Brown and Morris[117] report results of their work in extending Raoult’s investigations of the molecular weight of the carbohydrates. The process is carried on as follows:

A solution of the carbohydrate is prepared containing a known weight of the substance in 100 cubic centimeters of water. About 120 cubic centimeters of the solution are introduced into a thin beaker of about 400 capacity. This beaker is closed with a stopper with three holes. Through one of these a glass rod for stirring the solution is inserted. The second perforation carries a delicate thermometer graduated to 0°.05. The temperature is read with a telescope. The beaker is placed in a mixture of ice and brine at a temperature from 2° to 3° below the freezing point of the solution. The solution is cooled until its temperature is from 0°.5 to 1° below the point of congelation. Through the third aperture in the stopper a small lump of ice taken from a frozen portion of the same solution, is dropped, causing at once the freezing process to begin. The liquid is briskly stirred and as the congelation goes on the temperature rises and finally becomes constant. The reading is then taken. The depression in the freezing point, controlled by the strength of the solution, should never be more than from 1° to 2°.

The molecular weights may also be determined by the boiling points of their solutions as indicated by the author,[118] Beckmann,[119] Hite, Orndorff and Cameron.[120]

The method applied to some of the more important carbohydrates gave the following results:

149. Birotation.—As is well known, dextrose exhibits in fresh solutions the phenomenon of birotation. The authors supposed that this phenomenon might have some relation to the size of the molecule. They, therefore, determined the molecular volume of freshly dissolved dextrose by the method of Raoult and found M = 180. The high rotatory power of recently dissolved dextrose is therefore not due to any variation in the size of its molecule.

The mathematical theory of birotation is given by Müller as follows.[121] In proportion as the unstable modification A is transformed into the stable modification B, the rotation will vary. Let ρ = the specific rotatory power of B and aρ = that of A, both in the anhydrous state. Let now p grams of the substance be dissolved in V cubic centimeters of solvent and observed in a tube l decimeters in length. The time from making the solution is represented by θ. The angle of rotation α is read at the time θ. Let x = the mass of A, and y = that of B, and the equation is derived.

But x + y = p

If now there be introduced into the calculation the final angle of rotation α_n, which can be determined with great exactness; we have

This equation gives the quantity x of the unstable matter which is transformed into the stable modification in the time θ.

It must be admitted that the quantity dx which is changed during the infinitely small time dθ is proportional to the mass x which still exists at the moment θ, whence dx = -Cʹxdθ where Cʹ represents a constant positive factor. From this is derived the equation

Integrating and calling x the quantity of matter changed to the stable form at the moment θ, corresponding to a rotation α₀, we have

and taking into consideration the equation given above, and substituting common for superior logarithms we get

Experience has shown that such a constant C really exists, and its value can be easily calculated from the data of Parcus and Tollens.[122] The mean value of C from these data is 0.0301 for arabinose; 0.0201 for xylose; 0.0393 for rhamnose; 0.0202 for fucose; 0.00927 for galactose; 0.00405 for lactose; 0.00524 for maltose, and for dextrose, 0.00348 at 11° to 13° and 0.00398 from 13° to 15°. The constant C as is well known, increases as the temperature is raised.

The constant C, at a given temperature, measures the progress of the phenomenon of the change from the unstable to the stable state. It will be noticed that among the sugars possessing multirotation properties the pentoses possess a much higher speed of transformation than the others.

150. Estimation of Pentose Sugars and Pentosans as Furfurol.—The production of furfurol by distilling carbohydrates with an acid has already been mentioned. Tollens and his associates have shown that with pentose sugars, and carbohydrate bodies yielding them, the production of furfurol is quantitive.

The production and estimation of furfurol have been systematically studied by Krug, to whose paper the reader is referred for the complete literature of the subject.[123] The essential principles of the operation are based on the conversion of the pentoses into furfurol by distilling with a strong acid, and the subsequent precipitation and estimation of the furfurol formed in the first part of the reaction.

The best method of conducting the distillation is as follows:

Five grams of the pentose substance are placed in a flask of about a quarter liter capacity, with 100 cubic centimeters of hydrochloric acid of 1.06 specific gravity. The arrangement of the apparatus is shown in Fig. 46. The flame of the lamp is so regulated as to secure about two cubic centimeters of distillate per minute.

The distillate is received in a graduated cylinder and as soon as thirty cubic centimeters are collected, an equal quantity of hydrochloric acid, of the strength noted, is added to the distilling flask, allowing it to flow in slowly so as not to stop the ebullition. The process is continued until a drop of the distillate gives no sensible reaction for furfurol when tested with anilin acetate. The test is applied as follows: Place a drop of the distillate on a piece of filter paper moistened with anilin acetate. The presence of furfurol will be disclosed by the production of a brilliant red color. Usually about three hours are consumed in the distillation, during which time a little less than 400 cubic centimeters of distillate is obtained. The distillate is neutralized with solid sodium carbonate and, in order to have always the same quantity of common salt present, 10.2 grams of sodium chlorid are added for each fifty cubic centimeters of water necessary to make the total volume to half a liter.[124]

The reactions with pentosans probably consist in first splitting up of the molecule into a pentose and the subsequent conversion of the latter into furfurol according to the following equations:

151. Determination of Furfurol.—The quantity of furfurol obtained by the process mentioned above may be determined in several ways.

