the most amazing and mysterious objects revealed by the telescope. In them thousands of suns are grouped together in enormous systems which are nearly spherical in form, the stars being comparatively rare in the outermost parts of a cluster, but becoming progressively closer together toward the centre. But so great are their distances from us that in the case of the great majority of the globular clusters the combined light of their multitude of component suns is not sufficient to render them visible to the unaided eye. Only two of these objects are thus visible: the incomparable southern cluster Omega Centauri, which appears as a faint blurred star; and the largest and brightest of the northern clusters, Messier 13, in the constellation of Hercules, which can be distinguished by a keen, eye as an extremely faint, hazy spot of light.
It will therefore be readily understood that the globular clusters can be satisfactorily seen and photographed only with very large and powerful telescopes; indeed, in many of these clusters the component stars, especially near the centre, are massed so closely together that even the largest modern telescopes fail to show them well separated. Nothing can be more impressive and awe-inspiring, when searching through the depths of the heavens with a powerful telescope, than to come unexpectedly upon an object of this class. All around the stars are comparatively rare; here they are massed together in amazing richness. The component stars are usually so small that the impression of immense distance is given to the observer. Furthermore, the conclusion is irresistible that the apparent proximity of the stars composing these clusters is real; in other words, these stars are intimately related physically; they form an actual group, and are not merely arranged accidentally, one nearly behind the other, in the line of vision.
The observer cannot see these wonderful assemblages of suns without wishing to know something of their real meaning,—of their physical nature. Are the components large, comparable in size with our own sun and with the brilliant stars which can be seen with the unaided eye? If so, the distances of the clusters from the earth must be, in general, much greater than those of the bright stars. It now seems not improbable that this is the case, since even in the very dense clusters the motions of the component stars with reference to each other appear extremely slow,—so slow that in the great majority of cases no motion whatever can be detected by measurement with the largest telescopes, even in a long period of years.
In this work of searching for relative change of position among the components of the star-clusters photography is of immense value. With the great telescopes and refined methods now used in this work the relative positions of thousands of stars in a given cluster are recorded upon the photographic plate with extreme accuracy and with an exposure of the plate during only three or four hours. Such a record, once made, is permanent, and will serve for the most refined measurement, and for comparison with similar photographs made in future years.
Let us now consider how these photographs are made. No camera of the usual kind is employed; the telescope itself is used as a great camera, the photographic plate being placed at the focus of the large lens, or, in the case of a reflecting telescope, at the focus of the concave mirror. The stars in many of the clusters are so faint that even when their light is concentrated at the focus of a large lens or mirror, and when the most sensitive photographic plates are used, several hours' exposure must be given to secure a strong and distinct photograph. The tube of the telescope is therefore mounted on suitable axes, and is slowly and accurately moved by clockwork, so that it follows the diurnal motion of the celestial objects, from east to west, across the sky.
But with the very large telescopes which are necessary to photograph the dense star-clusters satisfactorily, this motion of the telescope given by the driving-clock, smooth and accurate as it is, is not sufficiently perfect to give sharp photographs. It is necessary to devise some means by which the astronomer can watch throughout the entire time of the exposure of the photographic plate, can detect any minute movements of the image of the celestial object from its proper position, and can
