will serve as a saccharimeter, if the graduated arc be replaced by a scale on which the corresponding sugar percentages can be read off directly. Instruments, however, on a different optical principle have been constructed expressly for determining solutions of sugar. Of these so-called saccharimeters, now so extensively used in trade, the most important are the following:

(a.) The Soleil-Ventzke-Scheibler Saccharimeter.

§ 77. This very ingenious instrument was originally devised in the year 1848, by the Paris optician Soleil,1 and more recently improved by Soleil and Duboscq.2 In Germany, various alterations were made in the instrument by Ventzke, who introduced a different scale, and also by Scheibler, who devised important improvements in the mechanical arrangements.

4

The optical principle of the instrument is based upon the following facts discovered by Biot :-1. That when a polarized ray is transmitted through several media possessing rotatory power in different directions, their separate activities may become either partially or wholly neutralized, according to the lengths of the media. 2. That the rotatory dispersion of cane-sugar is the same as that of quartz. White day or lamp-light is used with the instruThe optical parts are as shown in Fig. 55, the course of the

ment.

light being from right to left. Starting from the right-hand side, we have:

A. The so-called regulator for restoring the sensitive tint, consisting of a rotating Nicol's prism a, and a plate of quartz

1 Soleil: Comptes Rend., 24, 973; 26, 162.

2 Soleil and Duboscq: Idem. 31, 248.

3 Ventzke: Erdm., Journ. für prakt. Chem. 25, 84; 28, 111.

4 Scheibler: Zeitsch. des Vereins für Rübenzuckerindustrie, 1870, 609.

b (dextro-rotatory or lævo-rotatory), ground perpendicu-
larly to the axis.

B. The polarizer, for which an achromatized calc-spar prism, a
Sénarmont's prism, or double quartz prism can be used. It
is arranged with its principal section vertical. The polar-
ized extraordinary ray is allowed to proceed along the axis,
but the sidewards-refracted ordinary ray is intercepted by
a diaphragm. The face of the prism towards A is ground
convex, so that light may emerge approximately parallel.
C. The bi-quartz, composed of two plates, the one of dextro-
rotatory, the other of lævo-rotatory quartz, fitted accurately
together. They may be either 3.75 millimetres or 7.5
millimetres thick, and are fixed in a brass case, so that the
line of junction remains vertical. (A front view is shown
above C.)

D. The experimental tube.

E. The so-called rotation-compensator. This consists of a plate of quartz c, which may be dextro-rotatory, in which case the other two plates, d, must be made of lævo-rotatory quartz. The latter are ground to a wedge shape, and are made to slide over one another, so that their combined thickness may be made either equal to that of c, or greater or less, the distance moved to effect any particular adjustment being shown by an attached scale.

The plate c may be made of left-handed quartz, but in that case the wedges must be right-handed.

F. The analyzer, which may consist of an achromatized calc-spar prism. Its principal section must be arranged parallel to that of the polarizer B, in case the thickness of the bi-quartz of the instrument, C, is 3.75 millimetres, and perpendicular thereto when the thickness is 7.5 millimetres.

G. A small Galilean telescope, consisting of objective e and eyepiece f. The latter is to be adjusted so that the line of junction of the plates of the bi-quartz C is sharply defined

§ 78. In order clearly to understand the action of the several parts, let us suppose first that we are merely dealing with the polarizer B and analyzer F, and that these are arranged with principal sections parallel, and the field at its maximum of illumination. Let the active bi-quartz C be now introduced between B and F; the

white light coming from B will thus be rotated, and suffer decomposition into its component coloured rays. Now of the emergent rays, those whose plane of polarization is at right angles to that of the analyzer will not be transmitted; and should these be the yellow rays, the remainder will, in transmission, combine to a pale lilac mixed tint which, with the slightest alteration of the plane of polarization, passes either into pure red or pure blue. This intermediate colour has been already referred to in § 47 as the sensitive or transition tint. With the polarizer and analyzer undisturbed in their parallel position, the sensitive tint will appear when the bi-quartz C has the thickness requisite to rotate the yellow rays exactly 90° to the right or left. This requires a thickness of 3.75 millimetres, since, according to Biot, a thickness of 1 millimetre of quartz rotates mean yellow rays through an angle of 24°, whence we get the proportion, 24: 1 = 90: 3·75. If, on the other hand, the corresponding planes of polarizer and analyzer were set at right angles to each other, a rotation through 180° would be required to eliminate the yellow rays, and the bi-quartz must then have a thickness of 7.5 millimetres. In one half of the bi-quartz the sensitive tint is produced by dextrorotation, in the other by lavo-rotation, and as the thickness of the two sides is equal, the tints produced will be the same.