As Furfuramid.—When ammonia is added to a saturated solution of furfurol, furfuramid, (C₅H₄O)₃N₂, is formed. In order to secure the precipitate it is necessary that the furfurol be highly concentrated and this can only be accomplished by a tedious fractional distillation. This method, therefore, has little practical value.

As Furfurolhydrazone.—Furfurol is precipitated almost quantitively, even from dilute solutions, by phenylhydrazin. The reaction is represented by the equation:

152. Volumetric Methods.—Tollens and Günther have proposed a volumetric method which is carried out as follows:[125] The distillation is accomplished in the manner described. The distillate is placed in a large beaker, neutralized with sodium carbonate and acidified with a few drops of acetic. Phenylhydrazin solution of known strength is run in until a drop of the liquid, after thorough mixing, shows no reaction for furfurol with anilin acetate. The reagent is prepared by dissolving five grams of pure phenylhydrazin and three of glacial acetic acid in distilled water, and diluting to 100 cubic centimeters. The solution is set by dissolving from two-tenths to three-tenths gram of pure furfurol in half a liter of water and titrating with the phenylhydrazin as indicated above. The quantity of the pentose used has a great influence on the result.

With nearly a gram of arabinose about fifty per cent of furfurol were obtained while when nearly five grams were used only about forty-six per cent of furfurol were found. With xylose a similar variation was found, the percentage of furfurol, decreasing as the quantity of pentose increased. The method, therefore, gives only approximately accurate results.

153. Method of Stone.—Another volumetric method proposed by Stone is based on the detection of an excess of phenylhydrazin by its reducing action on the fehling reagent.[126] A standard solution of phenylhydrazin is prepared by dissolving one gram of the hydrochlorate and three grams of sodium acetate in water and completing the volume of the liquor to 100 cubic centimeters. This solution contains 1.494 milligrams of phenylhydrazin in each cubic centimeter, theoretically equivalent to 1.328 milligrams of furfurol. The reagent is set by titrating against a known weight of furfurol. Pure furfurol may be prepared by treating the crude article with sulfuric acid and potassium dichromate, and subjecting the product to fractional distillation. The distillate is treated with ammonia and the furfuramid formed is purified by recrystallizing from alcohol and drying over sulfuric acid. One gram of this furfuramid is dissolved in dilute acetic acid and the volume completed to one liter with water.[127] The phenylhydrazin solution being unstable, is to be prepared at the time of use.

The titration is conducted as follows: Twenty-five cubic centimeters of the distillate obtained from a pentose body, by the method described above, are diluted with an equal volume of water, a certain quantity of the phenylhydrazin solution added to the mixture from a burette and the whole heated quickly to boiling. The flask is rapidly cooled and a portion of its contents poured on a filter. The filtrate should have a pale yellow color and be perfectly clear. If it become turbid on standing, it should be refiltered. Two cubic centimeters of the clear filtrate are boiled with double the quantity of the fehling reagent. If phenylhydrazin be present, the color of the mixture will change from blue to green. By repeating the work, with varying quantities of phenylhydrazin, a point will soon be reached showing the end of the reaction in a manner entirely analogous to that observed in volumetric sugar analysis.

In practice the volumetric methods have given place to the more exact gravimetric methods described below.

154. Gravimetric Methods.—The distillation is carried on and the volume of the distillate completed to half a liter as described above. Chalmot and Tollens then proceed as follows:[128] Ten cubic centimeters of a solution of phenylhydrazin acetate, containing in 100 cubic centimeters twelve grams of the phenylhydrazin and seven and a half grams of glacial acetic acid dissolved and filtered, are added to the distillate and the mixture stirred with an appropriate mechanism for half an hour. The furfurolhydrazone at the end of this time will have separated as small reddish-brown crystals. The mixture is then thrown onto an asbestos filter and the liquid separated with suction. The suction should be very gradually applied so as not to clog the felt. The precipitate adhering to the beaker is washed into the filter with 100 cubic centimeters of water. The precipitate is dried at about 60° and weighed. As a check the hydrazone may be dissolved in hot alcohol, the filter well washed, dried and again weighed. To obtain the weight of furfurol the weight of hydrazone found is multiplied by 0.516 and 0.025 added to compensate for the amount which was held in solution or removed by washing. Less than one per cent of pentose can not be determined by this method since that amount is equalled by the known losses during the manipulation.

