Imagini ale paginilor

mixing with the blue, already come, would make the star change from blue to green ; while the red, arriving last of all, and joining themselves to the existing green, would at length make the star appear white; and, if it preserved the same lustre, it would ever after continue white.

If the star, however, is shown for only an instant of time, as an electric spark, then we should see it varying through each of the different colours, blue, yellow, and red, separately and distinctly. If, for example, the blue ray was to traverse the space between the star and ourselves in three years, the yellow in three years and one week, and the red in three years and two weeks, and supposing the above to apply only to the central portion of each coloured ray, which should gradually vary with filaments of different velocities so as to join insensibly with those of the neighbouring one, then, three years after the striking of this stellar spark, we should see a blue star appear in the sky, and last for one week, then the star would appear yellow during another week, and red during another; after which it would be lost altogether. If there are actual separations between the different colours, as is more than hinted at by the discovery of the black bands in the spectrum, then the star, after appearing of one colour, might disappear for a time before the next colour began to arrive.

Again, if a star which has existed for ages be on a sudden extinguished, the rays last emitted will be the couriers to announce the fact; and, supposing the star to have been white, three years afterwards in the above particular example), the last of the blue rays having arrived before the last of the others, the blue will be deficient in the star, and from white it will become orange; after a week all the yellow ones will have come in, and the star will be red; and, when the final rays of this colour have arrived, it will totally disappear.

But if the star shines permanently, and has so shone from time immemorial, then, whatever might be the difference of time elapsing between the blue and red rays shot out from the star at the same instant reaching us, we should see the star white; for blue and yellow and red rays of different dates of emission would all be reaching our eyes together.

This case can be exemplified by looking through a prism at a white surface of unlimited extent and equal brightness, when it will be seen as white as before ; for the multitudinous spectra formed by all the component points of the whole surface overlaying each other, the red of one coming to the blue and yellow of others, will form white light, as completely as if the three colours of one point be concentrated together again. Here was Goethe's error: he gazed at a white wall through a prism, and, finding it white still, kicked at Newton's theory to produce an absurd one of his own. But had he looked at the edges of the wall—which is a similar case to the birth or death of a star-he would have seen the blue half of the spectrum on one side, and the red on another : everything, in fact, with a sensible breadth, will have coloured borders, blue on one side and red on the other. If one part of the wall, however, be brighter than another, the strong blue of that, thrown on to the fainter red of another, will give that a bluish tinge, and vice versa : and so with the stars, if their brightness should alter, or in the common but singularly erroneous parlance, their magnitudes vary, the strong blue of a bright epoch arriving with the faint red of a dull period, will make blue appear to us as the predominating colour; will cause indeed the star's light to appear decidedly blue at one time, and, mutatis mutandis, red at another, although all the while the star's colour may not really have altered at all; but may have been really, and would have appeared to observers close by, as white as ever, varying only in quantity and not in quality. Real alterations in colour may doubtless occur, but evidently may often be only consequences of alterations in brightness which may be brought about by many regular and periodical phenomena, and certainly do not require the introduction of any such startling reason as the conflagration which was lugged in to explain the tints through which the variable star of 1572 passed, as it gradually died out of the sky, where it had so suddenly appeared a few months previously. Of this, at least, we may be certain, that there are periodical variations in the brightness of the stars, and that some alteration of colour should thereby be produced; but whether to a sensible extent or not, is only to be determined by experiment. B Persei has been selected by Arago as a favourable instance for testing this matter by observation, because it changes so very rapidly in brightness in a short space of time; but, though he did not succeed in detecting any alteration of colour, we must not despair; for, while on the one hand his means of determining the colour seem to have had no sensible degree of exactness, it is easily possible to assume such a difference of velocities for the various coloured rays of the star, and such a distance for them to traverse, as should completely annihilate the expected good effect of the quickness and frequent recurrence of the changes in this particular star; many other stars might indeed be picked out where the natural circumstances are more promising, while the perfection of the means of observation would allow of many more still being made subservient to the inquiry.

