turn leads to a central chamber connected with the vacuum pump or water aspirator. Each suction tube passing from the filter to the vacuum header may be closed by means of the stopcock so that the filtering process for any individual unit may be interrupted without interference with the others. The central vacuum chamber is connected with a mercury manometer or other vacuum gage. (c) PREPARATION AND FILTRATION OF THE SOLUTION WITH ASBESTOS [42] A generous supply of solution should be available, the amount to be taken depending upon the depth of color. For darker products, 100 g of solution may be sufficient, whereas 200 g or more may be neces FIGURE 76.-Peters and Phelps filter battery. sary for white sugars. The stability of transparency is, in large part, dependent upon high density. It is therefore considered necessary that the dry-substance concentration of the solution be not lower than 60 Brix, which is compatible with reasonably rapid filtration. Time is saved by dissolving and diluting on a weight basis. The required weights of sample and water are calculated as follows: Then g solution required XBrix of solution Brix of sample g sample to be taken. g water to be added=g solution-g sample. For nearly dry products of relatively high purity, such as 96 test raw sugars, a sufficiently close approximation results if polarization is substituted for Brix. The calculated amount of sample to provide a 60-Brix solution is weighed into a tared flask on a rough balance and weights equivalent to the required water are added to the pan. A small amount of hot distilled water is added and the flask is placed in a water bath heated to 50° C and shaken to promote rapid solution. Hot water is added to the flask a little at a time until the sample is dissolved. The flask is dried on the outside and returned to the balance pan where dilution to the required weight may be completed. Purified dry asbestos is added to the solution in amount depending upon the quantity and character of the suspended matter present. Usually 0.5 g will suffice, but when slimy material is present and the solution is refractory in filtration, 1 to 2 g should be used. The flask is loosely stoppered and shaken gently at first until the warm air is expelled. The stopper is then tightened and the flask is shaken vigorously to mix the contents thoroughly and to permit the asbestos fibers and the suspended matter to become entangled. The flask is returned to the bath and the solution is allowed to warm for a short time. Although it is possible to obtain a satisfactory filtrate with a single filtration, particularly in the case of easy filtering solutions, it is customary to filter twice. Two filters are therefore prepared as directed in section (b), p. 325. The first, or preliminary filter, immediately before filtration, is warmed by washing with a small amount of hot distilled water, which is aspirated from the asbestos pad as thoroughly as possible. The warm solution is then added and a few milliliters filtered to displace the water remaining in the pad when the suction is closed off and a clean dry receiver is substituted. When the filtration is resumed the remainder of the solution, or as much as the filter will hold, is added. The pad is to be kept covered with solution during the filtration, at the end of which the suction is stopped before the pad becomes uncovered. The receiver is detached and returned to the bath for further warming while the second filter is being rinsed with hot water and drained. The second filtration is performed exactly as the first but no asbestos is added to the first filtrate. The main portion of the filtrate is collected in a clean receiver. This filtrate is cooled and adhering condensed water is wiped from the neck of the bottle which is then closed with a clean, dry stopper and shaken to mix the contents throughly. The refractometric Brix of the optical filtrate is determined and c (g dry substance per 1 ml) is obtained by reference to table 114, p. 632). The solution, if not too dark, is now ready for photometric observation. (d) PREPARATION OF DIATOMACEOUS EARTH AND FILTRATION Diatomaceous earth, also known as infusorial earth or kieselguhr, on account of its availability and ease of application, is preferred in factory laboratories for the clarification of sugar solutions for colorimetry. As noted by Balch [57], variable quantities of earth must be used to obtain satisfactory clarification, depending upon the nature and amount of suspended matter present as applying to different grades of sugar products. Zerban and Sattler [58] found that the same statement applies to products of a single grade (raw sugar), and that there is a gradual falling off in-logt in some cases with increasing quantities of earth used, and that more earth than the quantities tested (5 g to 50 ml of 60 Brix) would have to be used to reach a limiting effect. Experience at this Bureau [42] and as reported elsewhere [59] has led to the belief that diatomaceous earth has a distinct decolorizing effect upon sugar solutions beyond the removal of visible turbidity, depending upon the quantity of earth used and the number of filtrations. On the basis of time consumed in obtaining satisfactory filtrates, a comparison of clarification with asbestos or with diatomaceous earth probably leaves little to choose, much depending upon the development of technique. For technical purposes, where an error of a few percent is of little consequence, and where the employment of earth is more convenient, its use is recommended, particularly for the filtration of thin juices. Diatomaceous earth should be treated for the removal of interfering substances, as described below. Methods for purifying diatomaceous earth by means of dilute hydrochloric acid have been described by Zerban and Sattler [58] and by Honig [60]. In the method employed at this Bureau, 100 g of earth in an Erlenmeyer flask is treated with 400 ml of strong hydrochloric acid and 40 ml of strong nitric acid. The contents of the flask are mixed and heated for 30 minutes on the steam bath. The mixture is then diluted with 2 liters of cold water, and the earth is allowed to settle for a few minutes when the liquid is decanted. The earth is then filtered with suction in a Büchner funnel and washed with relays of dilute hydrochloric acid (20-percent concentrated acid by volume) until the filtrate shows absence of iron, then with distilled water for removal of hydrochloric acid. The product is dried at 110° C. and Sattler. Zerban and Sattler boil 75 g of earth with a mixture of 100 ml of concentrated hydrochloric acid and 900 ml of distilled water and filter hot, then wash thoroughly with hot water. This treatment is repeated three times, and the material is dried and ignited in a muffle furnace. FIGURE 77.