Factor.—To convert the furfurol found into pentoses, the following factors are used:

155. Method Of Krug.—In conducting the determination of furfurol, according to the method of Chalmont and Tollens just noticed, Krug observed that the filtrate, after standing for some time, yielded a second precipitate of furfurol hydrazone. Great difficulty was also experienced in collecting the precipitate upon the filter on account of the persistency with which it stuck to the sides of the vessel in which the precipitation took place.[129] In order to avoid these two objections, Krug modified the method as described below and this modified method is now exclusively used in this laboratory.

After the precipitation of the furfurol hydrazone, it is stirred vigorously, by means of an appropriate mechanical stirrer, for at least half an hour and then allowed to rest for twenty-four hours. On filtering after that length of time the filtrate remains perfectly clear and no further precipitation takes place. After the filtration is complete and the beaker and filtering tube well washed, no attempt is made to detach the part of the filtrate adhering to the beaker but the whole of the precipitate, both that upon the filter and that adhering to the sides of the beaker, is dissolved in strong alcohol, from thirty to forty cubic centimeters being used. The alcoholic solution is collected in a small weighed flask, the alcohol evaporated at a gentle heat and the last traces of water removed by heating to 60° and blowing a current of dry air through the flask. After weighing the precipitate of furfurol hydrazone, obtained as above, the calculation of the weight of pentose bodies is accomplished by means of the usual factors.

156. Precipitation of Furfurol with Pyrogalol.—Furfurol is thrown out of solution in combination with certain phenol bodies by heating together in an acid solution. Hotter has proposed a method for the determination of furfurol based on the above fact.[130] The furfurol is obtained by distillation in the manner already described and hydrochloric acid is added if necessary to secure twelve per cent of that body in a given volume. The furfurol is thrown out of an aliquot portion by heating with an excess of pyrogalol in closed tubes for about two hours at 110°. The reaction takes place in two stages, represented by the following equations:

The aliquot part of the distillate used should not contain more than one-tenth of a gram of furfurol. The precipitate formed in this way is collected on an asbestos felt, dried at 103° and weighed. The weight obtained divided by 1.974 gives the corresponding amount of furfurol. There is some difficulty experienced in loosening the precipitate from the sides of the tubes in which the heating takes place, but this defect can be overcome by heating in covered beakers in an autoclave.

157. Precipitation with Phloroglucin.—Instead of using pyrogalol for the precipitating reagent phloroglucin may be employed. The method of procedure proposed by Councler for this purpose is given below.[131] The furfurol is prepared by distillation in the usual way. The volume of the distillate obtained is completed to half a liter with twelve per cent hydrochloric acid, and an aliquot portion, varying in volume with the percentage of furfurol is withdrawn for precipitation. This portion is placed in a glass-stoppered flask with about twice the quantity of finely powdered phloroglucin necessary to combine with the furfurol present. The contents of the flask are well shaken and allowed to stand fifteen hours. The precipitate is collected on an asbestos filter, washed free of chlorin, dried at the temperature of boiling water and weighed.

The theoretical quantity of precipitate corresponding to one part of furfurol, viz., 2.22 parts, is never obtained since the precipitate is not wholly insoluble in water. The actual proportions between the precipitate and the original furfurol vary with the amount of precipitate obtained.

When the weight of the precipitate is 200 milligrams and over, 2.12 parts correspond to one part of furfurol. When the weight of the precipitate varies from fifty to 100 milligrams, the ratio is as 2.05:1 and when only about twenty-five milligrams of precipitate are obtained the ratio is as 1.98:1.

The quantity of pentose bodies corresponding to the furfurol is calculated from the factors given by Tollens in a preceding paragraph.

The reaction which takes place with furfurol and phloroglucin is simply a condensation of the reagents with the separation of water. It is very nearly represented by the following formula:

It has been shown by Welbel and Zeisel,[132] that in the presence of twelve per cent of hydrochloric acid phloroglucin itself is condensed into dark insoluble compounds. When three molecules of furfurol and two molecules of phloroglucin are present, the bodies are both condensed and separated by continued action. When from one and a quarter to three parts of phloroglucin by weight are used to one part of furfurol, the weight of the precipitate obtained under constant conditions may serve sufficiently well for the calculation of the furfurol. The precipitates contain chlorin, which they give up even in the cold, to water. For these reasons the analytical data obtained by the method of Councler, given above, are apt to be misleading. It is probable also that similar conditions may to a certain extent prevail in the separation of furfurol with phenylhydrazin, and further investigation in this direction is desirable. For the present the very best method that can be recommended for the estimation of pentoses and pentosans is the conversion thereof into furfurol and the separation of the compound with phenylhydrazin acetate.

158. Estimation of Sugars by Fermentation.—When a solution of a hexose sugar is subjected to the action of certain ferments a decomposition of the molecule takes place with the production of carbon dioxid and various alcohols and organic acids. Under the action of the ferment of yeast Saccharomyces cerevisiae the sugar yields theoretically only carbon dioxid and ethyl alcohol, as represented by the equation:

C₆H₁₂O₆ = 2C₂H₆O + 2CO₂.

The theoretical quantities of alcohol and carbon dioxid obtained according to this equation are 51.11 percent of alcohol and 48.89 per cent of carbon dioxid.

