CHAPTER XVII.

ASTROPHYSICAL INSTRUMENTS.

So far we have been concerned with instruments which enable us to ascertain the positions, dimensions, and appearances of the various orders of heavenly bodies; but we can go further than this, and learn something of the physical and chemical constitutions of the glittering orbs by which we are surrounded. We can, for instance, bring instrumental aid to bear upon the determination of the brightnesses of the heavenly bodies, and by means of that powerful appliance of modern astronomy—the spectroscope—we can study the chemistry of all those bodies which shine by light of their own, and which are not so feebly luminous as to be out of our range.

PHOTOMETRY.—The naked eye was alone employed in observations of stellar brightness until quite recently. Each step in the advance of astronomical research, as in most other branches of science, however, depends upon the greater precision of observation which can be introduced, and so we now find the eye to be assisted in these inquiries by a _photometer_ of some kind or other. The general purpose of photometry will be familiar to all in connection with such practical matters as the determination of the illuminating power of coal gas. The methods here employed, however, are not directly applicable to the comparatively feeble light-sources which have usually to be dealt with in astronomical photometry.

As will be more fully explained in another part of this work, the stars visible to the naked eye are divided into six grades of magnitude. The brightest of them are classed as first magnitude, while those only just visible to the naked eye are of the sixth magnitude. Now that telescopes are used, this division of stars into magnitudes must be continued in some form or other, so as to include telescopic stars. From photometric comparisons it has been ascertained that the average star of the first magnitude may conveniently be reckoned 100 times as bright as a sixth magnitude star. Hence, the light-ratio corresponding to a difference of a single magnitude is 2·5. Thus, a star which is 2½ times less bright than one of the sixth magnitude ranks as seventh magnitude, and so on. Fractions of magnitudes are also necessary to express the results which can now be obtained.

LIMITING APERTURES.—For the reason that a large telescope enables us to see stars which are too dim to be visible in a smaller one, the brightnesses of stars may be compared with more or less satisfactory results by reducing the aperture of a telescope until the star in question ceases to be visible. This is called the method of _limiting apertures_, and in practice a telescope intended for this work would be provided with a series of diaphragms, or other arrangement for conveniently reducing the effective area of its object-glass. A telescope which has an object-glass 10 inches in diameter should just show stars of the fourteenth magnitude under favourable conditions; a star which could just be seen when this aperture was reduced to an inch would be of the ninth magnitude, and so on.

There are numerous reasons why this method fails to give satisfactory results, but one of the most important is that the image of a star becomes more diffuse with each reduction in the aperture of the telescope. At best it must evidently fail for a comparison of stars which are visible to the naked eye.

WEDGE PHOTOMETER.—One of the simplest and best methods of estimating star magnitudes is afforded by the _wedge photometer_. This is a strip of neutral-tinted glass about six inches in length, and a quarter to half an inch deep, tapering from one end to the other, so as to present a gradual reduction in depth of tint from the thick to the thin end. A similar wedge of clear glass, tapering the opposite way, is cemented to this, in order to get rid of prismatic action. Compensated in this way, and mounted in a suitable frame, the wedge is placed in front of the eye-piece of a telescope, and is pushed along until the star under examination is just extinguished. A scale is then read off, and from the results of a previous evaluation of the wedge in the laboratory, the corresponding star magnitude is easily deduced.

In order to eliminate the effects of differences in the state of the sky, the position of the wedge at which a standard star, such as Polaris, ceases to be visible, is determined, and then it is the difference of wedge readings upon which the final calculation is based.

The great value of the wedge in stellar photometry was demonstrated by the labours of the late Prof. Pritchard, to whom we owe the catalogue of the magnitudes of naked eye stars in the northern hemisphere known to the astronomical world as the “Uranometria Nova Oxoniensis.”

OTHER PHOTOMETERS.—Some photometers depend for their action upon comparisons with terrestrial sources of light. In some cases, an artificial star, consisting of a pinhole illuminated by a standard lamp, is brought into the same field of view as the star to be compared, and then, by polarising apparatus, the brightnesses of the two images are equalised. The amount of reduction of either of the stars is determined by a scale which measures the rotation of the polariscope, and in this way all the stars are compared with an artificial star of known brightness.

