CHAPTER XXVII.

THE CHEMISTRY OF THE STARS: CONSTRUCTION OF THE SPECTROSCOPE.

In the addition of chemical ideas to astronomical inquiries, we have one of the most fruitful and interesting among the many advances of modern science, and one also which has made the connection between physics and astronomy one of the closest.

To deal properly with this part of our book, as the constitution of one of the heavenly bodies can be studied in the laboratory as well as in the observatory, we have to describe physical instruments and methods, as well as the more purely astronomical ones.

In a now rare book published in London in the year 1653, that is to say, some years before Sir Isaac Newton made his important observations on the action of a prism on the rays of light—observations which have been so very rich in results—is given Kepler’s treatise on Dioptrics. From this one finds that the great Kepler had done all he could to try to investigate the action of a three-cornered piece of glass.

It has been considered, that, because Newton was the first to teach us much of its use, he was the first to investigate the properties of the prism. This is not so. Fig. 167 is an illustration taken from this book, by which Kepler shows that if we have a prism and pass light through it, we get three distinct results when a ray (F) falls on the prism. He shows that the first surface reflects a certain amount of light, (D I), and that this is uncoloured, because it does not pass through the glass, and that the remainder is refracted by the glass and part emerges at E, coloured like the rainbow. Then he goes on to show that the second surface of the prism also reflects some light internally, and that there is a certain amount of light leaving the prism at M, and going to K.

By means of a very few experiments Newton was able to show how much knowledge could be got by examination of the prism. The first proposition in Newton’s _Optics_ is an attempt to prove that light, which differs in colour, differs also in degree of refrangibility. We shall recollect from the fifth chapter what this term means, for it was there shown that whenever a ray of light enters obliquely a medium denser than that in which it had been travelling, it is bent towards the perpendicular to the surface, in fact it is refracted, and those rays which are most refracted by the same substance with the same angle are said to be more refrangible than others. Newton’s experiment was very simple. He took a piece of paper, one half of which was coloured red and the other half blue; and this was placed on a stand horizontally, in the light from a window, with a prism between it and the eye.

He went on to show, that when he allowed the beam of sunlight to fall upon the paper, strongly illuminating the red and blue portions, making at the same time all the rest of the room as dark as possible (so that the operation was not impeded by extraneous light), when he held a prism in a particular way, he found that the red and the blue occupied different positions when looked at through the prism. When the prism is held as shown, the red is seen below and the blue above. If the prism be turned with the refracting edge downwards, the red is seen above and the blue below. When the refracting edge is upwards, it is very clear that if the violet is seen uppermost it must be because the violet ray is more refracted, and when the red ray is uppermost, with the refracting edge of the prism downwards, it is because the red ray is the least refracted.

There are other experiments to which he alludes, and by which Sir Isaac Newton considered he had proved that lights which differ in colour differ also in degrees of refrangibility.

Newton at one step went to the sun, and his second theorem is “The light of the sun consists of rays of different refrangibility,” and then he enters into the proof by experiment. The light from the sun passes through a hole in the window-shutter and through the prism which throws a spectrum on a screen. We now see the full meaning of the different degrees of refrangibility. There he had a long band of light of all colours, the red at one end and the blue at the other, showing that the different colours are unequally refracted, or turned from their course. In this way Sir Isaac Newton determined whether the law, that light which differed in colour differed also in refrangibility, held true with regard to the sun; and he clearly showed that in this case also the light differs in refrangibility, in exactly the same way as the red light and the blue light had done in his experiment with the pieces of paper. He was soon able to prove to himself that the circular aperture was not the best thing he could use, because in the spectrum he had a circle of colour representing every ray into which the light could be broken up. If we put a bit of red glass in the path of the rays we get an image of the hole in red; if we use other coloured glasses, we have a circle for each particular colour; all these images overlap, and the sum total gives us an extremely mixed spectrum, something quite different from what is seen when we introduce a slight alteration, which curiously enough was delayed for a great many years.