When the yeast ferment acts on cane sugar the latter first suffers inversion, and the molecules of dextrose and levulose produced are subsequently converted into alcohol and carbon dioxid as represented below:

C₁₂H₂₂O₁₁ + H₂O = 2C₆H₁₂O₆
2C₆H₁₂O₆ = 4C₂H₆O + 4CO₂.

Cane sugar, plus the water of hydrolysis, will yield theoretically 53.8 per cent of alcohol and 51.5 per cent of carbon dioxid.

In practice the theoretical proportions of alcohol and carbon dioxid are not obtained because of the difficulty of excluding other fermentative action, resulting in the formation especially of succinic acid and glycerol. Moreover, a part of the sugar is consumed by the yeast cells to secure their proper growth and development. In all only about ninety-five per cent of the sugar can be safely assumed as entering into the production of alcohol. About 48.5 per cent of alcohol are all that may be expected of the weight of dextrose or invert sugar used. Only sugars containing three molecules of carbon or some multiple thereof are fermentable. Thus the trioses, hexoses, nonoses, etc., are susceptible of fermentation, while the tetroses, pentoses, etc., are not.

159. Estimating Alcohol.—In the determination of sugar by fermentation, a rather dilute solution not exceeding ten per cent should be used. A quantity of pure yeast, equivalent to four or five per cent of the sugar used, is added, and the contents of the vessel, after being well shaken, exposed to a temperature of from 25° to 30° until the fermentation has ceased, which will be usually in from twenty-four to thirty-six hours. The alcohol is then determined in the residue by the methods given hereafter.

The weight of the alcohol obtained multiplied by 100 and divided by 48.5, will give the weight of the hexose reducing sugar which has been fermented. Ninety-five parts of sucrose will give 100 parts of invert sugar.

Example.—Let the weight of alcohol obtained be 0.625 gram. Then 0.625 × 100 ÷ 48.5 = 1.289 grams, the weight of the hexose, which has been fermented; 1.289 grams of dextrose or levulose correspond to 1.225 of sucrose.

160. Estimating Carbon Dioxid.—The sugar may also be determined by estimating the amount of carbon dioxid produced during the fermentation. For this purpose the mixture of sugar solution and yeast, prepared as above mentioned, is placed in a flask whose stopper carries two tubes, one of which introduces air free of carbon dioxid into the contents of the flask, and the other conducts the evolved carbon dioxid into the absorption bulbs. In passing to the absorption bulbs the carbon dioxid is freed of moisture by passing through another set of bulbs filled with strong sulfuric acid. During the fermentation, the carbon dioxid is forced through the bulbs by the pressure produced, or better, a slow current of air is aspirated through the whole apparatus. The aspiration is continued after the fermentation, has ceased, until all the carbon dioxid is expelled. Towards the end, the contents of the flask may be heated to near the boiling-point. The increase of the weight of the potash bulbs will give the weight of carbon dioxid obtained. A hexose reducing sugar will yield about 46.5 per cent of its weight of carbon dioxid. The calculation is made as suggested for the alcohol process.

The fermentation process has little practical value save in determining sucrose in presence of lactose, as will be described in another place.

161. Precipitation of Sugars in Combination with the Earthy Bases.—The sugars combine in varying proportions with the oxids and hydroxids of calcium, strontium and barium. Sucrose especially, furnishes definite crystalline aggregates with these bases in such a way as to form the groundwork of several technical processes in the separation of that substance from its normally and abnormally associated compounds. These processes have little use as analytical methods, but are of great value, as mentioned, from a technical point of view.

162. Barium Saccharate.—This compound is formed by mixing the aqueous solutions of barium hydroxid and sugar. The saccharate separates in bright crystalline plates or needles from the warm solution, as C₁₂H₂₂O₁₁BaO. One part of this precipitate is soluble in about forty-five parts of water, both at 15° and 100°.

163. Strontium Saccharates.—Both the mono- and distrontium saccharates are known, viz., C₁₂H₂₂O₁₁SrO + 5H₂O and C₁₂H₂₂O₁₁2SrO.

The monosalt may be easily secured by adding a few of its crystals to a mixture of sugar and strontium hydroxid solutions.

The disaccharate is precipitated as a granular substance when from two to three molecules of strontium hydroxid are added to a boiling sugar solution. The reaction is extensively used in separating the sugar from beet molasses.

164. Calcium Saccharates.—Three calcium saccharates are known in which one molecule of sugar is combined with one, two and three molecules of lime respectively.

The monosaccharate is obtained by mixing the sugar and lime in the proper proportion and precipitating by adding alcohol.

The precipitate is partly granular and partly jelly-like, and is soluble in cold water. The dicalcium compound is obtained in the same way and has similar properties. Both, on boiling, with water, form the trisaccharate and free sugar.

The tricalcium saccharate is the most important of these compounds, and may be obtained directly by mixing freshly burned and finely ground lime (CaO) with a very cold dilute solution of sugar.