-Filter ap- For the clarification of solutions of washed raw paratus of Zerban or better-grade sugars, 100 to 200 ml of the 60Brix solution is treated in a suitable flask, with 2 percent of the purified earth based on the weight of solids. For badly contaminated products, more earth must be used. The flask is closed with a rubber stopper and vigorously shaken, and the mixture is filtered through a double layer of a good grade of paper, 5.5 cm in diameter, in a Büchner funnel. The paper should first be washed with a little water. When the filtrate has become clear, a clean, dry receiver is substituted and the filtration is completed. As suggested by Balch [57], receivers may be changed without the necessity of breaking the vacuum if the Büchner funnel is connected to a fractional distillation receiver designed for distillation under reduced pressure. To complete the filtration without breaking the vacuum, Zerban and Sattler use the apparatus shown in figure 77. Balch states that the more resistant the suspended material is to filtration, the more earth must be used which may amount to as much as 10 percent on solids, and for cane juices or lime-defecated juices the maximum will undoubtedly be necessary. Honig [60] treats 200 ml of 55-Brix solution of sugar in a 300-ml Erlenmeyer flask with 2 g of prepared Hyflo Supercel. After mixing, the solution is filtered on a Büchner funnel through a hardened paper that has been slightly moistened beforehand. The suction is turned on and the entire amount of the mixture is added to the filter. The filtrate is refiltered five times through the original layer of filter-aid, finally being collected in a clean receiver. 7. DILUTION OF COLOR [40, 42] A dilution of color is necessary when the saccharine product is too dark to permit good photometric readings, particularly when observations are to be made in the blue portion of the spectrum. Dilution with water alone causes separation of colloids which produce turbidity. The dilution may be accomplished, however, by the addition of a concentrated, highly decolorized sugar solution to the turbid dark solution, or by mixing a known weight of white sugar with the sample, and diluting. To prepare the diluent sirup, a 60-Brix solution of the best obtainable grade of white sugar is heated in a bath to 90° C. Decolorizing carbon equivalent to 2 percent of the weight of sugar solids is added and the mixture is warmed and shaken during 15 minutes. Prepared kieselguhr (see section 6 (d), p. 327) 5 percent on solids, is added and mixed with the solution, which is then filtered through paper in a large Büchner funnel, cloudy first-runnings being returned to the filter. The solution is refiltered through an asbestos pad. The color remaining in the sirup after a single carbon treatment depends upon the color of the original sugar. Usually there is detectable light absorption in the blue part of the spectrum, but this should be so slight as to be negligible in the final results. It is possible, by repeated carbon treatment, to prepare solutions that absorb no mercury blue-violet in a 20-cm thickness of liquid column. The dilution of color with sirup may be carried out on a volume or a weight basis. The volume basis is more rapid, but the weight basis is more precise, since deduction may be made for any absorption by the diluent. When dry white sugar is used, the dilution is necessarily on a weight basis. In the following description of the procedure the same colored product and the same diluent sugar were used, the latter being a specimen sold as pure sucrose. In procedures (a) and (b) the diluent solution was not treated with decolorizing carbon but was filtered through asbestos. (a) VOLUME BASIS The two solutions, which should be at the same temperature (20° to 25° C), are pipetted into a flask, the pipettes used being graduated to contain a definite volume. Both pipettes are rinsed with the solution in the flask, after it is mixed, in order that the adhering solution may be of the same concentration as that in the flask. The mixture is filtered through asbestos at room temperature, and the transmittancy is measured with the photometer. The following example illustrates the calculations. Colored solution, refractometric Brix Colored solution, c, from table 114.. Diluent solution, volume taken. 60. 3 0. 7768 g/ml. 20. 0 ml. 100. 0 ml. The mixture then contains one-sixth of its volume of colored solution, and c, the colored dry substance of the mixture, therefore beomes 0.7768/6=0.1295. This procedure is longer than (a) but is capable of yielding more precise results, since a correction is introduced for any absorption by the diluent sirup. It is therefore necessary to know the dry-substance content of both colored solution and diluent and log t of the diluent at the wave lengths selected. The same values are then determined for the mixture and the corrections are applied. A portion of the colored solution is weighed in a tared flask or beaker, and the diluent in the desired amount is added and the whole reweighed and mixed, the weight of added diluent being found by difference. The mixture is warmed to 60° C, filtered with asbestos, cooled, and the refractometric Brix is determined. The transmittancy of the mixture in appropriate cells is then read at the selected wave lengths with the photometer. The manner of applying the corrections is illustrated by the following example, in which the values for the diluent sirup are first determined. Example. In the diluent sirup Brix=60.9, corresponding to c=0.7867 (table 114, p. 632). Cell length, b=20 cm. TA-560=0.887, Tx-436=0.699. From table 129 then we obtain -log T-560=0.0521, -log Tx-436=0.1555. Substituting in the Lambert-Beer equation, we have In the colored solution, Brix=60.3 corresponding to c=0.7768. Weight of mixture after adding ailuent. Weight of dry substance in colored solution=60.3 percent of 7.4862 44.6748 12. 415 g. 73. 480 g. 61.065 g. 12.415=7. 4862 g. -37. 1886 g. = 44. 6748 g. = 16. 75 Percentage of colored dry substanceAfter filtration and cooling, the refractometric Brix of the final filtrate was 62.0, or c=0.8049. Assuming that the proportions of colored and diluent dry substance do not change upon concentration, we have in the final filtrate c = 16.75 percent of 0.8049=0.1348 and diluent dry substance, cp=0.8049-0.1348=0.6701. The photometric readings of this mixture in a 2-cm cell were T-560=0.72 and TAA-4380.33. Corresponding to log T-560 0.1427 and -log T-436=0.4815. Substituting in the Lambert-Beer equation, we have = |