Even as early as 1785, however, Herschel perceived traces of a tendency which completely invalidated the supposition of any approach to an average uniformity of distribution. This was the action of what he called a "clustering power" in the Milky Way. "Many gathering clusters"[35] were already discernible to him even while he endeavoured to obtain a "true _mean_ result" on the assumption that each star in space was separated from its neighbours as widely as the sun from Sirius. "It appears," he wrote in 1789, "that the heavens consist of regions where suns are gathered into separate systems"; and in certain assemblages he was able to trace "a course or tide of stars setting towards a centre," denoting, not doubtfully, the presence of attractive forces.[36] Thirteen years later, he described our sun and his constellated companions as surrounded by "a magnificent collection of innumerable stars, called the Milky Way, which must occasion a very powerful balance of opposite attractions to hold the intermediate stars at rest. For though our sun, and all the stars we see, may truly be said to be in the plane of the Milky Way, yet I am now convinced, by a long inspection and continued examination of it, that the Milky Way itself consists of stars very differently scattered from those which are immediately about us." "This immense aggregation," he added, "is by no means uniform. Its component stars show evident signs of clustering together into many separate allotments."[37] The following sentences, written in 1811, contain a definite retractation of the view frequently attributed to him:-- "I must freely confess," he says, "that by continuing my sweeps of the heavens my opinion of the arrangement of the stars and their magnitudes, and of some other particulars, has undergone a gradual change; and indeed, when the novelty of the subject is considered, we cannot be surprised that many things formerly taken for granted should on examination prove to be different from what they were generally but incautiously supposed to be. For instance, an equal scattering of the stars may be admitted in certain calculations; but when we examine the Milky Way, or the closely compressed clusters of stars of which my catalogues have recorded so many instances, this supposed equality of scattering must be given up."[38]