The compound crystallizes with three or four molecules of water. When precipitated as described above, however, it has a granular, nearly amorphous structure, and the process is frequently used in the separation of sugar from beet molasses.

In the laboratory but little success has been had in using even the barium hydroxid as a chemical reagent, and therefore the reactions mentioned above are of little value for analytical purposes. In separating sugar from vegetable fibers and seeds, however, the treatment with strontium hydroxid is especially valuable the sugar being subsequently recovered in a free state by breaking up the saccharate with carbon dioxid. The technical use of these reactions also is of great importance in the beet sugar industry.

165. Qualitive Tests for the Different Sugars.—The analyst will often be aided in examining an unknown substance by the application of qualitive tests, which will disclose to him the nature of the saccharine bodies with which he has to deal.

166. Optical Test for Sucrose.—The simplest test for the presence of sucrose is made with the polariscope. A small quantity of the sample under examination is dissolved in water, clarified by any of the usual methods, best with alumina cream, and polarized. A portion of the liquor is diluted with one-tenth its volume of strong hydrochloric acid and heated to just 68°, consuming about fifteen minutes time in the operation. The mixture is quickly cooled and again polarized in a tube one-tenth longer than before used; or the same tube may be used and the observed reading of the scale increased by one-tenth. If sucrose be present the second reading will be much lower than the first, or may even be to the left.

167. Cobaltous Nitrate Test for Sucrose.—Sucrose in solution may be distinguished from other sugars by the amethyst violet color which it imparts to a solution of cobaltous nitrate. This reaction was first described by Reich, in 1856, but has only lately been worked out in detail. The test is applied as follows:

To about fifteen cubic centimeters of the sugar solution add five cubic centimeters of a five per cent solution of cobaltous nitrate. After thoroughly mixing the two solutions, add two cubic centimeters of a fifty per cent solution of sodium hydroxid. Pure sucrose gives by this treatment an amethyst violet color, which is permanent. Pure dextrose gives a turquoise blue color which soon passes into a light green. When the two sugars are mixed the coloration produced by the sucrose is the predominant one, and one part of sucrose in nine parts of dextrose can be distinguished. If the sucrose be mixed with impurities such as gum arabic or dextrin, they should be precipitated by alcohol or basic lead acetate, before the application of the test. Dextrin may be thrown out by treatment of the solution with barium hydroxid and ammoniacal lead acetate. The reaction may also be applied to the detection of cane sugar in wines, after they are thoroughly decolorized by means of lead acetate and bone-black. The presence of added sucrose to milk, either in the fresh or condensed state, may also be detected after the disturbing matters are thrown out with lead acetate. The presence of sucrose in honey may also be detected by this process. The reaction has been tried in this laboratory with very satisfactory results. The amethyst violet coloration with sucrose is practically permanent. On boiling the color is made slightly bluish, but is restored to the original tint on cooling. Dextrose gives at first a fine blue color which in the course of two hours passes into a pale green. A slight flocculent precipitate is noticed in the tube containing the dextrose. Maltose and lactose act very much as dextrose, but in the end do not give so fine a green color. If the solutions containing dextrose, lactose and maltose be boiled, the original color is destroyed and a yellow-green color takes its place. The reaction is one which promises to be of considerable practical value to analysts, as it may be applied for the qualitive detection of sucrose in seeds and other vegetable products.[133]

168. The Dextrose Group.—In case the carbohydrate in question shows a right-handed rotation and the absence of sucrose is established by the polariscopic observation described above, the presence of the dextrose group may be determined by the following test.[134]

Five grams of the carbohydrate are oxidized by boiling with from twenty to thirty cubic centimeters of nitric acid of 1.15 specific gravity, and then at gentle heat evaporated to dryness with stirring. If much mucic acid be present, as will be the case if the original matter contained lactose some water is added and the mixture well stirred, and again evaporated to dryness to expel all nitric acid. The residue should be of a brown color. The mass is again mixed with a little water and the acid reaction neutralized by rubbing with fine-ground potassium carbonate. The carbonate should be added in slight excess and acetic acid added to the alkaline mixture, which is concentrated by evaporation and allowed to stand a few days. At the end of this time potassium saccharate has formed and is separated from the mother liquid by pouring on a porous porcelain plate. The residue is collected, dissolved in a little water and again allowed to crystallize, when it is collected on a porous plate, as before, and washed by means of an atomizer with a little aqueous spray until it is pure white and free of any oxalic acid. The residual acid potassium saccharate may be weighed after drying and then converted into the silver salt. The potash salt for this purpose is dissolved in water, neutralized with ammonia and precipitated with a solution of silver nitrate. The precipitate is well stirred, collected on a gooch and washed and dried in a dark place. It contains 50.94 percent of silver. All sugars which contain the dextrose group yield silver saccharate when treated as above described. Inulin, sorbose, arabinose and galactose yield no saccharic acid under this treatment, and thus it is shown that they contain no dextrose group. Milk sugar, maltose, the dextrins, raffinose and sucrose yield saccharic acid when treated as above and therefore all contain the dextrose group.

