of sugar. Of this solution ten cubic centimeters are placed in a porcelain dish together with a considerable excess of fehling solution. When no sucrose is present, the mixture may be heated to the temperature of boiling water and kept at that temperature for a few minutes until all the reducing sugar is oxidized. There should be enough of the copper solution used to maintain a strong blue coloration at the end of the reaction. A greater uniformity of results will be secured by using in all cases a considerable excess of the copper solution. When sucrose or other non-reducing sugars are present, the temperature of the reaction should not be allowed to exceed 80° and the heating may be continued somewhat longer. At this temperature the copper solution is absolutely without action on sucrose. The precipitated suboxid is allowed to settle, the supernatant liquid poured off through a filter and the suboxid washed thoroughly a number of times by decantation with hot water, the washings being poured through the filter. This process of washing is greatly facilitated by decanting the supernatant liquid from the porcelain dish first into a beaker and from this into a third beaker and so on until no suboxid is carried off. Finally the wash water is poured through a filter-paper bringing as little as possible of the suboxid onto the paper. The suboxid on the filter-paper and in the beakers is next dissolved in a solution of ferric sulfate made strongly acid with sulfuric; or in a sulfuric acid solution of ammonia ferric sulfate which is more easily obtained free from impurities than the ferric sulfate. When all is dissolved from the beakers the solution is poured upon the suboxid which still remains in the porcelain dish. When the solution is complete it is washed into a half liter flask and all the vessels which contain the suboxid are also thoroughly washed and the wash waters added to the same flask. The whole is rendered strongly acid with sulfuric and made up to a volume of half a liter.

The process carried out as directed, when tested against pure sugar, gives good results, not varying from the actual content of the sugar by more than one-tenth per cent below or three-tenths above the true content. The distinct pink coloration imparted to the solution by the permanganate solution as soon as the iron is all oxidized to the ferric state marks sharply the end of the reaction. In this respect this process is very much to be preferred to the usual volumetric processes depending upon the coloration produced with potassium ferrocyanid by a copper salt for distinguishing the end of the reaction. It is less convenient than the ordinary volumetric process by reason of the somewhat tedious method of washing the precipitated cuprous oxid. When a large number of analyses is to be made, however, the whole can be washed with no more expenditure of time than is required for a single sample. One analyst can, in this way, easily attend to fifty or a hundred determinations at a time.

In the application of the permanganate method to the analysis of the juices of sugar cane and sorghum it is directed to take 100 cubic centimeters of the expressed juice and clarify by the addition of twenty-five cubic centimeters of basic lead acetate, diluted with water, containing enough of the lead acetate, however, to produce a complete clarification. It is not necessary to remove the excess of lead from the filtrate before the determination. Ten cubic centimeters of the filtrate correspond to eight cubic centimeters of the original juice. For percentage calculation the specific gravity of the original juice must be known. Before the addition of the alkaline copper solution, from fifty to seventy-five cubic centimeters of water should be added to the clarified sugar juice and the amount of fehling solution used in each case should be from fifty to seventy-five cubic centimeters. The heating at 75° should be continued for half an hour in order to insure complete reduction and oxidation of the sugar. The sucrose can also be estimated in the same juices by inverting five cubic centimeters of the clarified juice with five cubic centimeters of dilute hydrochloric acid, by heating for an hour at a temperature not above 90°. Before adding the acid for inversion, about 100 cubic centimeters of water should be poured over the five cubic centimeters of sugar solution. The washing of the suboxid and the estimation of the amount reduced are accomplished in the manner above described.

This method has been extensively used in this laboratory and with very satisfactory results. The only practical objection which can be urged to it is in the time required for filtering. This fault is easily remedied by adopting the method of filtering through asbestos felt described in the next paragraph.

For the sake of uniformity, however, the copper solution should be boiled for a few minutes before the addition of the sugar in order to expel all oxygen, the sugar solutions should be made with recently boiled water and the precipitation of the suboxid should be accomplished by heating for just thirty minutes at 75°. At the end of this time an equal volume of cold, recently boiled, water should be added and the filtration at once accomplished.

122. Modified Permanganate Method.—The permanganate method as used by Ewell, in this laboratory, is conducted as follows: After the precipitate is obtained, according to the directions given in the methods described, it is thoroughly washed with hot, recently boiled water, on a gooch. The asbestos, with as much of the precipitate as possible, is transferred to the beaker in which the precipitation was made, beaten up with from twenty-five to thirty cubic centimeters of hot, recently boiled water, and from fifty to seventy-five cubic centimeters of a saturated solution of ferric sulfate in twenty-five per cent sulfuric acid are added to the beaker and then poured through the crucible to dissolve the cuprous oxid remaining therein. If the precipitate be first beaten up with water as directed, so that no large lumps of it remain, there is no difficulty in dissolving the oxid in the ferric salt; while if any lumps of the oxid be allowed to remain there is great difficulty. After the solution is obtained, it is titrated with a solution of potassium permanganate of such a strength that each cubic centimeter is equivalent to 0.01 gram of copper.

