since under these circumstances the poppy absorbs nearly the whole of the light which falls upon it. The colour of white objects is due to the fact, that they absorb none of the rays; whilst the absence of colour of black objects, arises from their power of absorbing rays of every colour.

Dichromatism.

305. There are certain substances, which appear to be of one colour when white light is passed through a stratum of moderate thickness, but which appear to be of a different colour when the thickness is large. For example, light transmitted through glass coloured with cobalt, appears to be blue when the thickness is small, purple when the thickness is greater, and deep red when the thickness is large. If on the other hand, a solution of chlorophyll (that is, the green colouring matter of leaves), in alcohol be employed, the transmitted light appears to be of a bright emerald green when the thickness is small, and red when the thickness is large. This phenomenon is called dichromatism.

306. Dichromatism was first explained by Sir J. Herschel in the following manner. Cobalt glass is obviously opaque to yellow and green light; and although it is capable of transmitting blue and red light, it absorbs blue light to a greater extent than red. Let us consider a thin stratum of the substance, whose thickness is da; let I, I+ dI be the intensities of the light which enters and emerges from the stratum, and let q represent the proportion of light of unit intensity, which is absorbed by a stratum of unit thickness. Then qIda is the proportion of light of intensity I, which is absorbed by a stratum of thickness da; whence

where A is the intensity of the light incident upon the substance. This is the value of the intensity of the light after it has passed through a stratum of thickness . The quantity q is called the coefficient of absorption of the substance, and is dependent upon the colour of the incident light.

x.

307. Let us now denote the values of the quantities for blue and red light by the suffixes b and r. Then for cobalt glass, q may

be regarded as infinite for all colours but blue and red, but is finite for these colours, hence

Since the blue rays are entirely absorbed when the thickness is considerable, it follows that qu>q. Since the glass appears to be blue when the thickness is small, it follows that A > A,. This must be interpreted to mean, that out of the rays of different colours which fall upon the glass, a far greater number of blue rays are capable of being transmitted than of red. In other words, cobalt glass transmits a very large portion of blue extremity of the spectrum and very little of the red, but the coefficient of absorption of the blue rays is greater than that of the red.

308. In the case of chlorophyll, the coefficient of absorption is very large for all rays but green and red, and is greater for green than red; but this substance is capable of transmitting a larger portion of the green part of the spectrum, than of the red.

Anomalous Dispersion.

309. We have already pointed out, that the order of the colours in the solar spectrum is violet, indigo, blue, green, yellow, orange, red; and that the violet is the most refracted, and the red is the least. There are however certain substances, in which the order of the colours in the spectrum produced by refracting sunlight through them, is different from that produced by glass, and the majority of transparent media. The dispersion produced by such substances is called anomalous dispersion.

310. Anomalous dispersion appears to have been first observed by Fox Talbot' about 1840, but the discovery excited no attention. It was next observed by Leroux' in 1862, who found that vapour of iodine refracted red light more powerfully than violet. This substance absorbs all colours except red and violet, and it was observed that the order of the colours in the spectrum, beginning at the top, was red, then an absorption band, and then violet. The indices of refraction, as determined by Hurion3, are

See Proc. R. S. E. 1870; and Tait, Art. Light, Encycl. Brit.

2 C. R. 1862; and Phil. Mag. Sept. 1862.

Journ. de Phys. 1st Series, Vol. vii. p. 181.

ANOMALOUS DISPERSION.

297

311. Anomalous dispersion is most strongly marked in solutions of the aniline dyes in alcohol. Christiansen' discovered in 1870, that it was produced by fuchsine, which is one of the rose aniline dyes; for when sunlight was passed through a prism containing a solution of this substance, it was found that the order of the colours was indigo, green, red and yellow, the indigo being The amount of anomalous dispersion increases

the least deviated.

with the concentration of the solution, as is shown in the following table of the indices of refraction.

From this table, we see that the line D is more refracted than any of the others, and that the violet is less refracted than the red.

312. Kundt afterwards showed, that blue, violet and green aniline, indigo, indigo-carmine, cyanine, carmine, permanganate of potash, chlorophyll and a variety of other substances exhibited anomalous dispersion. The following table gives some of the indices of refraction found by him.

