CHAPTER III.

ANALYSING SUNLIGHT.

THE researches of telescopists have revealed many important facts respecting the Sun's constitution. Studied thoughtfully, these researches enable us to answer many questions which at first sight seem to require more powerful modes of inquiry. But, undoubtedly, the science of solar physics is too vast and too difficult to be satisfactorily treated by telescopic research alone. The condition of the Sun is so different from that of any bodies we can experiment on, that his mere aspect --and the telescope can show us nothing more-is insufficient to tell us what his constitution may be. The picture of the Sun presented by the most powerful telescope resembles a book, full of meaning indeed, but written in an unknown language. The spectroscope is the means by which that unknown language has been in part interpreted.

Let us consider what spectroscopic analysis really is. It is scarcely possible to treat of any astronomical subject in the present day without describing the most powerful of all instruments of astronomical research; but in the case of the Sun it would be hopeless to

attempt such a course. Spectroscopic analysis forms. the very basis of all our ideas respecting solar physics. We must thoroughly understand the mode in which it teaches, the meaning of its teaching, and the extent to which its teaching may be relied on, otherwise our views will be vague and unsatisfactory, depending rather on the statements of others than on any clear apprehension of their truth on our own part.

For this reason I shall spare no pains to make the explanation of spectroscopic analysis which follows as simple, and, at the same time, as complete as possible. There is, in reality, nothing difficult in the subject; but it cannot be denied that, considering its simplicity, it is not nearly so well or so widely understood as it might be, a circumstance the more to be regretted because the whole history of recent scientific researches is a sealed book to those who do not clearly recognise the nature of the instrument by which those researches have been effected.

He

Newton was the first who showed that white light is a compound of light of many different colours. proved this by his investigation of an experiment of Grimaldi's-illustrated in fig. 21. Here A в represents the course of a pencil of solar light† passing through

* Grimaldi was, however, the first who discovered the effect of passing sun-light through a prism. (See his Physico-mathesis de lumine, prop. xxx. et seq.)

† I have purposely modified Newton's figure, because many misunderstand a figure in which the pencil A B is shown with its proper divergency. They confuse the dispersing effect of the prism with the optical effects produced on a diverging pencil of pure light. It need hardly be said that it would be quite as great a mistake to neglect the

H

a circular aperture in a screen s s'. The prism P is so placed as to intercept the light. It will be well to consider the prism as placed with the base E F uppermost and horizontal.*

Now, if the prism were removed, the light would fall at i and make a small elliptical image there. And if the solar light were simple instead of being composed of rays of many different colours, it would follow such

a course as is indicated by the bent bright line, and form a small elliptical image at i-this image being white. like that at i, and resembling the latter image in shape.†

effect of divergency altogether. In fact, many important optical considerations are associated with the divergency (however small this may be made) of the pencil analysed by the prism. But the two matters are best kept apart, at least in the case of the beginner. In the above description the object has been to direct the reader's special attention to the dispersive action of the prism, and therefore no attention is paid to the divergency of the pencil.

*E G is supposed equal to GF, so that these lines are equally inclined to the vertical.

It would indeed be absolutely identical in shape with the image at i if the exact course indicated in the figure were followed; but if the

But instead of this, Newton found, as Grimaldi had found before him, that a streak of light was formed as VR (VR is exaggerated in length), the streak being violet at the highest point and thence changing through indigo, blue, green, yellow, and orange, to red at the lowest point. Neither above nor below was the streak well defined, but passed gradually into darkness. At the sides, however, the streak was well defined, and in breadth equal to the horizontal breadth of the figure at i. It thus formed a rainbow-tinted streak or ribbon.

It appears from this experiment that light consists of rays of all the colours of the rainbow, that the violet rays are the most bent by the action of a prism, the red rays least, the others in the order named above. This happens with prisms of all refracting angles and of whatever substance. Hence the rays forming the violet part of the spectrum are often called-without further description-the most refrangible rays, while the red rays are called in the same way the least refrangible rays. This mode of speaking, and the expressions rising out of it, should be carefully noted.

Now, the streak of light seen by Newton showed no breach of continuity. Newton appears to have suspected the possibility that by a change in the conditions of his experiment the streak would show gaps. In other words, he suspected that light of all degrees of

single image fell at v or R this would not be the case. It need hardly be remarked that for pure light the course of the beam would depend on the refracting angle G of the prism.

refrangibility, between the light which forms the extreme violet and the extreme red of the streak, may not be present in the solar beam. But he did not succeed in proving this, though he employed apertures of different shape whereby to admit the light. It is clear that if there were simple violet light, and simple indigo light, and so on in the solar beam, a succession of small coloured images would be formed as shown in the figure at V, I, &c., and between these images dark spaces would be seen. Newton's experiments led him to the conclusion that an infinite number of images, shifting by indefinite gradations from v to R, exist along the streak, and so cause the colour to vary insensibly from violet to red as observed in his first experiment.

Wollaston was the first who succeeded in showing that there are gaps in the spectral streak.

It is clear that the circular aperture in the above experiment is not suitable for determining whether rays of all degrees of refrangibility are included in a beam of solar light. In fig. 21 an image of the aperture is represented at V, another at I, another at B, and so on; though of course the spectrum in Newton's experiment showed no such separate images. Now, it is perfectly obvious that if instead of the seven images represented in the figure there were twenty or thirty along the spectrum v R, there would be no means of knowing that the spectrum was made up of these twenty or thirty distinct images, for they would overlap, and so show a continuous streak of light. Wollaston found that when, instead of a circular, triangular,