In triplicate determinations made by this method the precipitates obtained required after solution in the ferric salt, 28.7, 28.9, 28.6 cubic centimeters of potassium permanganate solution, respectively. For the quantities taken this was equivalent to an average percentage of reducing sugars of 4.19. The percentage obtained by the gravimetric method was 4.26.

The method seems to be sufficiently accurate for all ordinary purposes and is extremely rapid.

The permanganate solution used should be standardized by means of metallic iron, but in ordinary work it is also recommended to standardize by check determinations of reducing sugars in the same sample by the gravimetric method.

123. Determination of Reducing Sugar by the Specific Gravity of the Cuprous Oxid.—Gaud proposes to determine the percentage of reducing sugar from the specific gravity of the cuprous oxid. The manipulation is carried out as follows:

In a porcelain dish are placed fifty cubic centimeters of the alkaline copper solution and an equal quantity of water and the mixture maintained in ebullition for two or three minutes. The dish is then placed on a boiling water-bath and twenty-five cubic centimeters of a reducing sugar of approximately one per cent strength added at once. The reduction is thus secured at a temperature below 100°, which is an important consideration in securing the minimum decomposing effect of the alkali upon the sugar. The dish is kept upon the water-bath for about ten minutes when the reduction is complete and the supernatant liquor should still be intensely blue. The precipitate is washed by decantation with boiling water, taking care to avoid the loss of any of the cuprous oxid. The washing is continued until the wash waters are neutral to phenolphthalein. The cuprous oxid is then washed into a pyknometer of from twenty to twenty-five cubic centimeters capacity, the exact content of which has been previously determined at zero. It is filled with boiling water, the stopper inserted, and after cooling the flask is weighed. Let P be the weight of the pyknometer plus the liquid and the precipitate, the total volume of which is equal to the capacity of the flask at the temperature at which it was filled, that is Vₜ = V₀ [1 + 3β(t-t₀)].

This formula is essentially that given in paragraph 51, for calculating the volume of a pyknometer at any temperature, substituting for 3β, γ the cubical expansion of glass, viz., 0.000025.

The specific gravity of the dry cuprous oxid is Δ = 5.881 and let the specific gravity of water at the temperature of filling, which can be taken from any of the tables of the density of water, be d. The total weight p of the precipitated suboxid may then be calculated by the following formula:

The density of water at 99°, which is about the mean temperature of boiling water for laboratories in general, is 0.95934, and this may be taken as the weight of one cubic centimeter for purposes of calculation in the formula above.

In order to obtain exact results, it is important that the weight P be reduced to a vacuum. The weight of cuprous oxid not varying proportionally to the weight of reducing sugar, it is necessary to prepare a table showing the principal numerical values of the two, in order to be able to calculate easily all the possible values, either directly from the table or by appropriate interpolations. Following are the chief values which are necessary for the calculation:

It is claimed by the author that the above method is both simple and rapid and can be applied with an error of not more than one-thousandth if the corrections for temperature and pressure be rigorously applied.[85]

124. The Copper Carbonate Process.—While the copper solutions which have been mentioned in previous paragraphs have only a slight action on sucrose and dextrin yet on prolonged boiling even these bodies show a reducing effect due probably to a preliminary change in the sugar molecules whereby products analogous to dextrose or invert sugar are formed. In order to secure a reagent, to which the sugar not reducing alkaline copper solutions might be more resistant Soldaini has proposed to employ a liquor containing the copper as carbonate instead of as tartrate.[86] This solution is prepared by adding to a solution of forty grams of copper sulfate one of equal strength of sodium carbonate. The resulting copper carbonate and hydroxid are collected on a filter, washed with cold water, and dried. The reaction which takes place is represented by the following formula:

2CuSO₄ + 2Na₂CO₃ + H₂O =
CuCO₃ + CuO₂H₂ + 2Na₂SO₄ + CO₂.

The dry precipitate obtained, which will weigh about fifteen grams, is placed in a large flask with about 420 grams of potassium bicarbonate and 1400 cubic centimeters of water. The contents of the flask are heated on a steam-bath for several hours with occasional stirring until the evolution of carbon dioxid has ceased. During this time the liquid is kept at the same volume by the addition of water, or by attaching a reflux condenser to the flask. The potassium and copper compounds at the end of this time will be found dissolved and the resulting liquor will have a deep blue color. After filtration the solution is boiled for a few minutes and cooled to room temperature. The volume is then completed to two liters. A more direct method of preparing the solution, and one quite as effective, consists in adding the solution of the copper sulfate directly to the hot solution of potassium bicarbonate and heating and shaking the mixture until the copper carbonate formed is dissolved. After filtering the volume is made as above. The proportions of reagents employed are placed by Preuss at 15.8 grams of crystallized copper sulfate and 594 grams of potassium bicarbonate.[87] The soldaini reagent is extremely sensitive and is capable of detecting as little as half a milligram of invert sugar. The presence of sucrose makes the reagent more delicate, and it is especially useful in determining the invert sugar arising during the progress of manufacture by the action of heat and melassigenic bodies on sucrose.

