Other telegraphic marvels might well find a place here. I might speak of the wonders of submarine telegraphy, and of the marvellous delicacy of the arrangements by which messages by the Atlantic Cable are read, and not only read, but made to record themselves. I might dwell, again, on the ingenious printing telegraph of Mr. Hughes, which sets up its own types, inks them, and prints them, or on the still more elaborate plan of the Chevalier Bonelli “for converting the telegraph stations into so many type-setting workshops.” But space would altogether fail me to deal properly with these and kindred marvels. There is, however, one application of telegraphy, especially interesting to the astronomer, about which I must say a few words: I mean, the employment of electricity as a regulator of time. Here again it is the principle of the system, rather than details of construction, which I propose to describe. Suppose we have a clock not only of excellent construction, but under astronomical surveillance, so that when it is a second or so in error it is set right again by the stars. Let the pendulum of this clock beat seconds; and at each beat let a galvanic current be made and broken. This may be done in many ways—thus the pendulum may at each swing tilt up a very light metallic hammer, which forms part of the circuit when down; or the end of the pendulum may be covered with some non-conducting substance which comes at each swing between two metallic springs in very light contact, separating them and so breaking circuit; or in many other ways the circuit may be broken. When the circuit is made, let the current travel along a wire which passes through a number of stations near or remote, traversing at each the coils of a temporary magnet. Then, at each swing of the pendulum of the regulating clock, each magnet is magnetized and demagnetized. Thus each, once in a second, draws to itself, and then releases its armature, which is thereupon pulled back by a spring. Let the armature, when drawn to the magnet, move a lever by which one tooth of a wheel is carried forward. Then the wheel is turned at the rate of one tooth per second. This wheel communicates motion to others in the usual way. In fact, we have at each station a clock driven, _not_ by a weight or spring and with a pendulum which allows one tooth of an escapement wheel to pass at each swing, but by the distant regulating clock which turns a driving wheel at the rate of one tooth per second, that is, one tooth for each swing of the regulating clock’s pendulum. Each clock, then, keeps perfect time with the regulating clock. In astronomy, where it is often of the utmost importance to secure perfect synchronism of observation, or the power of noting the exact difference of time between observations made at distant stations, not only can the same clock thus keep time for two observers hundreds of miles apart, but each observer can record by