One of the most notable achievements in this field of astronomical work is that of Professor Pickering of the Harvard College Observatory, who invented and made splendid use of the so-called _meridian photometer_. Here the telescope has two object-glasses of equal aperture side by side, and in front of each is a silvered flat mirror inclined at an angle of 45° to the optic axes. The telescope is supported in an east and west direction, so that one mirror reflects the Pole Star into its object-glass, while the other can be rotated so as to reflect any other star which is on the meridian into the second object-glass. Again, by a polariscope at the eye end of the telescope the images of the two stars are made of equal brightness, and the readings give the data for calculating the required magnitude.

Photographs of the stars are also largely employed for the estimation of magnitudes, stars of different magnitudes being represented on the photographs by spurious discs of different sizes. If all stars gave out light of the same quality, the photographic method would be very trustworthy, but as the colours of the stars vary, the photographic and visual magnitudes are not invariably in agreement A bright, reddish star, such as Betelgeuse, would photographically be only equivalent to a white star which was much less bright to the naked eye.

THE PRISMATIC SPECTROSCOPE.—Reference has already been made in these pages to the wonderful field of astronomical research which has been opened up by the discovery of the action of a triangular glass prism upon rays of light, and the subsequent improvements in the method of utilising this effect.

A prismatic spectroscope may be regarded as an arrangement which will enable us to get a pure spectrum, and to observe it to the best advantage. The light to be analysed is admitted through a narrow aperture called the _slit_, which is placed at the focus of a double convex lens. Emerging from this _collimator_, as a parallel beam, the rays pass through the prism, and after deviation and dispersion they fall upon another double convex lens, which brings them to a focus in the form of a spectrum. An eye-piece may then be employed to view the spectrum, or a sensitive plate may be placed at the focus to photograph it.

In a simple form of spectroscope the prism is supported at the centre of a graduated circular plate, to which the collimator is firmly fixed, while the observing telescope is attached to an arm pivoted at the centre of the plate. A vernier moving with the telescope indicates the position, on a scale of degrees, of any colour brought to the centre of the field of view.

The best results are obtained when the rays of light emerge from the prism at the same angle at which they enter it, in which case the prism is said to be at _minimum deviation_, for the reason that the deflection of the rays from their original path is then the least possible. As lights of different colours are refracted unequally, it is clear that the prism can only be at minimum deviation for rays of one particular colour at any instant. Frequently, however, there is an automatic arrangement by which, as the observing telescope is moved so as to bring different colours into the field of view, the prism is turned so as to be at minimum deviation for the colour actually under observation.

The appearances observed in the spectroscope are a series of images of the aperture through which the light is admitted. If the source of light be yellow, such as that of a spirit lamp flame when common salt is introduced, a yellow image of the aperture will be seen, and so on for other monochromatic radiations. When a white light is observed, images of every gradation of colour are formed, and in such a “continuous spectrum” the separate images cannot be recognised. The form of aperture most widely adopted is a narrow straight slit with parallel sides. In this case there is the least possible confusion, because the several images of the slit appear as so many spectrum “lines.”

For observations of the sun, where the light is so intense, a great number of prisms, each drawing out the spectrum into a longer band, may be employed, so that the lines of the spectrum may be widely separated, and the peculiarities of each more closely investigated. For the fainter bodies, however, the instrument must generally be one of comparatively small dispersion, so that the light may not be spread out into invisibility. It will be evident that the longer the spectrum the greater will be the chances of accurate measurements.

Another way of obtaining great dispersion is to use prisms of the new dense Jena glass, one of which is equal to three or four of the flint glass prisms in general use.

There are various forms of the prismatic spectroscope. In some of them reflecting prisms are introduced to turn the rays back through the dispersive train, so as to get increased dispersion without increasing the number of prisms. In the so-called _direct vision spectroscope_, prisms of different kinds of glass are combined so that the rays of light leave them in nearly the same direction that they enter. Here the collimator and observing telescope are in the same straight line, and this is a great convenience in certain classes of observation.

THE GRATING SPECTROSCOPE.—Sometimes, especially in instruments designed for solar observations, the prisms are replaced by what is called a diffraction grating. Usually this consists of a piece of highly polished speculum metal, upon which is ruled a great number of equidistant parallel scratches or lines. A portion of the light falling upon the grating is simply reflected, while the remainder is spread out into two series of beautiful spectra, one on each side of the directly reflected beam. The two nearest to the directly reflected beam are called spectra of the first order, while following these are spectra of the second, third, and fourth orders; the length of spectrum increasing in each case, and all being available for observation if the light dealt with be sufficiently bright. The production of these spectra is due to the interference of light waves.