125. The Analytical Process.—As in the case of fehling solution a great many methods of conducting the analysis with the soldaini reagent have been proposed. The general principle of all these processes is the one already described for the alkaline copper tartrate solution, viz., the addition of the reducing sugar solution to the boiling reagent, and the determination of the end of the reaction by the disappearance of the copper.[88]

Practically, however, these methods have had no general application, and the use of the soldaini reagent has been confined chiefly to the determination of invert sugar in presence of a large excess of sucrose. For this purpose the sugar solution is not added until the blue color of the reagent has been destroyed, but on the other hand, the reagent has been used in excess, and the cuprous oxid formed collected and weighed as metallic copper. The weight of the metallic copper found, multiplied by the factor 0.3546, gives the weight of invert sugar in the volume of the sugar solution used. According to Preuss, the factor is not a constant one, but varies with the quantity of invert sugar present, as is seen in the formula y = 2.2868 + 3.3x + 0.0041x², in which x = the invert sugar, and y the metallic copper.[89]

126. Tenth Normal Copper Carbonate Solution.—In the study of some of the solutions of copper carbonate, proposed for practical work, Ettore Soldaini was impressed with the difficulty of dissolving so large a quantity of carbonate in the solvent employed.[90] The solution recommended by Bodenbender and Scheller,[91] in which forty grams of the crystallized copper sulfate were used, failed to disclose an equivalent amount of copper in the reagent ready for use. For this reason a tenth-normal copper solution is prepared by Soldaini containing the equivalent of 3.464 grams of copper sulfate in one liter. The reagent is easily prepared by adding slowly the dissolved or finely powdered copper salt to a solution of 297 grams of potassium bicarbonate, and after complete solution of the copper carbonate formed, completing the volume to one liter. With this reagent as little as one-quarter of a milligram of reducing sugar can be easily detected. For the quantitive estimation of sugar a solution of the above strength is to be preferred to the other forms of the soldaini reagent by reason of the ease of direct comparison with standard fehling solutions.

The analytical process is conducted with the tenth-normal solution, prepared by Soldaini and described above, as follows: Place 100 cubic centimeters of the reagent in each of several porcelain dishes heat to boiling, and add little by little the sugar solution to one dish until the blue color has disappeared. Having thus determined nearly the exact quantity of sugar solution required for the copper in 100 cubic centimeters of the reagent the whole of the sugar solution is added at once, varying slightly the amounts added to each dish. The boiling is continued for fifteen minutes, and the contents of the dishes poured on filters. That filtrate which contains neither copper nor sugar represents the exact quantity of sugar solution which contained fifty milligrams of dextrose.

127. Relation of Reducing Sugar to Quantity of Copper Suboxid Obtained.—The relation of the quantity of copper reduced to the amount of sugar oxidized by the copper carbonate solution has been determined by Ost, and the utility of the process thereby increased.[92] The solution used should have the following composition: 23.5 grams of crystallized copper sulfate, 250 grams of potassium carbonate, and 100 grams of potassium bicarbonate in one liter. Without an indicator the end reaction is distinctly marked by the passage of the blue color into a colorless solution. Ost affirms that this solution is preferable to any form of fehling liquor because it can be kept indefinitely unchanged; it attacks sucrose far less strongly, and an equal quantity of sugar precipitates nearly double the quantity of copper. The boiling requires a longer time, as a rule ten minutes, but this is a matter of no importance, when the other advantages are taken into consideration. The relations of the different sugars to the quantity of copper precipitated are given in the table in the next paragraph.

128. Factor for Different Sugars.—For pure dextrose the relation between sugar and copper reduced has been determined by Ost, and the data are given in the table below. The data were obtained by adding to fifty cubic centimeters of the copper solution twenty-five cubic centimeters of sugar solutions of varying strength and collecting, washing, and reducing the cuprous oxid obtained in a current of hydrogen in a glass tube by the method described further on.

The boiling in all cases was continued, just ten minutes, although a slight variation from the standard time did not produce so great a difference as with fehling reagent. In the case of dextrose, when fifty milligrams were used with fifty cubic centimeters of the solution, the milligrams of copper obtained after six, ten and twenty minutes’ boiling were 164.6, 165.5, and 166.9 respectively.[93]

The data differ considerably from those obtained by Herzfeld, but in his experiments the boiling was continued only for five minutes, and this is not long enough to secure the proper reduction of the copper.[94]

Table Showing the Quantity of Copper
Reduced by Different Sugars.

