Having got the error, say three (0·3)-tenths of a second slow, this is the arrangement for correcting it. The pendulum is suspended by a spring extra long, and a long arm goes across the clock pier, and the pendulum spring passes through a fine slit in the middle of it, and the left end (of said arm) turns on a pivot, while the right end rests on a cam, which can be turned by a handle outside the clock-case. Turning the handle one way raises the arm, and with that lengthens the acting length of the pendulum spring, and turning the other way, lowers it and shortens the pendulum, but so slightly that it takes fifteen minutes of the quickened rate of the pendulum, when shortened, to add the required 0·3 seconds to the indications of the clock.”[14] The sidereal clock is used in many ways besides the purpose of giving a basis from which we can at any time get solar time, the distribution of which forms the subject of our next chapter. ----- Footnote 13: _Nature_, April 1, 1875. Footnote 14: This plan was devised and executed by Mr. Sang, C.E., Edinburgh. CHAPTER XVIII. “GREENWICH TIME” AND THE USE MADE OF IT. We have now described the method of obtaining and keeping true Greenwich time by means of transit observations, and the next thing is to distribute it either by controlling or driving other clocks electrically, or by sending electric signals at known times for persons to set their clocks right. Nearly all, if not quite the whole, of the mean-time clocks in the Observatory are driven by a current controlled by the standard clock, as also is a seconds relay, _a_ Fig. 129. The clock controls, by currents sent every second by the relay, one or two clocks in London, by special wires. So long ago as the year 1840 Sir Charles Wheatstone read a paper before the Royal Society in which he described an apparatus for controlling any number of clocks by one standard clock at a distance away. The principle was, that at each beat of the standard clock an electric current was sent from it through a wire to the clocks to be worked by it or governed; and this current made an electro-magnet attract a piece of iron each time it was sent; and this piece of iron moved backwards and forwards two pallets, something like those of an ordinary clock, which turned a wheel, and so worked the clock. Instead of a spring or weight being used to work it regulated by the pallets, the pallets moved the clock themselves, and of course keep time with the standard clock. Sir Charles Wheatstone in this most valuable pioneer paper, indicates several modifications of this plan. He proposed to the Astronomer-Royal to test his method by using the then new telegraph line to Slough, but the idea was not carried out. This method of _driving_ clocks by electricity naturally required considerable battery power, and in the more modern systems the clocks are simply _controlled_, and not _driven_, by electric currents.