All gratings produce exactly similar spectra, so that the distances between identical lines as seen with one grating are always strictly proportional to their distances as seen with any other. With prisms, the relative separation of colours is by no means constant; a prism made of one kind of glass may, for example, separate the green and yellow more than another prism made from different material, while the separation of yellow and red might be the same in both cases. The grating spectrum accordingly affords a constant standard of reference, and what is called the “normal solar spectrum” is the spectrum of the sun mapped with the various dark lines in the relative positions shown by a grating spectroscope.

Prof. Rowland, of John Hopkins University, has introduced a form of grating spectroscope, in which the grating is ruled on a concave spherical surface of speculum metal. After passing through the slit the rays of light fall directly upon this concave surface, and are brought to a focus after reflection, so that no lens except the eye-piece used for visual observations is required. Several of these gratings, having mostly a radius of curvature of about 21 feet, and a ruled surface of about 5½ inches x 2 inches, with 20,000 lines to the inch, are in use at the present time. Some idea of the difficulties to be faced in making these magnificent aids to research maybe gathered from the following remarks of Mr. J. S. Ames:—“It takes months to make a perfect screw for the ruling engine, but a year may easily be spent in search of a suitable diamond point.... When all goes well it takes five days and nights to rule a 6 inch grating having 20,000 lines to the inch. Comparatively no difficulty is found in ruling 14,000 lines to the inch.”

With the aid of these wonderful gratings, the solar spectrum can be photographed with perfect definition, and extending over a total length of several yards. Thousands of the tell-tale Fraunhofer lines are rendered visible in this way.

MEASUREMENT OF SPECTRA.—The spectra of many substances, including hydrogen and iron, are so characteristic as to be recognisable at a glance by an experienced observer, but one must as a rule resort to measurement for the identification of lines, or for the purpose of locating unknown lines for future reference. One of the simplest methods of measurement is that of reading the position of the observing telescope upon a graduated circle, when the line is seen at the centre of the field. If supplemented by a micrometer eye-piece, for differential measures with regard to known spectra, this method is extremely convenient. As recorded on arbitrary scales of this character, the position of the same line would be represented by a number which would be different for every instrument, and it is therefore necessary to reduce all measurements to a common scale; that now universally adopted is the natural one of wave-lengths. The position of a line in the spectrum depends upon the length of the waves constituting the rays of light which produce it, so that a measure of wave-length completely specifies the situation of a line whatever spectroscope maybe employed. Light waves are excessively minute, but by the use of the diffraction grating they can be measured with great accuracy. So small are they, that the most convenient unit of wave-length is the ten-millionth part of a millimetre[5]—or tenth metre, as it is technically named. Expressed in this way, the wave-length of the glorious red line seen in the spectrum of hydrogen is 6563·07, while that of the blue line characteristic of the same gas is 4861·51.

When the positions of a certain number of lines of known wave-length have been read off on the scale of any spectroscope, the required wave-lengths of other lines are ascertained by a graphical interpolation, or by calculation. Elaborate tables of the wave-lengths of the lines in the spectra of the sun and chemical elements have been prepared by various investigators, and these are in constant demand by all workers in the field of astrophysics.

THE TELESPECTROSCOPE.—For the examination of the spectra of the heavenly bodies, a spectroscope is attached to the eye end of a telescope from which the eye-piece has been removed, such a combination forming a _telespectroscope_. The slit is placed at the principal focus of the object-glass of the main telescope, and an image of the object to be observed is thus produced upon it. If the sun be under observation, any special part of it, such as a sun-spot or the chromosphere, may be separately investigated by bringing the corresponding part of the image upon the slit.

In the case of the sun, moon, comets, planets, or nebulæ, the image is one of sensible size and the spectrum lines have a perceptible length. With a star, however, the image is only an illuminated dot upon the slit, and the spectrum would have no appreciable breadth, so that all but the strongest lines would in general fail to show themselves. Accordingly, when observing star spectra, a cylindrical lens is placed in front of the slit, so that the stellar image is drawn out into a bright line, and the necessary breadth of spectrum and length of the spectrum lines are secured.