129. Pavy’s Process.—The well-known solubility of cuprous oxid in ammonia led Pavy to adopt a copper reagent containing ammonia in the volumetric determination of reducing sugars.[95] In Pavy’s process an alkaline copper solution is employed made up in the usual way, to which a sufficient quantity of ammonia is added to hold in solution all the copper when precipitated as cuprous oxid. The solution used by Pavy has the following composition: One liter contains

For use 120 cubic centimeters of the above reagent are mixed with 300 of ammonia of specific gravity 0.88, and the volume completed to one liter with distilled water. Twenty cubic centimeters of this reagent are equivalent to ten milligrams of dextrose or invert sugar when added in a one per cent solution.

In the use of ammoniacal copper solution, care must be taken that all the liquids employed be entirely free of oxygen and that the contents of the flask in which the reduction takes place be in some way excluded from contact with the air. Pavy secured this by conducting the reduction in a flask closed with a stopper carrying two holes; one of these served for the introduction of the burette carrying the sugar solution and the other carried a tube dipping into a water seal by means of a slit rubber tube, which would permit of the exit of the vapors of steam and ammonia, but prevent the regurgitation of the water into the flask.

The complete decoloration of the copper solution marks the end of the reaction. The usual precautions in regard to the length of the time of boiling must be observed.

It is easy to see that in the Pavy process the quantity of ammonia in the solution is rapidly diminished during the boiling and this has led to the suggestion of other methods to exclude the air. Among these have been recommended the introduction of a current of hydrogen or carbon dioxid. One of the best methods of procedure is that proposed by Allen, who recommends covering the copper solution by a layer of paraffin oil (kerosene).[96]

130. Process Of Peska.—Peska has also independently made use of Allen’s method of covering the solutions with a layer of paraffin oil and finds it reliable.[97] The copper reagent employed by him has the following composition:

The copper sulfate is dissolved in water, the ammonia added, and the volume completed to half a liter with distilled water. A second solution containing half a liter is made by dissolving 34.5 grams of potassium-sodium tartrate and ten grams of sodium hydroxid and completing the cool solution to half a liter with distilled water. In all cases the water used in making up the above solutions must be freshly boiled to exclude the air.

For the titration, fifty cubic centimeters of each of the above solutions are taken, mixed and covered with a layer of paraffin oil half a centimeter in depth. The reduction is not accomplished at a boiling temperature, but at from 80° to 85°. The manipulation is conducted as follows:

The mixed solutions are placed in a beaker, covered with oil, and heated to 80°. The temperature is measured by a thermometer which also serves as a stirring rod. The sugar solution is run down the sides of the beaker from a burette of such a shape as to be protected from the heat. After each addition of the sugar solution the mixture is carefully stirred, keeping the temperature at from 80° to 85°. The first titration is made to determine approximately the quantity of sugar solution necessary to decolorize the copper. This done, the actual titration is accomplished by adding at once the total amount of sugar solution necessary to decolorize, less about one cubic centimeter. Any sugar solution adhering to the side of the beaker is washed down by distilled water, the contents of the beaker well stirred, and the temperature kept at 85° for two minutes. The rest of the sugar solution is then added in quantities of one-tenth of a cubic centimeter until the decoloration is completed. The total time of the final titration should not exceed five minutes. The sugar solution should be as nearly as possible of one per cent strength. If a lower degree of strength be employed a larger quantity of the sugar is necessary to reduce a given quantity of copper.

In the case of dextrose, when a one per cent solution is used, eight and two-tenths cubic centimeters, corresponding to 80.2 milligrams of dextrose, are required to reduce 100 cubic centimeters of the mixed reagent. On the other hand, when the sugar solution is diluted to one-tenth of a per cent strength 82.1 milligrams are required.

With invert sugar slightly larger quantities are necessary, the reducing power being as 94.9 to 100 as compared with dextrose. With a one per cent strength of invert sugar it is found that eighty-four milligrams are required to reduce 100 cubic centimeters of the mixed reagent and when the strength of the invert sugar is reduced to one-tenth per cent 87.03 milligrams are required.

131. Method Of Allein and Gaud.—Allein and Gaud have proposed a further modification of the ammonia process which consists essentially in the suppression of rochelle salt and fixed caustic alkali and the entire substitution therefor of ammonia. Ammonia acts with much less vigor upon sugars than the caustic alkalies, and it is therefore claimed that the decomposition of the sugar due to the alkali is reduced to a minimum when ammonia is employed.[98] The copper solution is made as follows:

Dissolve 8.7916 grams of electrolytic copper in ninety-three grams of concentrated sulfuric acid diluted with an equal volume of water. Complete the resulting solution to one liter with concentrated ammonia. Ten cubic centimeters of this solution are equal to fifty grams of dextrose.

It is recommended that the reduction be accomplished in an atmosphere of hydrogen, but it is apparent that the use of kerosene is permissible in this case, and on account of its greater simplicity it is to be recommended as the best means of excluding the oxygen. The reduction is accomplished at a temperature of about 80°.

It is also proposed to reoxidize the copper by substituting a current of air for the hydrogen at the end of the reaction, and thus use the same copper a number of times. The danger of loss of ammonia, and the difficulty of determining when the oxidation is complete, render this regeneration of the reagent undesirable.

