105.]

DECOMPOSITION OF LIGHT BY ABSORPTION.

179

was then examined by means of a good prism. The light which passes through the thinner strata yields a spectrum generally differing but little from that of daylight; but that which has traversed greater depths of liquid exhibits a rapid disappearance of certain portions of the rays, whilst other rays are but little affected. Fig. 78, 1, shows the spectrum obtained by transmitting a beam of daylight through a dilute solution of a salt of cobalt, which appears to be of a delicate rose colour to the unaided eye. The same salt in more concentrated solutions appears to be of a rich blue, and exhibits a spectrum shown at fig. 78, 2, which represents the appearance of the spectrum furnished by a strong solution of cobaltous chloride in alcohol. The letters correspond to those of Fraunhofer's lines (106), the right-hand side of the figures indicating the red end of the spectrum; the lower part of the figures showing the effect of the thinnest strata of liquid.

It was ascertained from an extensive series of observations made in this manner, that when the salts formed by the union of a coloured base with different colourless acids were examined, the compounds of the same base nearly always exhibited a similar absorbent action upon the spectrum. Even in dichromic media, or solutions which, under certain circumstances, appear to the unaided eye to transmit light of one tint, and, under certain other circumstances, to transmit light of a different tint, this law generally holds good. An exemplification of this fact is seen in the case of the chromic salts, some of which exhibit a green colour when in solution, others a red or purple hue. Now all these salts furnish a spectrum, the general form of which is shown in fig. 79, in which the indigo

and the green rays are soon cut off, whilst the red and bluish-green rays are comparatively little affected. By some salts, such as chromic acetate, the green rays are absorbed much more rapidly than the red, and hence these solutions have, even in thin layers, a red colour: others, as the chromic chloride, are green when their solutions are seen in thin layers, but look red or purple when viewed in considerable mass by transmitted light.

Some salts, even though their solutions have but little colour, furnish very characteristic spectra. This is particularly the case with solutions of didymium, which are of a feeble rose colour, but they exhibit two very black lines, one in the yellow, the other in the green. These lines are visible in the spectrun even when the solution is very dilute, and they may be employed to indicate the presence of small quantities of didymium in solutions of lanthanum and ceriun, in which no such lines occur.

In artificial flames it is very generally the case that certain

180

DISPERSIVE POWER-SENSITIVENESS.

[105.

colours are present in smaller proportion than others, and are even sometimes altogether wanting. Strontic nitrate, for instance, gives a red tint to burning bodies; and the prism shows that in such light the blue and violet, or more refrangible rays, are singularly deficient. Common salt produces in burning bodies a nearly pure and homogeneous yellow light, which may be used to illustrate the observations just made upon the cause of colour in natural objects. The brilliant colours of insects or of the plumage of birds, strike strangely on the eye when seen in this yellow light.

(105 a) Dispersive Power.-The prismatic analysis of white light, which has just been considered, is not due entirely to the refracting power of the body by which it is effected. Another element of great importance is the dispersive power, which is independent of refraction. Two substances may possess the same mean refractive index, in consequence of which the mean deviation of the rays transmitted will be the same, and yet the spectra which they furnish may be of very unequal lengths. If a hollow prism be made of plates of glass, and filled with oil of cassia, the spectrum which it produces will be more than double the length of that furnished by a similar prism of flint glass. The dispersive power of oil of cassia is much greater than that of flint glass, especially for the more refrangible rays from F to H, and hence there is a great difference in the length of the two spectra, though the mean refracting powers of the two media do not differ materially. The comparative lengths of these spectra, as obtained from prisms of equal angles, are given in fig. 8o. No. I is the spectrum of oil of cassia; 2 that of flint glass.

FIG. 80.

In the construction of achromatic lenses, two media which differ in dispersive power are employed; by this means the fringe of colours, which is always perceptible around the margin of an object viewed by an ordinary lens of high magnifying power, is removed.