Let the compensator E now be put in its place, the quartz wedges being so adjusted that their combined thickness is exactly equal to that of the quartz plate c. As the rotatory powers of c and d act in opposite directions, they neutralize each other, and the sensitive tint still occupies the field of vision. This position of the wedges corresponds with the zero-point of the scale. If, however, we shift the relative position of the wedges, the transmitted coloured rays will suffer rotation, and the analyzer will eliminate those of which the planes of polarization are perpendicular to its own. The other rays which pass on combine to produce a new chromatic mixture, which will, necessarily, be unlike for the rays transmitted by the respective halves of the bi-quartz C. Thus, if the compensator be so adjusted that its action is dextro-gyrate, the rays contributed by the dextrorotatory half of the bi-quartz will undergo an increase, and those coming from the lævo-rotatory half a decrease, in the amount of their rotation. Seen through the analyzer, the two halves will thus appear differently coloured-green and blue predominating in one, red and orange light in the other; and these tints will change places when the action of the compensator is lævo-gyrate.

Lastly, having again set the compensator to zero and reproduced the sensitive tint, let a tube filled with an optically-active solution be introduced. A splitting of the field of vision into two different coloured halves will once more occur. By sliding the quartz wedges, d, so as to produce rotation opposite to that of the solution, a position may be found where the action of the latter is annulled, and this will be indicated when the halves of the field of vision again exhibit uniformly the sensitive tint. The rotatory power of the substance introduced can thus be measured by the amount of change in the combined thickness of the quartz wedges, d, as indicated by the amount of adjustment required to bring into view the sensitive tint.

For the proper action of the compensator, it is, however, requisite that the active solution should have the same dispersive power as quartz, otherwise the effects of dispersion due to the solution will not be exactly neutralized by the opposite dispersion of the quartz. As before stated, § 18, this is the case with cane-sugar; but there are many substances which do not fulfil this condition, as, for instance, cholesterin, and in such cases, at least with strong rotations, the compensator can no longer be adjusted so as to restore a perfectly uniform colour to the field of vision. With Soleil's instrument, exact data are therefore only attainable in the case of substances whose rotatory dispersion does not materially differ from that of quartz.

The sensitive tint makes its appearance as above described when ordinary white light is used. When, on the contrary, lamp-light is employed, which contains the coloured rays in somewhat different proportions, there appears instead not the blue-violet, but a reddish tint, and a similar alteration may occur when the active solution is itself coloured. In this way arises the need for yet another addition to the instrument, namely, the regulator A, by which we can still produce the sensitive tint. It consists of a rotating Nicol prism, a, and a quartz plate, b, both placed in front of the polarizer B. When light which has been polarized by the Nicol falls on the quartz plate, it undergoes rotatory dispersion, and according to the position of the Nicol in reference to the polarizer, B, so will certain rays suffer total extinction by the polarizer or be transmitted by it with diminished intensity. In this way we are able to weaken the red and yellow rays of lamp-light, and so admit to the bi-quartz, C, a selected mixture of rays, which, in consequence of the rotation there experienced, reproduces the sensitive tint.

As

the polarizer, B, brings the transmitted rays into one plane of polarization, the two halves of the bi-quartz will appear coloured alike, the tint being variable at will by turning the Nicol a. A similar process is adopted in dealing with coloured solutions. The colour of the latter should, however, never be more than faint, otherwise too much light will be lost by absorption.

The two portions forming the regulator can be placed at the eye-end instead of the light-end if preferred, as is done in instruments of French make.

$ 79. The external form of Soleil's saccharimeter with Scheibler's improvements is shown in Fig. 56, which represents the instrument as supplied by Berlin opticians.1

On a brass stand rests, with horizontally rotating motion, a blackened metal trough, hh, in which the experimental tube R is

laid, the upper half of the trough serving as a cover, which can be shut down during observations so as to exclude extraneous light. The end of the trough next the light-in the Fig. the right-hand end-is connected with a brass tube enclosing at one end the bi-quartz D, at the other the polarizer C. Into this tube fits the rotatory tube, A B, containing the regulator with its Nicol at A, and the quartz plate at B. The rotation of the tube is effected by wheel and pinion movement, worked by means of a rod with the milled-head L.

The ocular part of the instrument contains at G the quartz plate of the compensator, and at EF the two quartz wedges. Each of the wedges is cemented to a similar wedge of glass, so as to form

1 May be obtained of Schmidt and Haensch, Stallschreiberstrasse 4, Berlin; Dr. Steeg and Reuter, Homburg v. d. Höhe.