1 Pogg. Ann. Vol. CXLI. p. 479; and Phil. Mag. March, 1871, p. 244.

2 Ibid. Vol. CXLIII. p. 250. See also Wiedermann, Ber. der Sächs. Gesell. math.phys. Cl. Vol. 1. 872; G. Lundquist, Nova acta reg. Soc. Sc. Upsaliensis [3] Vol. IX. Part II. (1874); Jour. de Physique, Vol. III. p. 352 (1874).

3 Pogg. Ann. Vols. CXLII. p. 163; CXLII. pp. 149, 259; CXLIV. p. 128; CXLV. p. 164.

4 Pogg. Ann. Vol. cxLv. p. 67.

The preceding table gives a general idea of the condition of the spectrum, and shows that for cyanine the line E is the least refracted. In the case of cyanine and fuchsine, the lower end of the spectrum is blue, then comes an absorption band, and afterwards red and orange; so that the blue is least refracted, the green and some of the yellow are absorbed, and the orange is the most refracted. In the spectrum produced by permanganate of potash, there is a slight amount of anomalous dispersion between D and G; for Kundt found, that the indices of refraction for green and blue were 1.3452 and 13420 respectively, showing that the blue is less refracted, than the green in the neighbourhood of D. In the region between D and G there are also several absorption bands.

313. By means of his experiments, Kundt deduced the following law :

On the lower or less refrangible side of an absorption band, the refractive index is abnormally increased; whilst on the upper or more refrangible side, it is abnormally diminished.

In order to clearly understand this law, let us revert to the spectrum produced by fuchsine. In this substance the absorption is very strong between D and F, that is in the green portion of the spectrum; and on looking at the table, we see that the red and orange rays, which lie below the green in the spectrum produced by a glass prism, lie above it in the case of fuchsine; whilst the violet rays lie below the green. The refrangibility of the red and orange rays is therefore abnormally increased, whilst that of the violet is abnormally diminished.

Selective Reflection.

314. We have already pointed out, that the colours of natural bodies arise from the fact that they absorb certain kinds of light; there is however another class of substances, which strongly reflect light of certain colours, whilst they very slightly reflect light of other colours. The phenomenon exhibited by these substances is called selective reflection.

315. Selective reflection appears to have been first discovered

SELECTIVE REFLECTION.

299

by Haidinger'. It was subsequently studied by Stokes'; and the experiments of Kundt, which have already been referred to3, show, that it is exhibited by most substances which produce anomalous dispersion. In fact absorption, anomalous dispersion and selective reflection are so closely connected together, that they must be regarded as different effects of the same cause, and consequently ought to be capable of being explained by the same theoretical considerations.

316. The properties of substances, which exhibit selective reflection, may be classified under the following three laws:

I. Those rays which are most strongly reflected, when light is incident upon the substance, are most strongly absorbed, when light is transmitted through the substance.

II. When the incident light is plane polarized in any azimuth, the reflected light exhibits decided traces of elliptic polarization.

III. When sunlight is reflected, and the reflected light is viewed through a Nicol's prism, whose principal section is parallel to the plane of incidence, the colour of the reflected light is different from what it is, when viewed by the naked eye.

Since the reflected light is elliptically polarized, it follows that selective reflection is accompanied by a change of phase of one or both the components of the incident light.

317. The properties of substances, which exhibit selective reflection, resemble those of metals, as will be explained in the Chapter on Metallic Reflection. For in the first place, metals strongly absorb light, and powerfully reflect it; and in the second place light reflected by a metallic surface is always elliptically polarized, unless the plane of polarization of the incident light is parallel or perpendicular to the plane of incidence. The optical properties of these substances appear to occupy a position, intermediate between ordinary transparent media and metals; and on this account, selective reflection is sometimes called quasi-metallic reflection.

1 Ueber den Zusammenhang der Körperfarben, oder des farbig durchgelassenen, und der Oberflächenfarben, oder des zurückgeworfenen Lichtes gewisser Körper. Proc. Math. and Phys. Class of the Acad. of Sciences at Vienna 1852, and the papers there cited.

2 Phil. Mag. (4) vi. p. 393.

3 Ante, p. 297.