the end of this rod. The angle of the halfshade angle is indicated on an engraved dial, 3 cm in diameter, which is in constant view of the operator.

Another important improvement [2] has been made in the location. of the milled heads which move the quartz wedges. Heretofore polariscope builders have mounted these heads on the ends of vertical rods, thereby forcing the hand and arm of the observer into a cramped and unnatural position while making the setting. It will be observed that the milled heads are at right angles to the customary position, thereby overcoming this objection. The wedges can be instantly clamped rigid at any position of the scale by means of the clamps, C and N, figure 18, B.

Both the right- and left-rotating wedges are of unusual length, and in the present models read from 15° S to 105° S. Their scales are the regulation type ordinarily used on Fric saccharimeters. With two exceptions, the scale now used by manufacturers generally is of metal and reads by reflected light. Owing to the shortness of the wedge the reading is necessarily made by means of a magnifying telescope. A black line thus usually separates the vernier and the scale proper or may appear with usage. Owing to the fact that the eye must bridge this line and that the edges of the rulings are not sharp, interpolation of the vernier to hundredths is impossible, and in many cases it is very easy to make an error of 0.1° S. The Fric scale, which has been in use for some years, is made with the lines etched on glass, which permits of very sharp rulings. What is of even greater advantage, it is read by transmitted light, the black dividing line between the vernier and the scale being thus eliminated. Not only is this scale much easier on the eye of the observer, but it also permits of reading accurately to 0.01° S. It is illuminated by waste light collected by a 45° mirror, located in front of the polarizing system. Thus it is not necessary to have any extraneous light in the room, as all the light needed enters the instrument through the collecting lens in the end of the tube. In investigations utilizing the saccharimeter, where temperature corrections are to be made, it is necessary to know accurately the temperature of the quartz wedges. Polariscope builders do not generally make provision for this. A thermometer (10° to 40° C, in one-fifth degree) with a horizontal scale and with its bulb. between the quartz wedges has accordingly been mounted in a brass case on top of the metal housing containing the compensator. For all ordinary sugar testing, where the temperature of the room changes slowly, the reading of the thermometer is practically the temperature of the room. The observer is thus able to take the temperature of the wedges with the same facility that he reads the scale on his instrument, since the thermometer scale is in a similar position and is illuminated by the same light source.

In general, when a tube of liquid is placed in a polariscope, there is present a certain amount of haziness which seems to overlie the field of view. It is due to depolarized light and interferes with the focusing and the accurate matching of the halves of the field. By proper diaphragming, this has been reduced to such a low minimum that it may be said to be practically eliminated. When a tube is placed in the instrument, the observer sees a clear, sharply defined, circular field with no extraneous light. The observing telescope, as well as the two reading telescopes for the scale, has screw adjustments

which permit of accurate and rapid focusing. Still further improvement has resulted from making the base of the instruments exceptionally heavy and mounting it on rubber tips. The resulting inertia of the instrument and friction on its supporting bench prevent accidental shifting with reference to the light source.

A twofold object has been kept constantly in view in designing this instrument: (1) to produce a saccharimeter of great flexibility for regular commercial testing and to correct the defects of the ordinary instrument; (2) to provide the chemist with a white-light instrument

FIGURE 18.-A, Bates type adjustable-sensitivity saccharimeter with NBS polariscope lamp and bichromate filter.

suitable for research work. The result has been the production of an instrument of a greater range of adaptability than any saccharimeter heretofore built. In measuring rotations with the greatest possible accuracy, or when it is desired to make the settings with the least possible strain on the eye, the observer has only to change the halfshade angle until he has just sufficient light to bring the two halves of the field to the same intensity, without undue eyestrain. He then has for his eye an instrument so adjusted as to give the maximum sensitivity for making the setting, no matter what the character of the substance whose rotation is being measured.

2. BASIS OF SACCHARIMETER CALIBRATION

The specifications defining the 100° point of saccharimeters have been changed many times since Ventzke [3] defined the normal sugar solution in 1842.

Two markedly different saccharimeter scales are in use today, each purporting to give the correct percentage of sucrose in a sample. They are the French Scale and the International Sugar Scale. The former is used largely in France and the French Colonies; the latter is used generally throughout the rest of the world. The International Sugar Scale is the official scale of the International Commission for

FIGURE 18 (Continued)-B, front view of saccharimeter shown in figure 18, A. Uniform Methods of Sugar Analysis and was adopted at the Eighth Session of the Commission in 1932 at Amsterdam. Obviously any saccharimeter scale is, by definition, a basis on which the standardization of the saccharimeter may be brought about. In any consideration of the saccharimeter scale it is essential that there be kept clearly in mind the differentiation between mere definition of the scale and the numerical values of the constants upon which its correctness depends. The confusion which has long existed with regard to saccharimeter scales may be correctly attributed to the use of incorrect values for these constants.