(105 b) Sensitiveness to Heat.-Gladstone and Dale have published (Phil. Trans. 1858, 887) the results of an inquiry into the influence of temperature upon the refracting and dispersive powers of bodies upon light. They find that the index of refraction diminishes in every substance as the temperature increases. The degree of this sensitiveness to the effect of heat varies much in different substances; melted phosphorus and carbonic disulphide being the most sensitive, and water the least sensitive of the bodies experimented on by them. This sensitiveness, however, is independent of the refracting or the dispersive power of the substance; ether, for example, being much more sensitive than water to the action of heat, though the refracting and the dispersive powers of the two liquids are nearly the same. Those bodies which expand most by heat are generally the most sensitive. No sudden change of sensitiveness has, however, been observed on the approach of the liquid to the boiling point. The length of the spectrum also decreases as the temperature rises, the effect of heat being most marked in those substances which have the highest dispersive power. The dis

105 c.] INFLUENCE OF CHEMICAL COMPOSITION ON REFRACTION. 181

persive power is invariably diminished by rise of temperature, though not at the same rate as the refractive index, which is found to diminish in proportion as the density diminishes.

(105 e) Influence of Chemical Composition on Refraction.-Dale and Gladstone further found that the refractive index of any body, minus unity, divided by the density, gives nearly a constant quantity, which they term the specific refractive energy of the body. This result is confirmed by Landolt, who has made an extended series of experiments upon the influence of chemical composition upon refracting power. (Pogg. Annal. 1864, cxxii. 545, cxxii. 595) If be the refractive index of the line C, d the density of the compound, and P its molecular weight, the constant (") would be what Landoit terms the specific refractive energy, and P (") its equivalent of refraction. From his observations Landolt draws the following conclusions, which confirm and extend those of Gladstone and Dale :

d

1. Metamneric bodies have usually very nearly the same specific refractive energy, and consequently, the same equivalent of refraction. Gladstone and Dale, however, remark that though this is true where there is close chemical relationship, isomeric bodies are sometimes widely different in optical properties, as in the case of aniline and picoline.

2. In polymeric bodies the index of refraction and the density increase with each duplication of the formula, but the specific refractive energy diminishes slightly; the indices of refraction do not, therefore, increase exactly in multiple proportion.

3. In homologous series, with few exceptions, the index of refraction and the specific refractive energy increase as the terms ascend in the series, the dif ference for each addition of CH,, becoming progressively less the equivalent of refraction, on the other hand, increases uniformly for each addition of CH..

4 Ou comparing together compounds belonging to different series, the empirical formulæ of which differ only in the number of atoms of carbon, the density, and usually also the refractive index, is diminished by each additional atom of carbon: the specific refractive energy is, in some cases, increased, and in others, diminished, but the equivalent of refraction generally increases very uniformly.

5. When the proportion of hydrogen only is increased, the density, the index of refraction, the specific refractive energy, and the equivalent of refraction are all increased.

6. When the number of atoms of oxygen alone is increased, the density, the index of refraction, and the equivalent of refraction are all increased, whilst the specific refractive energy is diminished.

The equivalent of refraction, it appears, then, is always increased, whether hydrogen, oxygen, or carbon be added to the compound; but the amount of the increase is not uniform for each additional atom of the same element, being in some cases greater than in others, according as the chemical type of the compound varies.

In the case of mixtures, unattended by contraction, Dale and Gladstone find that the specific refractive energy is the arithmetical mean of that of its components; and they conclude, from the results of their entire investigation, that "every liquid has a specific refractive energy composed of the specific refractive energies of its component elements, modified by the manner of combination, and which is unaffected by change of temperature, and accompanies it when mixed with other liquids" (Phil. Trans. 1863, 337).

Dr. Gladstone (Phil. Trans. 1870, 9) has also found that the refraction equivalents of many gases and solids are not changed when these substances

182

FRAUNHOFER'S LINES IN THE SOLAR SPECTRUM.

[105 c.

are dissolved in water or alcohol, and that many elements retain their refrac tion equivalents even when in chemical combination, so that the refraction equivalent of a compound may often be calculated from those of its elements and the refraction equivalent of an element determined by the examination of one of its compounds with another element of which the refraction equivalent is known.

Some elements are found to have different refraction equivalents in different compounds, thus in ferrous salts that of iron is 12 0 and in ferric salts 20'1; again hydrogen in some compounds has an equivalent of 13 and in others of 35. The investigation of the refractive energy of compounds sometimes enables the chemist to obtain a glimpse of their internal structure, for instance, the hydrochlorate of camphene seems to be a compound of the formula C,H,,Cl and not a combination of hydrochloric acid with the hydrocarbon C12H,.