Sir Isaac Newton recognised the difficulties there were in getting a pure spectrum by means of a circular aperture, but although he used afterwards an oblong opening instead of a circular aperture, in which we had something more or less like what we now use, namely, a “slit”—a narrow line of light; he does not seem to have grasped the point of the thing, because in one of his theorems he says he also tried triangular openings. We shall show how important it is that we should not only have an oblong opening as proposed by Newton, but that that oblong opening should be of small breadth.

The moment we exchange the circular aperture for the oblong opening of Newton, we get a spectrum of greater purity, and, as in the case of the circular opening the purity depended on the size of the circle, so also in the case of the oblong opening the purity of the spectrum depends very much on the breadth of the oblong opening.

We thus sort out the red, orange, yellow, green, blue, and violet; they are no longer mixed as they are when we employ a circular opening. If we attempt the same experiment with red glass interposed we get something more decided than before; we have no longer a circular patch of light, but an oblong one in the red; in fact, the exact form of the aperture, or slit, through which we have allowed the light to pass through the prism and lens to form an image.

Now although Newton made these important observations on sunlight, he missed one of the things, in fact we may say _the_ thing, which has made sunlight and starlight of so much importance to Astronomy. The oblong opening which Newton used varied from one-tenth to one-twentieth of an inch in width; but Dr. Wollaston in 1812—we had to wait from 1672 till 1812 to get this apparently ridiculously small extension—used such a narrow slit as we have mentioned, and he found that when he examined the light of the sun with a prism before the eye, he got results of which Newton had never dreamt.

Dr. Wollaston not only found the light of the sun differing in refrangibility; but in the different colours of the solar light he found a number of dark lines, which are represented by the black lines across the spectrum in Fig. 169.

In the year 1814 Fraunhofer examined the spectrum by means of the telescope of a theodolite, directing it towards a distant slit, with a prism interposed. In this manner he observed and mapped 576 lines, the appearance of the spectrum to him being represented in Fig. 170. From this time they were called the “Fraunhofer lines.” It need scarcely be said that from the time of Wollaston until a few years ago these strange mysterious lines were a source of wonder to all observers who attempted to attack the problem. The difference between the simple prism and slit which Newton, Wollaston, and Fraunhofer used to map these lines, and the modern spectroscope, as used with or without the telescope, is due to a suggestion of Mr. Simms in 1830.

Let us refer to a modern spectroscope. Fig. 171 represents a form usually used for chemical analysis. The only difference between the spectroscope and the simple prism in Newton’s experiment is this, that in the one case the light falls directly from the slit through the prism on a screen and is viewed there; and in the other the eye is placed where the screen is, and looks through the prism and certain lenses at the slit.

The great improvement which Mr. Simms suggested was this simple one. He said, “It would surely be better that the light which passes through the prism or prisms independently of the number I use, should, if possible, pass through them as a parallel beam of light; and therefore, instead of putting the slit merely on one side of a prism and the eye on the other, I will, between the slit and the prism, insert an object-glass,” as shown in Fig. 172; so that the slit of the spectroscope is the representative of the hole in the shutter.

The slit is exactly in the focus of the little object-glass, C, or collimating lens, as it is called; so that naturally the light is grasped by this lens, and comes out in a parallel beam, and travels among the prism or prisms, quite irrespective of course of their number. This parallel beam, in order to be utilized by the eye after it has passed through the system of prisms, is again taken up by another object-glass and reduced from its parallel state into a state of convergence, and brought to a focus which can be examined by means of an eyepiece.

The red rays from the slit come to a focus at R, and the blue at B, forming there their respective images of the slit, and between B and R are a number of other images of the slit, painted in every colour that is illuminating it, thus forming a spectrum which is viewed by the eyepiece. In fact, the object-glass and eyepiece constitute a telescope, through which the slit is viewed, and the collimating lens makes the light parallel, just as if it had come from a distant object, and fit to be utilized in the telescope. This is the principle to be observed in the construction of every spectroscope.