132. Method of Gerrard.—The method of Gerrard does not depend upon the use of ammonia, but the principle involved is the same, viz., the holding of the separated cuprous oxid in solution and the determination of the end of the reaction by the disappearance of the blue color. As first proposed by Gerrard, the copper sulfate solution is made of double the strength usually employed and to each 100 cubic centimeters thereof, before use, three and three-tenths grams of potassium cyanid are added. This is sufficient to hold the precipitated cuprous oxid in solution.[99]

The original method of Gerrard is found difficult of execution and the author, in conjunction with Allen, has lately modified it and reduced it to a practical working basis.[100]

In the new method the ordinary fehling solution is employed and it is prepared for use in the following way: Ten cubic centimeters of the fehling solution, or half that quantity of each of the component parts kept in separate bottles, are placed in a porcelain dish with forty cubic centimeters of water and brought to the boiling-point. To the boiling liquid is added, from a pipette, a five per cent solution of potassium cyanid until the blue color just disappears, or only a very faint tint of blue remains, avoiding any excess of the cyanid. A second portion of the fehling solution equal to that first employed is added, and to the boiling mixture the solution of sugar is added, from a burette, until the blue color disappears. The contents of the dish should be kept boiling during the addition of the sugar solution. The volume used will contain fifty milligrams of dextrose. The sugar solution should be of such a strength as to contain no more than half a per cent of reducing sugar.

The principle of the preparation of the solution may be stated as follows: If to a solution of copper sulfate, potassium be added until the blue color disappears, a double cyanid of copper and potassium cyanid is formed according to the following reaction:

CuSO₄ + 4KCN = Cu(CN)₂.2KCN + K₂SO₄.

This double cyanid is a salt of considerable stability. It is not decomposed by alkalies, hydrogen or ammonium, sulfid. With mineral acids it gives a whitish, curdy precipitate. With fehling solution the same double cyanid is formed as that described above. If, however, fehling solution be present in excess of the amount necessary to form the double cyanid of copper, this excess can be used in the oxidation of reducing sugar and the colorless condition of the solution will be restored as soon as the excess of the fehling is destroyed. The double cyanid holds in solution the cuprous oxid formed and thus complete decoloration is secured.

133. Sidersky’s Modification of Soldaini’s Process.—In all cases where the sugar solutions are not too highly colored, Sidersky finds that the method of reduction in a large test tube, as practiced by Violette, is applicable with the copper carbonate solution.[101] For more exact work it is preferred to determine the quantity of copper reduced by an indirect volumetric method. The sugar solution, properly clarified and the lead removed if subacetates have been used, is made of such a volume as to contain less than one per cent of reducing sugars. In a flask or large test tube are placed 100 cubic centimeters of the copper solution, which is boiled for a short time and the sugar solution added, little by little, from a pipette, at such a rate as not to stop the ebullition. The boiling is continued for five minutes after the last addition of the sugar. The vessel is taken from the flame and 100 cubic centimeters of cold water added, the whole brought on an asbestos felt and the cuprous oxid washed with hot water until the alkaline reaction has disappeared. The residual cuprous oxid is dissolved in a measured quantity of set sulfuric acid, semi- or fifth-normal, a few particles of potassium chlorate added, and the mixture boiled to convert any cuprous into cupric sulfate. The reaction is represented by the following formula:

3Cu₂O + 6H₂SO₄ + KClO₃ = 6CuSO₄ + KCl + 6H₂O.

The residual sulfuric acid is titrated with a set alkali in excess, ammonia being preferred.

The solution of ammonia is made by diluting 200 cubic centimeters of commercial aqua ammonia with 800 of water. Its strength is determined by adding a little copper sulfate solution as indicator and then the set solution of sulfuric acid until the blue color disappears. The copper sulfate secured from the cuprous sulfate as described above is cooled, and a quantity of the ammonia, equal to twenty-five cubic centimeters of the set sulfuric acid, added. The excess of the ammonia is then determined by titration with the sulfuric acid, the disappearance of the blue color being the indication of the end of the reaction. The number of cubic centimeters of the set sulfuric acid required to saturate the ammonia represents the equivalent of cuprous oxid originally present. One cubic centimeter of normal sulfuric acid is equivalent to 0.0317 gram of metallic copper.

To determine the weight of invert sugar oxidized, multiply the weight of copper, calculated as above described, by the factor 0.3546.[102] For a general application of this method of analysis the relative quantities of copper reduced by different quantities of sugar must be taken into consideration.

While, as has already been stated, the copper carbonate process has heretofore been applied chiefly to the detection of invert sugar, it has merits which justify the expectation that it may some time supplant the fehling liquor both for volumetric and gravimetric work. Large volumes of the reagent can be prepared at once and without danger of subsequent change. The action of the reagent on the hexobioses and trioses is far less vigorous than that of the alkaline copper tartrate, and the end reactions for volumetric work are, at least, as easily determined in the one case as the other.

