60 ABSTRACTS OF CHEMLCAL PAPERS APPARATUS, ETC. Interferometer in the Brewery Laboratory. L. Adler and H. Luers. (Zeitsch. ges. Bmuw., 1916,39, 17-19,25-28,33-35,41-43; through J. Inst. Brewing, 1916,22, 504-509.)-The consfruction and d w operandi of the Lowe-Zeiss inter- ferometer are explained, and experiments are described, the object, of which was to ascertain to what extent the instrument can be usefully applied to analytical work in FIG.1. the brewery laboratory. Fig. 1 represents the instrument eiicased in a cylindrical he&insulating mantle, and supported on a stand, together with an accumulator for operaking the source of light, a box containing extra chambers, and a protective helmet for the instrument when not, in use. Fig. 2 is a diagrammatic representationAPPARATUS, ETC.61 of the working parts in plan (above) and elevation (below). The instrument is essentially one for measuring verg. small differences between the refractive indices of two liquidP. Light from the source B, following the path shown, traverses the two glass plates, PI and Pz, of tlhe compensator K , and after passing through the attemperating bath Tr, containing water, is reflected from the mirror 8, and observed throughlthe eye-piece OE. The field, as thus observed, exhibits a series of vertical interference bgnds, of which the middle ones are the most intense.When the re- K Th Tr- y- I B FIG. ‘L. cepfade W , for tthe liquids under examination is placed in the instrument, the upper half of the beam of light has to traverse this receptacle, which has parallel glass end- plates and is divided longitudjnally by a vertical partition into two chambers, so that one part of the upp& halE of the beam of light.traverses one chamber and the other part lmverses the other. If the two chambers contain liquids of exactly the same refractive index, the interference bands in the upper half of the field will correspond in position with those in the lower half, as in Fig.3; but if $he liquids differ in refractive index, one of them will cause a relative retardation of the light passing through it, and the bends in the upper part of the To restore the bands to their original position the relative retardation mentioned is compensated by slightly tilting the glass plate P, (so as to interpose a greater thick- ness of glass in the path of the lion-retarded portion of the beam of iight).The plate P, is tilted by raising or lowering the lever Kh by means of the screw M , with a graduated head. The number of divisions on the screw head, through which it has to be turned to restore the bands in the upper part of the field to their zero position, constitutes the interferometer reading, and is taken its a measure of the difference field will be displaced to the right or left.PIG. 3.62 ABSTRACTS OF CHEMICAL PAPERS in refractive index between the liquids (or gases) in the two chambers. According to the length of the chambers used (0.5 to 5 cm.), the instrument is from 7 to 60 times as sensitive as the immersion refraatometer, but its range of measurement is correspondingly less.The authors first applied the instrument to the determination of wort gravities. The worts were diluted tenfold, and compared with distilled water. The instrument was first calibrated by determining the readings for a series of worts whose gravity had been previously determined pyknometrically. By plotting the readings against the true gravities a straight-line graph was obtained, which was then used for finding the gravities of a number of other worts from their interferometer readings.The dilution of the worts was carried out with the greatest care. The results for 13 worts showed errors ranging from 0 to 1 *5 in the fourth place of decimals in the gravity (e.g., 1.03345 found instead of 1.0336 as determined pyknometrically).Similar experiments in which the worts were diluted only fourfold (and compared with water) gave results in which the average error was 5 to 6 in the fifth place of decimals of the gravity (e.g., 1-0336 found instead of 1.03365), and the maximum error between two and three times as great. Owing to the extreme care required to avoid intro- dqcing errors in diluting the wort, the determination of gravities by this method could not be regarded as more convenient than the pyknometric method.The authors therefore made observations with undiluted worts, and used for comparison a solution of monopotassium phosphate containing 10 grms. per 100 C.C. The results for 35 different worts gave a straight-line graph, with an average discrepancy in individual cases corresponding to 5 in the fifth place of decimals of the gravity (e.g., 1003375 instead of 1*0337), and a maximum discrepancy of nearly three times as much.With some practice, the determination of gravity in this way occupied five minutes at the outside, but with dark worts the readings were difficult and some- what uncertain, and in such cases it would probably be preferable to operate with the wort diluted fourfold and compare it with water.Marc has previously shown how the interferometer can be used for the determina- tion of dissolved colloids in liquids by taking readings before and after: treatment of the liquid with a substance such as barium sulphate which &dsorbs the colloids. I n some experiments of this kind with worts and beers the authors employed purified blood charcoal instead of barium sulphate, and used only a relatively small amount so as to avoid the removal of high-molecular crystalloids. Although by this method only 8, fraction of the total colloids present was removed, it was thought that useful comparative results would bt! obtained by adhering to a standard procedure.The procedure finally adopted was to treat 50 C.C.of wort with 0.25 grm. of charcoal for half an hour, with frequent agitation, and then filter. The difference between the interferometer readings before and after this treatment is designated the “ colloid- number.” If was found to vary considerably with different worts, but the data, obtained are not yet sufficiently numerous to warrant general conclusions.For the examination of waters the interferometer is much more useful than the refrwtometer, owing to its greater sensitiveness, and its possibilities in this direction have been pointed out by Marc and Sack (Kolloidchemiscie Beihefte, 1914,5). The interferometer reading of a water (compared with distilled wat.er) gives an indicationAPPARATUS, ETC. 63 of the amount of total solids present.The ratio of total hardness to permanent hardness may also be estimated by taking readings of the water in its origina1 con- dition and after it has been boiled, filtered, and diluted with distilled water to the original volume. The readings of different waters are not strictly comparable, but for waters of similar composition they furnish a very exact measure of the total solids.A comparison of the amounts of dissolved colloids in different waters may be made by the method of adsorption described above for worts; the inorganic salts present are not adsorbed b;y the charcoal under the conditions indicated. The pres- ence of colloidal matters in water in which barley has been steeped is easily shorn in this way, and it would probably be possible to use the interferometer as a check on the efficiency of the steeping and washing process.The extraordinary sensitiveness of the interferometer makes paesible the deter- mination of the concentration of very dilute solutions of any single substance when once the instrument has been calibrated for that particular substance. For example, the authors obtained a reading of 2 with a solution of 0.1 C.C.of & sodium hydroxide in 100 C.C. of water, using the 1 cm. chambers. The concentration of dilute solutions of furfural can also be determined, and the possibility of applying the interferometer to the estimation of pentosans is being investigated. The Zeiss instrument is provided with special chambers, 10 em. long, for obser- vations with gases, and it can be used for the estimation of any one const.ituent of gas mixtures.For this purpose it is necessary to fill one chamber with the gas under test, and the other with the same gas from which the constituent t.0 be determined has been removed. Thus, in the estimation of carbon dioxide in flue gas. the gas is dried and passed first fhrough one of the two chambers, then through a t.ube contain- ing soda-lime to absorb carbon dioxide, and then through the other chamber. After the gas has been flowing for a sufficient time 00 sweep out the chambers, the flow is stopped and a reading is i;aken. The results obtained by this method are slightly high, if the inst'rument has been calibrated by means of mixtures of carbon dioxide and air. This is because the soda-lime absorbs sulphur dioxide as well as carbon dioxide from the flue gas. The slight error may be avoided by removing the sulphur dioxide, by means of lead peroxide, before the gas is passed through the first chamber. The carbon dioxide can then be determined with a possible error of less than 0.1 per cent.