the General Electric Co. as a source for photomicrographic work. The latter is a 30-ampere arc in a small glass bulb, the incandescent tungsten electrode serving as the light source. It is operated by a special high-reactance transformer operating from the 110-volt alternating-current line. Both of these types can, by the use of proper optical systems, be made to give uniform polariscopic fields without the use of the ground glass, thereby greatly increasing the intensity available. Such sources, however, find only limited uses and for very special purposes. The 100-watt lamp described above in detail is adequate for all ordinary work. 5. CERTIFICATION OF QUARTZ CONTROL PLATES Quartz control plates for use in checking saccharimeters will be accepted by this Bureau for standardization with reference to the sugar value. The conditions as to mounting, purity of quartz, correctness of fabrication, etc., are given on page 57. This Bureau reserves the right to reject any plate showing defects which may render it unreliable or otherwise unsatisfactory in service. Certificates are issued showing the optical rotation of the plate for pure monochromatic light of two wave lengths, λ=5461 A and 5892.5 A, as well as the sugar value of the plate at 20° C in International Sugar Degrees. When desired, an accompanying table of temperature corrections will be furnished covering the range 15° to 30° C, by means of which saccharimeter readings made at any temperature within that range may be corrected to the reading that would have been obtained had the readings been made at 20° Č. 6. REFERENCES [1] H. Soleil, Compt. rend. 20, 1085 (1845); 21, 426 (1845); 24, 973 (1847); 26, 163 (1848). [2] F. J. Bates, Bul. BS 4, 461 (1908) S86; 5, 193 (1908) $98. [3] K. Ventzke, Erdmann's J. prakt. Chem. 25, 65–84 (1842); 28, 111 (1843). [4] Jean Baptiste Biot, Mém. acad. sci. 15, 101 (1838). [5] O. J. Broch, Dove Rep. Phys. 7, 113 (1846). Ann. chim. phys. 34, 119 (1852). [6] A. Girard and V. de Luynes, Compt. rend. 80, 1354 (1875). [9] E. Mascart and H. Benard, Ann. chim. phys. 17, 125 (1899). [10] H. Pellet, Ann. chim. phys. 23, 289 (1901). [11] F. J. Bates and F. P. Phelps, J. Research NBS 17, 347 (1936) RP916. [12] O. Schönrock, Geiger's Handbuch Physik 19, 705 (1928). [13] E. Roux, Bul. assn. chim, sucr. dist. 55, 404 (1938). [14] Proc. Third Session, International Commission for Uniform Methods of Sugar Analysis, Paris, (1900). [15] A. Herzfeld, Z. Ver. deut. Zucker-Ind. 50, 826 (1900). [16] O. Schönrock, Z. Ver. deut. Zucker-Ind. 54, 521 (1904). [17] Proc. Seventh Session, International Commission for Uniform Methods of Sugar Analysis, New York, 1912. [18] F. J. Bates and R. F. Jackson, Bul. BS 13, 67 (1916) S268. [19] Proc. Eighth Session, International Commission for Uniform Methods of Sugar Analysis, Amsterdam, 1932, Int. Sugar J. 35, 17, 62 (1933). [20] E. Gumlich, Wiss. Abhandl. physik.-tech. Reichsanstalt 2 212 (1895). [21] T. M. Lowry, Phil. Trans. 212, 288 (1912-1913). [22] T. M. Lowry and W. R. C. Coode-Adams, Phil. Trans. 226, 391 (1926-27) [23] B. Tollens, Ber. deut. chem. Ges. 10, 1403 (1877). [24] Raffaello Nasini and Vittorio Villavecchia, Public di Lab. chim. delle gabelle, Rome, p. 47 (1891). Roma, R. Acc. Lincei Rend. 7, 285-290 (1891); Gazz. chim. ital. 22, 97-104 (1892). [25] Hans Landolt, Das optische Drehungsvermogen, p. 420 (Friedrick vieweg und Sohn, Braunschweig, Germany; 1898). [26] R. F. Jackson, Bul. BS 13, 633 (1916) S293. V. TEMPERATURE CORRECTIONS AND CONTROL 1. QUARTZ-WEDGE SACCHARIMETER The question of temperature corrections for polarimetric apparatus, as well as for the optically active substances, is a difficult one. The literature on this subject is extensive, but it has not always been a simple matter to select the proper correction owing to the different results secured by various investigators. For the ordinary polarimeter used for measuring absolute rotations no instrument correction is necessary. It may be used at any temperature if care is taken to allow for any zero-point shift that may occur. However, a correction is unavoidable for saccharimeters, in which a quartz wedge is used to neutralize the rotation of the substance being tested if temperatures other than the standard are used. The rotation for a plate of quartz in the neighborhood of 20° C is given by the following: aa%a2% 0.000143 (t-20). (27) The linear coefficient of expansion of quartz perpendicular to the optic axis is 0.000013, and of the scale is 0.000018 if it is of nickelin, and 0.000008 if it is of glass. Thus the total temperature coefficient [1] for the ordinary quartz-wedge saccharimeter is or 0.000143-0.000013+0.000018 0.000148 (metal scale) (28a) 0.000143-0.000013+0.000008 0.000138 (glass scale). (28b) If the scales were etched on the wedges, the scale coefficient would become zero. Since the effect of the expansion coefficient 0.000148 is to lower the reading of the scale with an increase of temperature, the apparent polarization of any substance is lower than it should be and the reading at 20° (S20) is given by the following: S20=St+S, 0.000148 (t—20). (29) When a quartz control plate is read in a saccharimeter, this effect is not completely compensated. Since the temperature coefficient of the plate is 0.000143, the reading (S) of the plate is then S20 SS, 0.000005 (t-20). (30) The correction given by this equation changes sign if the scale is of glass, and is so small at all times as to be negligible in ordinary polarizations. 