134. Method Depending on Titration of Excess of Copper.—Instead of measuring the quantity of copper reduced, either by its disappearance or by reducing the cuprous oxid to a metallic state, Politis has proposed a method of analysis depending on the titration of the residual copper.[103] The reagents employed are:

The reaction is represented by the formula

2CuCl₂ + 4KI = Cu₂I₂ + 4KCl + I₂.

The analytical process is carried out as follows: In a 100 cubic centimeter flask are boiled fifty cubic centimeters of the copper solution, ten cubic centimeters of about one-tenth per cent reducing sugar solution are added, the boiling continued for five minutes, the flask filled to the mark with boiling water and its contents filtered. Fifty cubic centimeters of the hot filtrate are cooled, slightly acidified, potassium iodid solution added in slight excess; and the iodin set free determined by titration with sodium thiosulfate. The quantity of iodin obtained corresponds to the unreduced copper remaining after treatment with the reducing sugar. The number of cubic centimeters of thiosulfate used subtracted from twenty-five will give the number of cubic centimeters of the copper solution which would be reduced by five cubic centimeters of the sugar solution used.

Example.—In the proportions given above it was found that eleven cubic centimeters of thiosulfate were required to saturate the iodin set free. Then 25 - 11 = 14 cubic centimeters of copper solution reduced by five cubic centimeters of the sugar solution. Since one cubic centimeter of the copper solution is reduced by 0.0036 gram of dextrose the total dextrose in the five cubic centimeters = 0.0036 × 5 = 0.0180 gram.

The above method does not seem to have any practical advantage over those based on noting the disappearance of the copper and is given only to illustrate the principle of the process. While the titration of the iodin by sodium thiosulfate is easily accomplished in the absence of organic matter, it becomes difficult, as shown by Ewell, when organic matters are present, as they always are in the oxidation of a sugar solution. Ewell has therefore proposed to determine the residual copper by a standard solution of potassium cyanid, but the method has not yet been developed.[104]

135. General Principles.—In the preceding pages the principles of the volumetric methods of sugar analysis by means of alkaline copper solution have been set forth. They depend either on the total decomposition of the copper solution employed by the reducing sugar, or else on the collection and titration of the cuprous oxid formed in the reaction. In the gravimetric methods the general principle of the process rests upon the collection of the cuprous oxid formed and its reduction to metallic copper, the weight of which serves as a starting point in the calculations of the weight of reducing sugar, which has been oxidized in the solution.

The factors which affect the weight of copper obtained are essentially those which influence the results in the volumetric method. The composition of the copper solution, the temperature at which the reduction is accomplished, the time of heating, the strength of the sugar solution and the details of the manipulation, all affect more or less the quantity of copper obtained. As in the volumetric method also, the kind of reducing sugar must be taken in consideration, dextrose, levulose, invert sugar, maltose and other sugars having each a definite factor for reduction in given conditions. It follows, therefore, that only those results are of value which are obtained under definite conditions, rigidly controlled.

136. Gravimetric Methods of the Department of Agriculture Laboratory.—The process used in this laboratory is based essentially on the methods of Maercker, Behrend, Morgen, Meissl, Hiller and Allihn.[105] Where dextrose alone is present, the table of factors proposed by Allihn is used and also the copper solution corresponding thereto.

For pure invert sugar, the tables and solutions of Meissl are used. For invert sugar in the presence of sucrose, the table and process proposed by Hiller are used.

Figure 43. Apparatus for the Electrolytic Deposition of Copper.

The reduction of the copper solution and the electrolytic deposition of the copper are accomplished as follows:

The copper and alkali solutions are kept in separate bottles. After mixing the equivalent volume of the two solutions in a beaker, heat is applied and the mixture boiled. To the boiling liquid the proper volume of the cold sugar solution is added. This must always be less than the amount required for complete reduction. The solution is again brought into ebullition and kept boiling exactly two minutes. A two-minute sand glass is conveniently used to determine the time of boiling. At the end of this time an equal volume of freshly boiled cold water is added, and the supernatant liquor at once passed through a gooch under pressure. The residual cuprous oxid is covered with boiling water and washed by decantation until the wash water is no longer alkaline. It is more convenient to wash in such a way that, at the end, the greater part of the cuprous oxid is in the gooch. The felt and cuprous oxid are then returned to the beaker in which the reduction is made. The gooch is moistened with nitric acid to dissolve any adhering oxid and then is washed into the beaker. Enough nitric acid is added to bring all the oxid into solution, an excess being avoided, and a small amount of water added. The mixture is again passed under pressure through a gooch having a thin felt, to remove the asbestos and the filtrate collected in a flask of about 150 cubic centimeters capacity. The washing is continued until the gooch is free of copper, when the volume of the filtrate should be about 100 cubic centimeters. The liquid is transferred to a platinum dish holding about 175 cubic centimeters and the flask washed with about twenty-five cubic centimeters of water. From three to five cubic centimeters of strong sulfuric acid are added and the copper deposited by an electric current.