10 16"

10 17

(106) Fixed Lines in the Solar Spectrum; Fraunhofer's Lines; Bright Lines in Artificial Lights.-Newton, by admitting a beam of solar light through a small circular aperture into a darkened room, and allowing it to fall upon a triangular prism of glass, obtained the magnificent coloured image known as the solar spectrum, which shades off by insensible gradations from the least refracted red into the most refracted or violet portion of the light. But it does not appear that any one, till Wollaston's time, a century later, examined the effect of admitting the light through a narrow slit, with sides parallel to those of the prism, and viewing it directly, by placing the eye immediately behind the prism (Phil. Trans. 1802, 378). Wollaston found that the spectrum so obtained was not, as it appeared to be by the ordinary mode of examination, a continuous stripe of light, but that it was crossed at right angles to its length by dark bands, which he supposed divided the colours of the spectrum from one another.

It was not, however, till 1815, that these dark bands were carefully examined, when the celebrated German optician, Fraunhofer, published a minute description of them, accompanied by a careful map, in which he figured more than six hundred of these lines, which have ever since borne the name of Fraunhofer's lines. The more important of these lines he distinguished by the letters of the alphabet, and in the uppermost spectrum shown in figs. 81 and 83, a few of them are given as points of comparison with other spectra.

In order to observe these lines, the sun's light, after admission through a narrow vertical slit into a darkened room, was allowed to fall upon a prism placed with its axis parallel to the slit, and at a distance of about 24 feet (7 or 8 metres) from it. The prism was fixed before the object-glass of a telescope of low power, in such a manner that the angie formed by the incident light with the first face of the prism, was equal to that formed by the refracted beam with the second face, so that the position of the prism was that in which the light is subjected to the minimum amount of deviation. This is consequently generally spoken of

106.]

DARK AND BRIGHT LINES OF THE SPECTRUM.

183

as the position of minimum deviation. Under these circumstances Fraunhofer observed numberless vertical lines, varying in breadth and in strength in different parts of the spectrum. These bands were always visible, whatever was the solid or liquid medium used in the construction of the prism, and whether its refracting angle were great or small; and under all circumstances they preserved the same relative position in the respective coloured spaces in which they occur. This fixed position has enabled the optical observer to use these lines as points of reference by which the refractive indices of a great variety of bodies have been determined with precision.

When, however, the source of the light was varied, as if the flame of a candle, the light of the fixed stars, or the spark from the electrical machine was made use of, a different set of lines was in each case observed to occur.

Beyond this fact-viz., the dependence of the position of the lines upon the source of the light employed-Fraunhofer was unable to ascertain anything connected with their cause.

The inquiry thus launched by Fraunhofer has been followed in four principal branches of research, which may be described as relating to

a. Cosmical lines, or the black lines produced in the light of the sun, the planetary bodies, and the fixed stars.

b. Black lines produced by absorption, a class of phenomenal discovered by Sir D. Brewster, in his observations upon the red vapours of nitrous acid.

c. Bright lines produced by the electric spark when taken between different metallic conductors.

d. Bright lines produced by coloured flames or by the introduction of different substances into flame.

I shall enter into some detail upon this subject, which has acquired great interest and importance from the remarkable investigations of Kirchhoff and Bunsen.

a. The cosmical lines admit of partial reproduction by means of photographs of the spectra in which they occur. Most of these lines shown by the photograph are, however, invisible to the eye, as they occur in that part of the spectrum which is more refrangible than even the violet rays. Edmond Becquerel (Taylor's Scientific Memoirs, 1843, iii. 537) was the first who received the

The moon and the planets, including Venus, Jupiter, Saturn, and Mars, exhibit lines corresponding with those of the Sun. In the case of Jupiter, Saturn, and Mars, additional lines are also visible, owing to an absorptive action, due, as is believed, to the atmospheres surrounding these planets (Huggins and Miller, Phil. Trans. 1864, 421). Sirius shows different lines, and Castor others somewhat similar. Amongst the lines of Procyon, Fraunhofer recognised the solar line D, and in those of Capella and Betelgeux, both D and b (Brewster's Edin. Journ. Science, 1828, viii. 7). (109 note.)