169. Levulose.—The levulose group of sugars, wherever it occurs, when oxidized with nitric acid, gives rise to tartaric, racemic, glycolic and oxalic acids, which are not characteristic, being produced also by the oxidation of other carbohydrates. A more distinguishing test is afforded by the color reactions produced with resorcin.[135] The reagent is prepared by dissolving half a gram of resorcin in thirty cubic centimeters each of water and strong hydrochloric acid. To the sugar solution under examination an equal volume of strong hydrochloric acid is added, and then a few drops of the reagent. The mixture is gently warmed, and in the presence of levulose develops a fire-red color. Dextrose, lactose, mannose and the pentoses do not give the coloration, but it is produced by sorbose in a striking degree, and also by sucrose and raffinose since these sugars contain the levulose group.

170. Galactose.—The galactose which arises from the hydrolysis of milk sugar is readily recognized by the mucic acid which it gives on oxidation with nitric acid.[136] The analytical work is conducted as follows: The body containing galactose or galactan is placed in a beaker with about sixty cubic centimeters of nitric acid of 1.15 specific gravity for each five grams of the sample used. The beaker is placed on a steam-bath and heated, with frequent stirring, until two-thirds of the nitric acid have been evaporated. The residual mixture is allowed to stand over night and the following morning is treated with ten cubic centimeters of water, allowed to stand for twenty-four hours, filtered through a gooch, and the collected matter washed with twenty-five cubic centimeters of water, dried at 100° and weighed. The mucic acid collected in this way will amount to about thirty-seven per cent of the milk sugar or seventy-five per cent of the galactose oxidized. Raffinose yields under similar treatment, about twenty-three per cent of mucic acid, which proves that the galactose group is contained in that sugar. Raffinose, therefore, is composed of one molecule each of dextrose, levulose, and galactose.

171. Invert Sugar.—The presence of a trace of invert sugar accompanying sucrose can be determined by Soldaini’s solution, paragraph 124, or by boiling with methyl blue.[137] Methyl blue is the hydrochlorate of an ammonium base, which, under the influence of a reducing agent, loses two atoms of hydrogen and becomes a colorless compound. The test for invert sugar is made as follows: The reagent is prepared by dissolving one gram of methyl blue in water. If the sugar solution is not clear, twenty grams of the sugar are dissolved in water clarified by lead subacetate, the volume completed to 100 cubic centimeters, and the solid matters separated by filtration. The filtrate is made slightly alkaline with sodium carbonate to remove the lead. A few drops of soda lye solution are then added and the mixture thrown on a filter. To twenty-five cubic centimeters of the filtrate a drop of the methyl blue solution is added, and a portion of the liquor heated in a test tube over the naked flame. If, after boiling for one minute, the coloration disappear, the sample contains at least 0.01 per cent of invert sugar; if the solution remain blue it contains none at all or less than 0.01 per cent. The test may also be made with the dilute copper carbonate solution of Ost described further on.

172. Compounds with Phenylhydrazin.—Many sugars may also be qualitively distinguished by the character of their compounds with phenylhydrazin. In general, it may be said that those sugars which reduce fehling solution form definite crystalline compounds with the reagent named. If a moderately dilute hot solution of a reducing sugar be brought into contact with phenylhydrazin acetate, a crystalline osazone is separated. The reaction takes place between one molecule of the sugar and two molecules of the hydrazin compound, according to the following formula:

C₆H₁₂O₆ + 2C₆H₅N₂H₃ = C₁₈H₂₂N₄O₄ + 2H₂O + H₂.

The hydrogen does not escape but combines in the nascent state with the excess of phenylhydrazin to form anilin and ammonia.

The precipitation is accomplished as follows:

One part by weight of the sugar, two parts of phenylhydrazin hydrochlorate, and three parts of sodium acetate are dissolved in twenty parts of water and gradually heated on the water-bath.

The osazone slowly separates in a crystalline form and it is freed from the mother liquor by filtration, and purified by solution in alcohol and recrystallization. The crystals are composed of yellow needles, which are difficultly soluble in water and more easily in hot alcohol. The crystals are not decomposed by a dilute acid but are destroyed by the action of strong acids.

Dextrosazone.—The crystals melt at from 204° to 205°, reduce fehling liquor, and dissolved in glacial acetic acid are slightly left rotating.

Levulosazone.—This body has the same properties as the dextrose compound.

Maltosazone.—This substance melts at 206° with decomposition. It is left rotating. Its structure is represented by the formula C₂₄H₃₂N₄O₉.

Galactosazone.—This substance, C₁₈H₂₂N₄O₄, has the same centesimal composition as the corresponding bodies produced from dextrose and levulose. It is distinguished from these compounds, however, by its low melting point, viz., 193°.