137. Precipitating the Copper.—When no more nitric acid is used than indicated in the previous paragraph, it will not be necessary to remove it by evaporation. The platinum dishes containing the solutions of the cuprous oxid are arranged as shown in the figure for the precipitation of the copper by the electric current. Each of the supporting stands has its base covered with sheet-copper, on which the platinum dishes rest. The uprights are made of heavy glass rods and carry the supports for the platinum cylinders which dip into the copper solutions. The current used is from the city service and is brought in through the lamp shown at the right of the figure. This current has a voltage of about 120. After passing the lamp it is conducted through the regulator shown at the right, a glass tube closed below by a stopper carrying a piece of platinum foil, and above by one holding a glass tube, in the lower end of which is sealed a piece of sheet platinum connected, through the glass tube, with the lamp. The regulating tube contains dilute sulfuric acid. The strength of current desired is secured by adjusting the movable pole. A battery of this kind easily secures the precipitation of sixteen samples at once, but only twelve are shown in the figure. The practice here is to start the operation at the time of leaving the laboratory in the afternoon. The next morning the deposition of the copper will be found complete. The wiring of the apparatus is shown in the figure. The wire from the regulator is connected with the base of the first stand, and thence passes through the horizontal support to the base of the second, and so on. The return to the lamp is accomplished by means of the upper wire. This plan of arranging the apparatus has been used for two years, and with perfect satisfaction.

Where a street current is not available, the following directions may be followed: Use four gravity cells, such as are employed in telegraphic work, for generating the current. This will be strong enough for one sample and by working longer for two. Connect the platinum dish with the zinc pole of the battery. The current is allowed to pass until all the copper is deposited. Where a larger number of samples is to be treated at once, the size of the battery must be correspondingly increased.

138. Method Used at the Halle Station.—The method used at the Halle station is the same as that originally described by Maercker for dextrose.[106] The copper solution employed is the same as in the allihn method, viz., 34.64 grams of copper sulfate in 500 cubic centimeters, and 173 grams of rochelle salt and 125 grams of potassium hydroxid in the same quantity of water. In a porcelain dish are placed thirty cubic centimeters of copper solution and an equal quantity of the alkali, sixty cubic centimeters of water added and the mixture boiled. To the solution, in lively ebullition, are added twenty-five cubic centimeters of the dextrose solution to be examined which must not contain more than one per cent of sugar. The mixture is again boiled and the separated cuprous oxid immediately poured into the filter and washed with hot water, until the disappearance of an alkaline reaction. For filtering, a glass tube is employed, provided with a platinum disk, and resembling in every respect similar tubes used for the extraction of substances with ether and alcohol. The arrangement of the filtering apparatus is shown in Fig. 44. In the Halle method it is recommended that the tubes be prepared by introducing a platinum cone in place of the platinum disk and filling it with asbestos felt, pressing the felt tightly against the sides of the glass tube and making the asbestos fully one centimeter in thickness. This is a much less convenient method of working than the one described above. After filtration and washing, the cuprous oxid is washed with ether and alcohol and dried for an hour at 110°, and finally reduced to metallic copper in a stream of pure dry hydrogen, heat being applied by means of a small flame. The apparatus for the reduction of the cuprous oxid is shown in Fig. 45. The metallic copper, after cooling and weighing, is dissolved in nitric acid, the tube washed with water, ether and alcohol, and again dried, when it is ready for use a second time. The percentage of dextrose is calculated from the milligrams of copper found by Allihn’s table.

Figure 44. Apparatus for Filtering Copper Suboxid.

Figure 45. Apparatus for Reducing Copper Suboxid.

139. Tables for Use in the Gravimetric Determination of Reducing Sugars.—The value of a table for computing the percentage of a reducing sugar present in a solution, is based on the accuracy with which the directions for the determination are followed. The solution must be of the proper strength and made in the way directed. The degree of dilution prescribed must be scrupulously preserved and the methods of boiling during reduction and washing the reduced copper, followed. The quantity of copper obtained by the use of different alkaline copper solutions and of sugar solutions of a strength different from that allowed by the fixed limits, is not a safe factor for computation. It must be understood, therefore, that in the use of the tables the directions which are given are to be followed in every particular.

140. Allihn’s Gravimetric Method for the Determination of Dextrose.—Reagents:

Manipulation: Place thirty cubic centimeters of the copper solution (I), thirty cubic centimeters of the alkaline tartrate solution (II), and sixty cubic centimeters of water in a beaker and heat to boiling. Add twenty-five cubic centimeters of the solution of the material to be examined, which must be so prepared as not to contain more than one per cent of dextrose, and boil for two minutes. Filter immediately after adding an equal volume of recently boiled cold water and obtain the weight of copper by one of the gravimetric methods given. The corresponding weight of dextrose is found by the following table:

Allihn’s Table for the
Determination of Dextrose.