For photographing the spectra of the heavenly bodies it is simply necessary to replace the eye-piece by a small camera, and to expose a sensitive plate for a length of time dependent on the brightness of the spectrum. The spectrum of a terrestrial substance, such as hydrogen or iron, photographed in juxtaposition, is always a great convenience, and is essential for the investigation of stellar movements by the displacement of spectrum lines.

THE LICK STAR SPECTROSCOPE.—Among the most complete and perfect spectroscopes adapted for use with the telescope is that designed by Prof. Keeler for the great refractor of the Lick Observatory. It is illustrated in Fig. 55, and it will be at once evident that there are ample means for keeping the instrument under control. Towards the upper part of the diagram, on the left, is the eye end of the telescope, without the eye-piece. Two stout brass rods 3 inches in diameter and 6 feet long are attached by clamps to a revolving jacket which surrounds the end of the telescope tube, and on these the spectroscope is supported by clamps which allow of it being moved inwards or outwards from the focus of the telescope. The collimator of the spectroscope lies midway between the rods, and in order to facilitate the focussing of the image upon the slit, it has a small longitudinal movement independently of that of the whole spectroscope. The observing telescope is seen on the left of the diagram, while the grating rests on the circular graduated plate over which the observing telescope can be moved. The grating has 14,438 lines to the inch.

Three prisms can also be used with the spectroscope, two of them being single prisms of 30° and 60° refracting angles respectively, and the third a compound prism giving a very high dispersion. Two observing telescopes are provided, one being of extra power for use with the grating in solar spectroscopy

The instrument is generously supplied with the small refinements which contribute so largely to easy and successful manipulation. Among these are a diagonal eye-piece for viewing the image of the object on the slit plate, electrical illumination of the graduated scale and wires of the micrometer eye-piece, and an automatic arrangement for keeping the prisms at minimum deviation.

There is a small totally-reflecting prism covering half of the slit, by which the light from an electric spark, or other source of luminosity, can be made to pass through the spectroscope so as to produce a series of known reference lines which serve as so many mile-posts for the measurement of the spectrum of the celestial body under observation. The induction coil, seen to the right of the diagram, is for the purpose of producing these electrical sparks.

In mounting the spectroscope, which weighs no less than 200 pounds, the eye end of the great telescope tube is first supported by a prop, and the long rods are inserted. The spectroscope is then placed on the rods, and balancing weights equivalent to the weight of the spectroscope are removed from the lower part of the telescope tube.

[Illustration:

FIG. 55.—_The Spectroscope adapted to the Eye End of the Lick Telescope._ ]

THE OBJECTIVE PRISM.—It is a very remarkable fact that many of the recent advances in our knowledge of the spectra of stars have followed from the revival of a method first employed by Fraunhofer in 1814, in which the slit and collimating lens, forming part of an ordinary spectroscope, are dispensed with. The rays coming from a star being already parallel, and the star itself being a virtual slit without length, a large prism placed in front of the object-glass of a telescope makes a complete stellar spectroscope. A prism employed in this way is known as an _objective prism_.

In place of the image of a star, which would be seen in the absence of the prism, a spectrum without appreciable width appears at the focus of the telescope, and the spectrum lines will be represented by mere dots. To turn these dots into lines so that they may be better visible, a cylindrical lens must be employed in conjunction with the eye-piece.

It is to the application of photography, however, that we owe so much, and in this case the cylindrical lens is removed, while a small camera replaces the eye-piece of the telescope. In this form the instrument is often called a _prismatic camera_.

The prism is so arranged that the spectrum lies along the meridian passing through the star, and it is then only necessary to allow the driving clock to be slightly in error in order that the spectrum may trail a short distance perpendicular to its own length, and in this way broaden the photographed spectrum. On the proper regulation of the clock rate, and consequent “trail” of the spectrum across the plate parallel to itself, depends very largely the success of the photograph obtained. The spectrum of a bright star must obviously be made to travel more quickly than that of a fainter one, and a short exposure suffices. For the same clock rate, and in the same time, a star near the Pole will give a shorter trail than one nearer the Equator, and declination must therefore be taken into account in adjusting the clock error for this method of photography.

One great advantage of the objective prism in the photography of stellar spectra depends upon the fact that all the light passing through the object-glass is utilised in the production of the spectrum, whereas in an ordinary telespectroscope a large percentage of the light is lost in the jaws of the slit. The large focal length of the telescope also enables a long spectrum to be obtained even with a single prism of small angle.