The above comprise all the phenylosazones which are important from the present point of view. Sucrose, by inversion, furnishes a mixture of dextros- and levulosazones when treated with phenylhydrazin, while starch and dextrin yield the dextros- or maltosazone when hydrolyzed. Lactose yields a mixture of dextros- and galactosazones when hydrolyzed and treated as above described.

The reactions with phenylhydrazin are approximately quantitive and it is possible that methods of exact determination may be based on them in the near future.[138]

173. Other Qualitive Tests for Sugars.—The analyst may sometimes desire a more extended test of qualitive reactions than those given above. The changes of color noticed on heating with alkalies may often be of advantage in discriminating between different sugars. The formation of definite compounds with the earthy and other mineral bases may also be used for qualitive determinations. One of the most delicate qualitive tests is found in the production of furfurol and this will be described in the following paragraphs.

174. Detection of Sugars and Other Carbohydrates by Means of Furfurol.—The production of furfurol (furfuraldehyd) as noted in paragraph 150, is also used quantitively for the determination of pentose sugars and pentosans.

Furfurol was first obtained from bran (furfur), whence its name, by treating this substance with sulfuric acid, diluted with three volumes of water, and subjecting the mixture to distillation. Its percentage composition is represented by the symbol C₅H₄O₂, and its characteristics as an aldehyd by the molecular structure C₄H₃O,C-HO.

Carbohydrates in general, when treated as described above for bran, yield furfurol, but only in a moderate quantity, with the exception of the pentoses.

Mylius has shown[139] that Pettenkofer’s reaction for choleic acid is due to the furfurol arising from the cane sugar employed, which, with the gall acid, produces the beautiful red-blue colors characteristic of the reaction.

Von Udránszky[140] describes methods for detecting traces of carbohydrates by the furfurol reaction, which admit of extreme delicacy. The solution of furfurol in water, at first proposed by Mylius, is to be used and it should not contain more than two and two-tenths per cent, while a solution containing five-tenths per cent furfurol is found to be most convenient. The furfurol, before using, should be purified by distillation, and, as a rule, only a single drop of the solution used for the color reaction.

The furfurol reaction proposed by Schiff[141] appears to be well suited for the detection of carbohydrates. It is made as follows:

Xylidin is mixed with an equal volume of glacial acetic acid and the solution treated with some alcohol. Strips of filter paper are then dipped in the solution and dried. When these strips of prepared paper are brought in contact with the most minute portion of furfurol, furoxylidin is formed, C₄H₃OCH(C₈H₈NH₂)₂, producing a beautiful red color. In practice, a small portion of the substance, supposed to contain a carbohydrate, is placed in a test tube and heated with a slight excess of concentrated sulfuric acid. The prepared paper is then placed over the mouth of the test tube so as to be brought into contact with the escaping vapors of furfurol.

The furfurol reaction with α-naphthol for some purposes, especially the detection of sugar in urine, is more delicate than the one just described. This reaction was first described by Molisch[142], who, however, did not understand its real nature.

The process is carried on as follows: The dilute solution should contain not to exceed from 0.05 to one-tenth per cent of carbohydrates. If stronger, it should be diluted. Place one drop of the liquid in a test tube with two drops of fifteen per cent alcoholic solution of α-naphthol, add carefully one-half cubic centimeter of concentrated sulfuric acid, allowing it to flow under the mixture. The appearance of a violet ring over a greenish fringe indicates the presence of a carbohydrate. If the substance under examination contain more than a trace of nitrogenous matter, this must be removed before the tests above described are applied.

If the liquids be mixed by shaking when the violet ring is seen, a carmine tint with a trace of blue is produced. If this be examined with a spectroscope, a small absorption band will be found between D and E, and from F outward the whole spectrum will be observed. One drop of dextrose solution containing 0.05 per cent of sugar gives a distinct reaction by this process. It can be used, therefore, to detect the presence of as little as 0.028 milligram of grape sugar. This test has been found exceedingly delicate in this laboratory, and sufficiently satisfactory without the spectroscopic adjunct.

The furfurol reaction is useful in detecting the presence of minute traces of carbohydrates but is of little value in discriminating between the different classes of these bodies.

It is not practical here to go into greater detail in the description of qualitive reactions. The analyst, desiring further information, should consult the standard works on sugar chemistry. [143]

175. Detection of Sugars by Bacterial Action.—Many forms of bacteria manifest a selective action towards sugars and this property may in the future become the basis of a qualitive and even quantitive test for sugars and other carbohydrates. Our present knowledge of the subject is due almost exclusively to the researches of Smith, conducted at the Department of Agriculture.[144] Dextrose is the sugar first and most vigorously attacked by bacterial action, and by proper precautions the whole of the dextrose may be removed from mixtures with sucrose and lactose.

The development of other forms of micro-organisms which will have the faculty of attacking other and special forms of carbohydrates is to be looked for with confident assurance of success.

DETERMINATION OF STARCH.

