PNEUMATIC TELEGRAPHY. 
59 
As the pneumatic tubes go to the upper stories, the pres¬ 
sure of the column of water is added to that of the compressed 
air, and this method is invariably efficacious. 
We have said that the signals required for working the 
tubes are made by electric bells. There is nothing remarkable 
about these bells, except that, instead of a continuous tingling, 
they give only a single blow of the hammer each time a current 
is sent. At every repetition of the signal, a disc falls behind 
an opening in the case, and remains in that position until it is 
raised up again by hand. 
Keservoirs made of iron plates are used for containing com¬ 
pressed or rarefied air, so that no change of pressure is per¬ 
ceptible when a carrier is sent. 
For compressed air, the dimensions of these reservoirs are 
as follows :—1. A cylindrical reservoir, 8 feet 6 inches dia¬ 
meter, 11 feet 2 inches high, holding 630 cubic feet. 2. A 
reservoir of the same shape and section, but only 9 cubic feet 
high ; its capacity is 440 cubic feet. 
The total capacity is 1,070 cubic feet. 
These reservoirs are fixed and are provided with a stop-cock 
at the bottom, for emptying out the water resulting from the 
compression. This is done every night. 
The reservoirs for rarefied air are also two in number : one 
has a capacity for 630 cubic feet, and the other of 160 cubic feet, 
the total being 790 cubic feet. 
The steam-engine first erected for working the tubes of 
Mr. Clark’s system is a horizontal expansion engine, of 28 
horse-power. The diameter of the cylinder is 16| inches, and 
the stroke 2 feet 6 inches, the expansion begins at two-thirds 
of the stroke, and the maximum pressure is 40 lbs. per square 
inch, the pressure in the condenser being 7 lbs. per square 
inch, and the number of revolutions 48 per minute. The 
theoretical horse-power is 40, but the indicated horse-power is 
only 25*20. 
The blowing cylinder is in a line with the steam cjdinder, 
and is double acting, with ordinary valves. The air is, there¬ 
fore, compressed on one side of the piston and rarefied on the 
other. This simplifies the mechanism, as only one cylinder is 
required ; but the useful effect is diminished. For when the 
piston returns, after having reached the end of its stroke, the 