2. SUCROSE The influence of temperature on the specific rotation of sucrose has been studied by numerous investigators, of whom Dubrunfaut [2] was the first to discover that the constant decreased with increase in temperature. Unfortunately, subsequent determinations of this variation have not shown satisfactory agreement. Schönrock [3] at the Reichsanstalt carried out an elaborate investigation and found that for the normal sugar solution (p=23.701) the temperature coefficient (8) was independent of the wave length of the light used, but that it decreased with increase in temperature as follows: 10° C, 0.000242; 20° C, 0.000184; 30° C, 0.000121. These data show the change to be practically linear over the range 10° to 30° C, and have been combined into the following equation [4, p. 7]: ♪=−0.000184+0.0000063 (t−20° C) 3. COMBINATION OF CORRECTIONS FOR QUARTZ-WEDGE SACCHARIMETER AND SUGARS (31) If a proper temperature correction is to be applied to a polarization, it is necessary to add algebraically all the corrections applicable to the conditions under which the polarization is being made, or to determine the correction experimentally. (a) SUCROSE In the ordinary testing of sucrose, the solution is made up to volume and read at the same temperature, which in general is not the standardization temperature. It is desirable, therefore, to know the variation with temperature in the saccharimeter reading of a near normal sucrose solution. Among the factors that require consideration are the changes in the quartz-wedge system due to changes in temperature, changes in the specific rotation of sucrose, in the volume of solution, in the volume of the flask, and in the length of the tube used. Some of these act in opposite directions and thus partially cancel or compensate for each other. By a summation of the best known values of the separate coefficients which enter into the correction factor, we obtain the value 0.0309 per degree centigrade for a normal sucrose solution (100° S). Owing to its importance, this over-all temperature coefficient has been measured experimentally by a number of investigators. Their results are given in table 7, based upon a normal solution of sucrose. TABLE 7.-Saccharimeter temperature coefficient for sucrose The average value obtained is 0.0303 for a normal sugar solution (100° S). The value varies slightly for different instruments, probably in part because of slightly different differential expansions of the scale and wedge mountings. This correction factor is predicated upon the instrument and all apparatus, flasks, tubes, etc., having been originally calibrated at 20° C but used at some other temperature between 20° and 30° C. It is important and necessary, for the proper application of this temperature correction, that the solution be made up to volume at the same temperature as that at which it is to be read in the saccharimeter and that the entire saccharimeter also be at that same temperature. Under these conditions the polarization in sugar degrees at 20° C (S20) of an approximately normal sucrose solution made up to volume and polarized at the same temperature (t° C) is given by the following: S20=St+St 0.0003 (t−20° C). (32) This equation is sufficiently accurate for all ordinary polarizations, irrespective of type of tube or scale. (b) APPLICATION OF TEMPERATURE CORRECTION BY MEANS OF A QUARTZ CONTROL PLATE This correction is most conveniently and satisfactorily applied by means of a standardized quartz control plate. The proper use of such a plate gives an over-all correction which includes not only the above temperature correction but also corrects any accidental or irregular variations, such as those due to minor residual temperature differences in the quartz wedges, etc. The procedure in using a quartz control plate is quite simple and is as follows: When making a series of polarizations of sugar solutions on the saccharimeter, also make a series of readings on the standard quartz control plate. Observe the temperature. From the table of sugar values for the standard plate find the sugar value corresponding to the observed temperature. The difference between the observed reading of the plate and the sugar value obtained from the table is applied as a temperature correction to all polarizations made during that series. This is the procedure which has long been recommended by this Bureau whenever making polarizations where a standard 20° C constant-temperature room is not available. Because of its convenience and accuracy this practice was introduced by the Bureau into the various Customs laboratories at a very early date. Figure 23 is a facsimile copy of a table showing the sugar values for various temperatures of a quartz control plate, which was issued in 1907 as a part of the certificate of test for that plate. Prior to that date similar tables were issued and are still being issued whenever requested in connection with the certificates for quartz control plates. (See this Circular, p. 553, Fee Schedule 423e; Circular C44, Polarimetry, 2d edition (1918), p. 111, Fee Schedule 44d; Circular C44, 1st edition (1914), p. 96, Fee Schedule 44d). These tables are based upon the measured value of the plate at 20° C and the saccharimeter temperature coefficient 0.0003, as defined above. They are calculated from the equation St S20 [1+0.0003 (t−20° C)] (33) and are valid, of course, only under the conditions stated above, namely, that the solution be made up and polarized at the same temperature. (c) CORRECTIONS FOR USE IN TROPICAL COUNTRIES Temperature correction by the method outlined above is frequently utilized and gives excellent results in tropical countries, and it is |