390 THE ELECTRIC TELEGRAPH. 



mechanism, a bell is struck at tlie completion of each unit 

 of length (knot or kilometer).* 



No good electrical measurements are possible during the 

 passage out on account of the motion of the ship, and the 

 difficulty of keeping the apparatus dry ; the measurements 

 which are made are only of value qualitatively and approxi- 

 mately in the event of a fault occurring. The electrical 

 conditions of the cable are, however, always kept under 

 surveillance. 



Messrs. Siemens have constructed a galvanometer purposely 

 for use on board- ship. It consists of an astatic system of 

 magnetic needles on a vertical bar, moving in stone pivots, 

 and surmounted by an aluminium pointer. Each needle 

 turns in the centre of an independent coil of wire. Above 

 the glass cover of the dial-plate is a tall rod of brass carrying 

 a horizontal adjusting magnet, which in different positions 

 and at different distances from the magnet system increases 

 or diminishes its directive force, and with this the sensibility 

 of the instrument. 



Professor Thompson has succeeded in eliminating the 

 directive force of the earth's magnetism entirely from his 

 marine galvanometer by surrounding it with a heavy armour 

 of soft iron, which gives it the advantage of retaining its 

 constant of sensibility and zero point in whatever position 

 the ship's head may be put ; he also keeps his mirror magnet 

 steady by suspending it to a tightly drawn cocoon fibre. 



The measuring apparatus on board does not materially 

 differ from that used on shore, only, where it is possible, it is 

 made simpler. On the sea everything is damp, and with a 

 dampness caused by the particles of salt water carried by the 

 wind ; these particles of water are conductors, and provide a 



* In picking up a cable, the counter, or tell-tale, sometimes indicates a 

 greater length than that really taken in, the difference being caused by the 

 elasticity of the line, which is taken in under tension. As an instance of 

 this, a hempen rope with a grapnel let out over the drum of a break, appeared 

 by the tell-tale, 3,000 metres, but on being recovered, the index indicated a 

 length of 3,100 metres, notwithstanding the shrinking of the rope in the 

 water. Allowance for this has, therefore, always to be made when drawing 

 in a line. 