176. Constitution of Starch.—The molecule of starch is without doubt formed by the condensation of a large number of hexose bodies. On account of its great insolubility its molecular weight has not been determined with any degree of accuracy. Its formula may be expressed either as (C₆H₁₀O₅)ₙ or (C₁₂H₂₀O₁₀)ₙ. It is insoluble in cold water and other common solvents and does not pass into solution in any reagent without undergoing a change of structure. In hot water it forms a paste and when heated under pressure with water it undergoes a partial change and becomes soluble. Heated with acids or subjected to the action of certain ferments it suffers hydrolysis and is transformed into dextrin, maltose and dextrose. In analytical work an attempt is usually made to transform the starch entirely into dextrose, the quantity of which is then determined by some of the processes already given. All starches possess the property of giving an intensely blue color with iodin and this reaction serves to detect the most minute quantity of the material.

Starch grains derived from different sources are distinguished by differences in size and appearance. In most cases a careful examination of the starch particles will reveal their origin.[145] The greatest part of the cereal grains is composed of starch, the percentage ranging from sixty to eighty. Rice has the greatest percentage of starch in its composition of any substance. Certain root crops are also rich in starch, such as the potato, artichoke and cassava. Starch appears as one of the first products of vegetable metabolism, according to some authorities, preceding the formation of sugars. By reason of its greater complexity, however, it is more probable that the production of simple sugars precedes the formation of the more complex molecule. Starch granules are probably used as a food by the plant in the building of more complex structures and the excess of this food is stored in the seeds and in tubers.

177. Separation of Starch Particles.—Advantage is taken of the insolubility of the starch particles to secure their separation from the other vegetable structures with which they are associated. The substances containing starch are reduced to a pulp as fine as possible, and this pulp being placed in a fine cloth the starch particles are washed through the cloth with water. The milky filtrate carrying the starch is collected in an appropriate holder and, after some time, the particles subside. They may then be collected and dried. While this process is the one used commercially in the manufacture of starch, it can only give approximate data respecting the actual quantity of starch in a given weight of the sample. It is not quite possible by this method to get all the starch separated from the rest of the vegetable matter, and particles of foreign substances, such as cellulose and albuminoid matters, may pass through the filter cloth and be found with the deposited granules. It follows from this that the quantitive determination of the starch in a given sample by any direct method is only approximately exact.

178. Methods of Separation.—Hot acids cannot be safely employed to dissolve starch from its natural concomitants because other carbohydrate bodies become soluble under similar conditions. In such cases the natural sugars which are present should be removed by cold water and the starch dissolved from the residue by a diastatic ferment. Instead of this the sugars may be determined in a separate portion of the pulped material and the starch, together with the sugars, determined, and the quantity of sugar found deducted from the final result.

In these cases the final determinations are made on the sugars, after inverting the sucrose, and proceeding as directed for invert sugars in paragraph 141. The starch, after separation with diastase, is converted into dextrose by one of the methods to be given and the resulting dextrose determined by one of the approved methods.

179. Separation with Diastase.—Diastase or malt extract at a temperature of about 65° rapidly renders starch soluble. Cereals, potato meal and other starch-holding bodies are dried, first at a low temperature, and extracted with ether or petroleum to remove fat. The material is then rubbed up with water, boiled, cooled to 65°, and treated with malt extract (diastase) prepared as given below. One kilogram of ground green malt is mixed with one liter of glycerol and an equal quantity of water, and allowed to stand, with frequent shaking, for eight days. After that time the mixture is filtered, first through a small filter press and afterwards through paper. In case no filter press is at hand the mixture may be pressed in a bag and the liquor obtained, filtered. Malt extract obtained in this way will keep its diastatic properties for a long time. In its use, blank determinations must be made of the dextrose produced by treating equal portions of it with hydrochloric acid. For three grams of starchy material twenty-five cubic centimeters of the malt solution should be used and the mixture kept at 65° for two hours.[146]

180. Method in Use at the Halle Station.—The method of separating starch from cereals, potatoes and other starch-holding materials, employed at the Halle station, is essentially the same as already described.[147]

The malt extract used is prepared immediately beforehand, inasmuch as no preservative is added to it. It can be quickly prepared by digesting, for a short time at not above 50°, 100 grams of finely ground dried malt with one liter of water and separating the extract by filtration. This extract will keep only a few hours.

The material in which the starch is to be determined is dried and extracted with ether. From two to four grams of the extracted material, according to the amount of starch which it contains, are boiled for half an hour with 100 cubic centimeters of water, cooled to 65°, treated with ten cubic centimeters of malt extract and kept at the temperature named for half an hour. It is then again boiled for fifteen minutes, cooled to the temperature mentioned and again treated with malt extract as above. Two treatments with malt extract are usually sufficient to bring all the starch into solution. Finally it is again boiled and the volume completed to 250 cubic centimeters and thrown upon a filter. Two hundred cubic centimeters of the filtrate are converted into dextrose by boiling with hydrochloric acid, and the rest of the analysis is conducted in the usual manner. The dextrose value of the quantity of malt extract used must be determined upon a separate portion thereof, and the quantity of dextrose found deducted from the total amount obtained in the analysis.