The great improvement of giving to a telescope equatoreally mounted an automatic movement by connecting it with clockwork, was proposed in 1674 by Robert Hooke. Bradley in 1721 actually observed Mars with a telescope "moved by a machine that made it keep pace with the stars;"[337] and Von Zach relates[338] that he had once followed Sirius for twelve hours with a "heliostat" of Ramsden's construction. But these eighteenth-century attempts were of no practical effect. Movement by clockwork was virtually a complete novelty when it was adopted by Fraunhofer in 1824 to the Dorpat refractor. By simply giving to an axis unvaryingly directed towards the celestial pole an equable rotation with a period of twenty-four hours, a telescope attached to it, and pointed in _any_ direction, will trace out on the sky a parallel of declination, thus necessarily accompanying the movement of any star upon which it may be fixed. It accordingly forms part of the large sum of Fraunhofer's merits to have secured this inestimable advantage to observers. Sir John Herschel considered that Lassell's application of equatoreal mounting to a nine-inch Newtonian in 1840 made an epoch in the history of "that eminently British instrument, the reflecting telescope."[339] Nearly a century earlier,[340] it is true, Short had fitted one of his Gregorians to a complicated system of circles in such a manner that, by moving a handle, it could be made to follow the revolution of the sky; but the arrangement did not obtain, nor did it deserve, general adoption. Lassell's plan was a totally different one; he employed the crossed axes of the true equatoreal, and his success removed, to a great extent, the fatal objection of inconvenience in use, until then unanswerably urged against reflectors. The very largest of these can now be mounted equatoreally; even the Rosse, within its limited range, has been for some years provided with a movement by clockwork along declination-parallels.