The failures made here therefore may be regarded in the same light as those in the olden inquiry of finding the parallax of the fixed stars, viz. not as reasons for leaving off, but for trying again more energetically, more extensively, and with more accurate means than before; and, although I may not be prepared just at present to describe any perfectly satisfactory method of observation, still, as some amateurs desirous of pursuing the subject may like to see such hints as my experience has necessarily given rise to, presented in some rather more practical form, I have thrown them together as follows:

In any method of determining colours of stars, three possible sources of error have to be met : 1. The state of the atmosphere generally at the time in altering the colour of all the stars above the horizon ; 2. The effect of altitude in varying on different stars the apparent colour produced by the atmosphere; and 3. The effect on the eye of the necessary quantity of some sort or other of artificial light, for the purpose of writing down or examining the dimensions of the instrument, the face of the clock, &c. &c.

The first can only be eliminated by extensive observation of a number of stars, especially circumpolar ones, all through the year. Although the colours of some stars may vary in a small number of months, weeks, or even days, the mean of them all may be considered to be safely depended on for a tolerably

[ocr errors]

constant quantity; and each star should be examined and tested for its colour every night, by comparison with the mean of all the rest ; and where any decided variation appears to be going on through the year, that star should at once be excluded from the standard list, and its difference from the mean of the others stated as its colour for each night's observation.

The second source of error is to be met by observations of the same star through a large part of its path from rising to culminating, or of a number of stars of known colour at various altitudes, combined with a correction something similar to that of refraction, as varying in a proportion not far from the tangent of the zenith distance, and which would consequently require the altitude of every body observed to be noted, as a necessary element in reducing the observations.

Low stars, however, should be eschewed, and each observer should confine himself as far as possible to his zenith stars; for, in addition to the low ones being so much fainter to him, than to one to whom they are vertical, and in addition to the colouring and absorbing effect of the atmosphere increasing so excessively, low down on the horizon, the envelope acts so strongly there as a prism that, combined with the bad definition prevailing, I have sometimes seen a large star of a white colour really appear like a blue and red handkerchief fluttering in the wind : the blue and red about as intense and decided as they could well be. This shews the extreme importance of noting not only the altitude of the star, which determines also the degree of prismatic effect, but of distinguishing in the observation any difference between the upper and lower parts of the star. In the Sun and Moon, bodies of very sensible breadth, this effect is not so evident; the surface will still be white or coloured uniformly by the atmosphere, and the upper and lower borders will alone show the prismatic colours, half on one edge and the other half on the other, as in the case of the white wall mentioned above; but the star, being merely a point of light, is wholly acted on, and exhibits as complete a spectrum as could be contrived without any of the white or self-compensating intermediate portion.


Combined with this, is the colouring effect of the object glass, and any deficiency of its achromaticity; but these, being nearly the same on all the stars, will not affect the difference observed: yet the latter quality of the eyepiece will be of more consequence, unless the star be brought very rigorously into the centre of the field of view, and kept there the whole time that it is under observation. An achromatized eye-piece should be specially used.

The third difficulty may be best counteracted by using one eye for the field of the telescope, and the other for writing down, &c.; having the artificial light used for these purposes as faint, and making them as white, as possible, with various other little practical details which will best occur to each observer.

We then come to the grand difficulty: viz. the manner in which the colour is to be determined; the methods are two: first, by the senses; second, by instrumental means. The first is that which has been employed hitherto, and will doubtless still be the only method employed for a considerable time by amateurs; and, though so very vague, yet may—by the education and the practice of the senses, combined with the corrections above consideredbe carried to considerable perfection : but the education must be much more systematic and the practice much more constant than they have hitherto been.

Some certain standard of colours must be kept and constantly referred to : the colours of precious stones have been used for this purpose; but, though very proper in one point of view, as being by their brightness more comparable to stars than ordinary pigments are, yet astronomers in general have not much acquaintance with anything so valuable and costly; and, if they had, would find that the colour of each star is not certainly to be defined by its name, i. e. that under the same name many different colours may be found; and different observers will therefore be giving the same name to stars not resembling each other; in addition to which, there is not a sufficient range of colours amongst the precious stones to meet all the cases which occur in nature in the heavens, and they neither admit of being mixed, to form varieties of colour, nor of being modified, to show gradations in their own colour; a most

« ÎnapoiContinuă »