THE TELEGRAPH. 
160 
From the beginning to the end of the operations of construct¬ 
ing and laying a cable, and even after it is laid, electrical tests 
(of which we shall speak further on) are continuously applied. ^ 
The cable is conveyed on board by means of pulleys fixed at 
intervals on a rope stretching from the shore to the ship. The 
ships are usually supplied with flat-bottomed iron tanks, 
arranged along the centre and sides. The bottom plates of 
these are thicker than those of the sides, and the thickness of 
the latter diminishes from the bottom upwards. A hollow cone 
occupies the middle of the tank, and round this the cable is 
stowed in level layers. Any risk of entanglement between the 
layers is avoided by the use of the “ crinoline.” This appa¬ 
ratus consists (see fig. 103) of a series of concentric circles 
connected by ten or twelve radii, which compel the cable to 
arrange itself in regular layers when it is stowed in the tank. 
From this the cable is conducted by large cast-iron pipes to the 
paying-out machine. In its course the cable passes through 
wooden troughs in order to prevent the heat of the sun from 
melting the gutta-percha. In these troughs the cable passes 
between rollers which keep it properly stretched before it 
reaches the brake. And, seeing that the cable might break on 
board the ship, as indeed happened on the Great Eastern with 
the Atlantic cable of 1865, there were added, in 1866, six wheels 
between the tanks and the brake, and these wheels served as so 
many brakes that could press the cable between two grooved 
pulleys, so as to stop it and prevent it going overboard if it 
should part before reaching the paying-out machine. 
The paying-out machinery through which the cable passes 
before it reaches the pulley at the stern, whence it drops into 
the sea, affords the means of maintaining a uniform tension 
on the cable, and of regulating its immersion according to the 
depth, so that it may not be exposed to a strain beyond that at 
which its elongation would endanger the core. 
This amount of strain is nearly equal to the weight of a 
length of the cable reaching vertically from the stern pulley to 
the bottom of the sea, diminished by the weight of the water 
displaced by the cable. 
The force applied to the brakes to prevent the too rapid 
descent of the cable must therefore be only a little less than 
the weight of that length of cable that w T ould reach from the 