When the dispersion is only small, the spectra of stars as faint as the tenth or eleventh magnitude can be photographed by this method, so that sometimes as many as 200 spectra are registered with a single exposure. Here, again, the objective prism has an immense advantage over the telespectroscope.

Professor Pickering, of Harvard College, was among the first to recognise the value of the objective prism for the photography of stellar spectra, and the munificent endowment of this research, by Mrs. Draper, as a memorial to Dr. Henry Draper, has enabled him to produce the Draper catalogue of stellar spectra, giving the chief characteristics of the spectra of over 10,000 stars.

Professor Norman Lockyer, at South Kensington, has also been conspicuously successful in this department of astrophysical research. The chief instrument he employs is a photographic telescope of only six inches aperture, with an objective prism of 45° refracting angle. The spectra thus obtained show hundreds of lines in such stars as Arcturus, with very fine definition, so that they bear almost unlimited enlargement.

An objective prism of twenty-four inches aperture will form one of the accessories of the fine telescope which is now being erected at the expense of Dr. Frank McClean, for the Cape Observatory, and there can be no doubt that the use of this gigantic prism will add greatly to our knowledge of the chemistry of the fainter stars.

As yet there is no very practicable method of employing the objective prism for determining the velocities of stars in the line of sight from the displacement of spectrum lines, and herein lies its one great disadvantage as compared with the telespectroscope. The difficulty is to ensure that the spectrum always falls absolutely in the same position with respect to the terrestrial spectrum, which must be photographed alongside for purposes of measurements. It is true that the spectrum of an approaching star is somewhat shorter, and of a receding star slightly longer than that of one at rest relatively to the observer, but these changes are so small as to little more than indicate the direction of movement even when a large instrument is employed.

Under the direction of Professor Norman Lockyer, the objective prism was very successfully used for photographing the spectra of the solar surroundings during the total eclipses of 1893 and 1896. In place of the picture of the solar corona, which would appear in the absence of the prism, the prismatic camera shows a spectrum consisting of bright rings. If, for instance, the corona were wholly composed of hydrogen, there would be a picture of it in red, blue-green, blue, and violet, corresponding to the lines ordinarily seen in the spectrum of that gas. These rings thus indicate the chemical nature of the corona, and at the same time show, by their differing forms, the distribution of different gases throughout its extent. The spectra of the solar prominences and chromosphere are also depicted during the brief time of their visibility, during an eclipse, with such distinctness that a series of “snap shots” is all that is required to give a lasting record.

THE SPECTROHELIOGRAPH.—A special form of spectroscope—called the _spectroheliograph_—has been devised by Prof. Hale, of Chicago, for photographing the sun in monochromatic light. It consists of a spectroscope, arranged for photography, in which the slit can be made to travel by clock-work across the sun’s image, which is projected upon it by the telescope to which the instrument is attached. In front of the photographic plate there is a secondary slit, so that only a very restricted part of the spectrum reaches the sensitive film. The secondary slit is connected by mechanism with the primary one, so that as the latter traverses the sun’s image, the former exposes different parts of the photographic plate to the light which passes through it, and in this way builds up an image of the sun in monochromatic light, matters being so arranged that light of the same wave-length always falls upon the secondary slit. By utilising the brightest lines which appear in the spectrum of the solar prominences, monochromatic images of those interesting appendages to our luminary have been successfully photographed without waiting for a total solar eclipse.

THE BOLOMETER.—Besides the luminous effects of the spectrum, there are heating effects which can be measured by the _bolometer_, an instrument invented by Prof. Langley. A very thin strip of metal is connected with a delicate galvanometer, and is arranged so that it can be passed a long the whole spectrum. The electrical resistance of the strip varies according to its temperature, and the galvanometer at once signals any fluctuations which may occur. If, for instance, the strip comes to the place occupied by a dark line, there will be a notable fall of temperature. In this way, the bolometer is used to map the solar spectrum in the “infra-red” region—a part of the spectrum invisible to the eye, and of which we should otherwise have remained in ignorance.

ASTRONOMY

[Illustration:

DONATI’S COMET, OCTOBER 9, 1858. (FROM LANGLEY’S “NEW ASTRONOMY”.) ]

SECTION III.—THE SOLAR SYSTEM.

BY AGNES M. CLERKE.