Mr. Clarke, of Philadelphia, exhibited at the late fair of the Franklin Institute held in that city, a very useful improvement in the telegraph register. By the ordinary arrangement, the operator has to use a key for winding up the register, but by Mr. Clarke's plan the register is self-winding. The winding motion is obtained by an extra magnet being placed in the register, and the closing and breaking of the circuit causes a lever to vibrate. This lever has a click at its end, acting in a small steel ratchet wheel, whieh causes the latter to revolve and transmit its motion by wheel gearing to the shaft of a spring contained in a box, like a watch. A spring is used for a motive power to the train of wheels, instead of a weight, as in the ordinary register. There is also an arrangement by which it ceases winding when the spring is wound to the power necessary to revolve the train of wheels. This is effected by two points coming in contact, and establishing a cross-current, which cuts off the current from the winding magnet, until, by its running, it causes the two points to separate, when the current flows through the magnet again, and the winding is continued. Another advantage of this improvement consists in the tact of a uniformity pf motion throughout any number oi messages being obtained.