We have now given an idea of the general nature of the instrument depending on this important addition made by Mr. Simms, which is the basis of the modern spectroscope, and it is obvious that if we want considerable dispersion, we can either increase the number of prisms, or increase their dispersive power.

We have already shown in a previous chapter that the dispersion depends on the angle of the prisms, and that the calculations necessary for making the object-glass of a telescope were based upon an observation made by passing light through a prism of a particular angle made of the same glass as that of which the proposed object-glass was to be constructed. Then, again, we took the opportunity of showing that with very dense substances greater dispersion could be obtained. We showed how the prism of dense flint glass overpowered the dispersion of the prism of the crown glass, and how the combination gave us refraction without dispersion.

Fig. 173 is a drawing of a spectroscope containing four prisms. It is a representation of that used by Bunsen and Kirchhoff when they made their maps of the solar spectrum: it is so arranged that the light after passing through the slit goes through the collimating lens, and then through the prisms; it is afterwards caught by the telescope lens and brought to a focus in front of the eyepiece. It is very important, when we have many prisms, to be able to arrange them so that whether we use one part of the spectrum or the other, each prism shall be in the best condition for allowing the light to traverse it; that is to say, that it shall be in the position of _minimum deviation_, when the angles of incidence and emergence are equal, and each surface refracts the ray equally. They can be arranged so, that as the telescope is moved to observe a new part of the spectrum, every prism will be automatically adjusted.

To insure this the prisms are united to form a chain so that they all move together, and each has a radial bar to a central pin which keeps them at the proper angle.

There is another arrangement which is very simple, in which we get the condition of minimum deviation by merely mounting the prisms on a spring, and then moving the spring with the telescope, in the same way as the telescope moves the other automatic arrangement.

For some observations, especially solar observations, in which the light is very intense, it is extremely important, in fact essential, to reduce the brilliancy of the spectrum; and of course this enables us, in the case of the sun especially, to increase the dispersion almost without limit, by having a great number of prisms, or even using the same twice over, in the following manner:

On the spectroscope there is a number of prisms so arranged that the light comes from the slit, and travels through the lower portion of the prisms; it then strikes against the internal reflecting surface of a right-angled prism at the back of the last prism, Fig. 176, and is sent, up to another reflecting surface, and then comes back again through the same prisms along an upper storey, and then is caught by means of a telescope above the collimator, on the slit of which the sun’s image is allowed to fall.

This contrivance, suggested by the author and Prof. Young independently, is now largely used. Fig. 177 shows an ordinary spectroscope so armed. The light from the slit traverses the upper portions of the prisms; it is then thrown down by the reflecting prism seen behind the collimator, then, returning along the lower part, it is received by a right-angled prism in front of the object-glass of the observing telescope.

Instead of the rays of light being reflected back through the upper storey of the prisms, another method has been adopted; the last prism is in this case a half prism, and the last surface on which the rays of light fall is silvered; the rays then are returned on themselves, and, when the instrument is adjusted, come to a focus on the inside of the slit plate, forming there a spectrum, any part of which can, by moving the prisms, be made to fall on a small diagonal reflecting prism on one side of the slit, by which it is reflected to the eyepiece. In this arrangement the collimating lens becomes its own telescope lens on the return of the ray.

There is another form of spectroscope, called the _direct vision_, which is largely used for pocket instruments. The principle of it is that the light passing through it is dispersed but not turned from its course, just the reverse of the achromatic combination of the object-glass; a crown-glass prism is cemented on a flint one of sufficient angle that their deviative powers reverse each other but leave a certain portion of the flint-glass dispersion uncorrected; since, however, the dispersive power of the flint-glass is to a great extent neutralized, therefore, in order to make the instrument as powerful as one of the ordinary construction, a number of flint-glass prisms are combined with crown-glass ones, as shown in Fig. 178.

There is another form of direct-vision prism, called the Herschel-Browning, in which the ray is caused to take its original course on emerging by means of two internal reflections.