spoon is moved forward to the front edge of the flame, and the salt fusing and running to the point, volatilizes, and produces an intense yellow colour. Or the stem d may be provided with a small brass revolving collar, having several arms with holes, into which can be fitted platinum wires with beads of salt fused on. When one bead is consumed the next arm is brought round to the flame, and so on. Instead of common salt calcined soda may be used, but this, whilst volatilizing more slowly, has less illuminating power.

In making observations, the instrument should be set up at a distance of an inch or so from the flame, and a black screen placed behind the latter, so as to shut off extraneous light. The room should be darkened too, at least partially, as the observations in general are more satisfactory the darker the place is. The zeropoint must first be determined. For this purpose the tube is put in its place either empty or filled with water, and the analyzer Fig. 23. set to the position of greatest darkness. If the circular field of vision is at all large, there will not be perfect obscuration over the whole, but merely a vertical dark band, getting lighter towards the sides, as in Fig. 23, and this band must be brought, by backward and forward motions of the analyzer, as nearly as possible into the middle of the field. Repeating the adjustment several times, and taking the mean of the readings on the disc, we get the true zero-point of the instrument. To make the zero of the scale agree, at least approximately, therewith, we set the index against the mark, loosen the clamp e (Fig. 21), and rotate the polarizer until the dark band appears in the middle. Usually this correction is made by the instrument-maker himself. As before stated (§§ 4 and 5), there are two positions, 180° apart, at which the analyzing prism gives maximum darkness, and the zero-point of the second, which must lie somewhere about 180° on the scale, should similarly be accurately determined by a few observations.

If the tube, filled with active liquid, be now laid in the instrument, the analyzer having been previously set to zero, the field of vision will again appear bright, and in order to restore the black band it will be necessary to rotate the analyzer to the right in the case of a dextro-rotatory substance—that is, in the direction of the hands of a watch-and in the opposite direction, to the left, when the substance is lævo-rotatory.

The same order of phenomena will be observed if

1500

Fig. 24.

300

1

3

we start from a position 180° from the first. In case, however, we do not know beforehand the direction of rotation peculiar to the substance, the following considerations must be borne in mind :-Suppose that the plane of polarization, having originally the direction AB (Fig. 24), is. diverted to CD (at an angle of 30° from A B) by passing through an optically-active medium, the dark band will then appear when the index stands at 30° or 210°, and the substance may either be dextro-rotatory, 30°, or lævo-rotatory, 360° - 210° = 150°. In most cases the side on which the smaller amount of deviation occurs is the true direction of the rotation. We cannot, indeed, be so guided when the smaller of the two angles exceeds 90°, which, however, only happens with substances having very high rotatory power, or in using tubes more than 2 decimetres in length. In these cases, the question can easily be decided by examining the liquid in a tube only half the length of that originally used, or by diluting the solution to half its strength. The deviation should then be only half of the original amount, and it is thus easy to discover which is the direction in which this holds. If, for example, darkness now occurs when the analyzer lies in the direction EF that is, at 15° and 195°-the decrease shows the rotation to be dextro-rotatory, since the position measured for lævo-rotatory power would indicate an increase of rotation from 150° to 165°, which is absurd.

B

It is well to take observations on the filled tube at both positions, 180° degrees apart, as, owing to defective construction, the Nicol prisms may be somewhat eccentric, causing the observed angles to differ appreciably from each other. Any such source of error is accordingly eliminated by taking the mean of the two readings. Mitscherlich's instruments frequently have two opposite index arms, but as with fairly good graduation the difference between their readings does not amount toth degree, it is generally sufficient to use The differences between the angles in successive observations usually amount to several tenths of a degree, and the accuracy of the final result will, of course, be greater in proportion as the observations are more numerous. As an example we may give the following

one.

results:

Instead of the sodium flame, monochromatic light, obtained by placing a red glass slide in front of an ordinary gas lamp, was used by Biot and Mitscherlich. But in this way observation is rendered much more difficult through defect of brilliancy, besides which, the red light so produced does not correspond to any one distinct ray (see p. 46).

§ 47. When white day or lamp-light is used with Mitscherlich's instrument, the angle of rotation observed is that for mean yellow rays, and, as already stated (§ 18), is denoted by aj. For this purpose the most suitable form is a gas or petroleum lamp, fitted outside its glass chimney with a metal screen coated inside with white porcelain, and having a side opening (see Fig. 25).

Fig. 25.

When with an empty tube placed in the instrument the analyzer is set to zero there appears, exactly as with the sodium flame, a dark band with fainter margins, which, as before, must be brought into the middle of the field of vision. If the active liquid is now placed in the tube, the different coloured rays composing this white light will experience different degrees of rotation, so that we shall have the phenomenon of rotatory dispersion. And here we may have two cases:

1. When the liquid possesses feeble rotatory power and the dispersion is therefore trifling, by turning the analyzer the dark band may be made to reappear with a border of blue on one side and red on the other. Now, if the blue border is to the left of the observer and the red to his right the substance is dextro-rotatory; if vice versâ, it is lævo-rotatory;

and this independently altogether of whether the proper position has been found by turning the analyzer to the right or to the left. It is to be observed, however, that when a Mitscherlich instrument is provided with any kind of Galilean telescope, the foregoing conditions are reversed. The position of the dark band indicates the point of extinction of the yellow rays.

2. When, on the other hand, the rotatory power of the active liquid is high, the dark band appears broad and undefined, or else cannot be brought back by any movement of the analyzer at all. By turning the latter we get merely a succession of colours, produced by the analyzer extinguishing, according to the position of its principal section, certain of the unequally-rotated coloured rays, and allowing the rest to pass on with different intensities, thus producing a succession of colour-mixtures. With solutions in which the angle of rotation for any ray is less than 90°, the sequence of coloured tints, when the analyzer is turned from the initial zero-point, is as follows:

As a point of reference for the analyzer, that position is chosen where the transition from blue to red stands exactly in the middle of the field of vision. This being the point of extinction of the yellow rays, the observed angles will be a¡.

With substances of still higher rotatory and dispersive powers, intermediate tints make their appearance; for example, between the blue and the red a reddish-violet, which, on the least touch of the analyzer, passes into one or the other. This is known as the sensitive or transition tint, and appears when the position of the analyzing Nicol is exactly such as to bar the passage of the mean yellow rays. This position also gives the angle aj.

For the relation between the angles of rotation a; and a, see § 18. Observations taken with white light cannot be made so exactly as those with the sodium flame; the former light is therefore only employed when the latter is not available.

$48. Mitscherlich's larger Instrument for Observations at Constant Temperature. In exact researches it is requisite, as already stated (§ 22), that the temperature should not only be known, but be constant during the period of observation. This can only be effected by surrounding the tube with water. Moreover, in examining substances of feeble rotatory power, it is necessary to employ tubes of considerable length, sometimes a whole metre long, to obtain rotation-angles of sufficient magnitude. Fig. 26 represents an instrument fulfilling these conditions, constructed at the works of Dr. Meyerstein, of Fig. 26.

Göttingen, and in use in the chemical laboratory of the Polytechnic School at Aachen.

Starting from the end next the light, the instrument consists of the following parts, resting loosely upon a frame formed of two strong iron bars QQ:

1. A fixed tube A, containing the polarizing Nicol, a convex lens of long focus, and a diaphragm with a square aperture of 5 millimetres side. Affixed to the same support as the tube is a circular dark screen, a, to shut off extraneous light.

2. A glass bottle with parallel walls B,1 filled with bichromate

1 May be obtained of Dr. J. G. Hofmann, 29, Rue Bertrand, Paris.