Probably no industrial commodity is so widely utilized in international trade as sucrose and its associated products. Therefore, it is of great importance that there be but one sugar scale, that is, one basis of standardization of saccharimeters in use throughout the world. Such a consummation must ultimately be brought about in view of the importance of the matter to international trade and especially because of the necessity of reporting the results of research. workers in unmistakable values. Until this can be done, and under present conditions, it is therefore of special importance, insofar as the inherent characteristics of the French and the International Sugar Scales will permit, that the two scales give results that are in agreement within the experimental error with which saccharimeter measurements can be made. It is fundamental that tests of a sample of sucrose should show an identical sucrose content regardless of the scale used.

The specific rotatory power or specific rotation [a] is, by definition [4], proportional to the rotation divided by the weight per volume, that is,

where [al is the specific rotation at 20° C measured with the D line of sodium, a the observed rotation, e the concentration expressed in grams per 100 ml of solution, and the length in decimeters.

Hence in the formulation of any saccharimeter scale based upon the above equation, some more or less arbitrary choice of value for either one or the other of these two quantities, a and c, must be made before the other can be fixed.

The French chose to let the 100° point of their scale be fixed by a definitely selected rotation, a=21°40', and to determine by experiment, based upon the law stated above, what weight of sugar (which was called the normal weight) was required to produce that rotation.

The Germans, on the other hand, following Ventzke, chose to select a normal weight, c, and determine by experiment what rotation, a, corresponded thereto and to let that rotation fix the 100° point of their scale.

It is clear that in establishing both scales actual physical measurements must be made, both of the weights and of the rotations. It therefore makes no difference in the final analysis whether the weight be considered fundamental and the rotation consequential or whether the rotation be considered fundamental and the weight consequential. In the abstract both lead to the same result, namely a relation or ratio between rotation and weight (per standard volume).

However, practical considerations, such as established usages, choice of values which determine the openness of the scale, effect of concentration on specific rotation, and other factors not involving the fundamental law itself, may operate to make one particular choice of scale preferred over another.

(a) THE FRENCH SUGAR SCALE

It is of interest to recall that the French Sugar Scale was originally based upon the rotation of 1 mm of quartz for sodium light to fix the 100° point. This constant was determined by Broch [5] to be 21°40' (21.667°) for sodium light. The usual procedure has been to lay off 65° on a circle and divide the 65° segment into 300 equal parts, taking the 100th division mark as the 100° point. Instruments were constructed on this basis, and when it was later found that the rotation for 1 mm of quartz was somewhat higher, the original value (21°40′) was retained in order to avoid confusion of scales. This procedure necessitated the abandonment of the original definition of the 100° point for the French Sugar Scale. The 100° point thus has always been and still is fixed by the absolute rotation, 21°40'. In this sense the French Sugar Scale has never been changed. However, the estimate of the amount of sugar required to produce this rotation, namely the value of the normal weight, has undergone many changes. Over 20 different values have been assigned to the normal weight, ranging from 16.0 to 16.51 g. In 1875 the value of Girard and de Luynes [6], 16.19 g, was adopted as the official weight and remained so for more than 20 years. In 1885 [7] Sidersky called attention to the error in the then official normal weight of 16.19 g, and was criticised for questioning the correctness of the official normal weight [8]. Nevertheless, Sidersky had the courage of his convictions and 11 years later, in 1896, his value was made the official value. This value (16.29 g) was based upon two considerations [8]: (1) A calculation of the normal weight, using Tollens' value, 66.5°, for the specific rotation of sucrose, whereby he obtained the value 16.295 g; and (2) the direct comparison of a Schmidt & Haensch and a Laurent saccharimeter by means of a quartz plate, which he read in both instruments. He assumed that the Schmidt & Haensch instrument was correct, since it was supposed to have been directly calibrated by means of pure sucrose, and calculated the weight which would have to be used with the Laurent instrument to give the correct percentage of sugar. He found for this value 16.30 g, which confirmed his calculated value, 16.295 g. It was not known at that time, as it is now, that the Schmidt & Haensch instruments of those days were in error by about 0.2 percent, the 100° point then corresponding to an absolute rotation of 34.68° instead of 34.620°, as at present. Had this error been known at that time, Sidersky's experiment would have yielded a result of 0.2 percent lower, or 16.27 g. Also, all subsequent determinations of the French normal weight have yielded results slightly lower than 16.29 g. Following the action of the IIme Congres International de Chimie Appliquée (Paris, 1896) in adopting 16.29 g as the official normal weight for the French Sugar Scale, a commission was appointed for the revision of the saccharimetric normal weight. This commission decided that further experimental work was required in order to establish the normal weight on a firm foundation. This work was undertaken by Mascart and Benard [9] at the Collége de France and also by Pellet [10] at the Sorbonne.

As a result of these experiments the correctness of the value 16.29 g was considered to be amply confirmed. However, it has been recently pointed out [11] that had the values actually found by these observers been used instead of the official value which it was considered they