• (A) = Milligrams of copper.
• (B) = Milligrams of dextrose.

141. Meissl’s Table for Invert Sugar.—Invert sugar is usually the product of the hydrolysis of sucrose. The following table is to be used when the hydrolysis is complete, i. e., when no sucrose is left in the solution. The solution of copper sulfate and of the alkaline tartrate are made up as follows: 34.64 grams of copper sulfate in half a liter, and 173 grams of rochelle salt and 51.6 grams sodium hydroxid in the same volume. The quantity of sugar solution used must not contain more than 245 nor less than ninety milligrams of invert sugar.

In the determination twenty-five cubic centimeters of the copper solution and an equal volume of the alkaline tartrate are mixed and boiled, the proper amount of sugar solution added to secure a quantity of invertose within the limits named, the volume completed to 100 cubic centimeters with boiling water, and the mixture kept in lively ebullition for two minutes. An equal volume of recently boiled cold water is added and the cuprous oxid at once separated by filtration on asbestos under pressure, and washed free of alkali with boiling water. The metallic copper is secured by one of the methods already described.

Table for Invert Sugar by Meissl and Wien.[107]

• (A) = Milligrams of copper.
• (B) = Milligrams of invert sugar.

142. Table for the Determination of Invert Sugar (Reducing Sugars) in the Presence of Sucrose.—The method adopted by the Association of Official Agricultural Chemists is essentially that proposed by Meissl and Hiller.[108] Prepare a solution of the material to be examined in such a manner that it contains twenty grams of the mixed sugars in one hundred cubic centimeters, after clarification and the removal of the excess of lead. Prepare a series of solutions in large test tubes by adding one, two, three, four, five etc. cubic centimeters of this solution to each tube successively. Add five cubic centimeters of the mixed copper reagent to each, heat to boiling, boil two minutes and filter. Note the volume of sugar solution which gives the filtrate lightest in tint, but still distinctly blue. Place twenty times this volume of the sugar solution in a 100 cubic centimeter flask, dilute to the mark, and mix well. Use fifty cubic centimeters of the solution for the determination, which is conducted as already described, until the weight of copper is obtained. For the calculation of the results use the following formulas and table of factors of Meissl and Hiller:[109]

Z indicates the vertical column, and the ratio of R to I, the horizontal column of the table, which are to be used for the purpose of finding the factor (F) for calculating copper to invert sugar.

Example:—The polarization of a sugar is 86.4, and 3.256 grams of it (W) are equivalent to 0.290 gram of copper. Then:

100 - R = 100 - 95.1 = 4.9 = I

R : I = 95.1 : 4.9

By consulting the table it will be seen that the vertical column headed I = 150 is nearest to Z, 145, the horizontal column headed 95: 5 is nearest to the ratio of R to I, 95.1: 4.9. Where these columns meet we find the factor 51.2, which enters into the final calculation:

Meissl and Hiller’s Factors for the Determination of
More Than One Per Cent of Invert Sugar.

143. Table for the Estimation of Milk Sugar.—The solutions to be used for this table are the same as those employed in the preceding table for the estimation of invert sugar. The milk sugar is supposed to be in a pure form in solution before beginning the analysis. The method to be employed for milk will be given in the part devoted to dairy products.

In the conduct of the work twenty-five cubic centimeters of the copper solution are mixed with an equal quantity of the alkaline tartrate mixture, and from twenty to one hundred cubic centimeters of the sugar solution added, according to its concentration. This solution should not contain less than seventy nor more than 306 milligrams of lactose. The volume is completed to 150 cubic centimeters with boiling water and kept in lively ebullition for six minutes. The rest of the operation is conducted in the manner already described. From the weight of copper obtained the quantity of milk sugar is determined by inspecting the table. It is recommended to use such a weight of milk sugar as will give about 200 milligrams of copper.

Table for Determining Milk Sugar.

• (A) = Milligrams of copper.
• (B) = Milligrams of milk sugar.

144. Table for the Determination of Maltose.—The copper and alkaline solutions employed for the oxidation of maltose are the same as those used for invert and milk sugars.

In the manipulation twenty-five cubic centimeters each of the copper and alkali solutions are mixed and boiled and an equal volume of the maltose solution added, which should not contain more than one per cent of the sugar. The boiling is continued for four minutes, an equal volume of cold recently boiled water added, the cuprous oxid separated by filtration and the metallic copper obtained in the manner already described. The weight of maltose oxidized is then ascertained from the table.

• Example. Weight of impure maltose taken, ten grams to a liter:
• Quantity used, twenty-five cubic centimeters:
• Weight of copper obtained 268 milligrams:
• Weight of maltose oxidized 237 milligrams:
• Weight of impure maltose taken 250 milligrams:
• Percentage of maltose in sample 94.8.

Table for Maltose.

• (A) = Milligrams of copper.
• (B) = Milligrams of maltose.