THE PNEUMATIC DESPATCH. 463 
ingenious for all that. We have run gossiping on so far; 
we must now have recourse to our diagrams and drawings. 
And it will be well to say, none of these — saving the carriage 
- — are portraits, or drawn to scale. They are sketch-diagrams, 
designed expressly to convey, in the most intelligible manner, 
correct explanations of principles. Fig. 1 is the despatch- 
tube — short in our drawing, but almost as long as you please, 
practically. The usual rate of propulsion through it for 
lengths under five miles would be some twenty-five or thirty 
miles an hour ; forty miles have been got, perhaps very 
much more might be; but such power has not as yet been 
ordinarily applied. In longer distances the speed would ne- 
cessarily be diminished, and indeed will diminish propor- 
tionately with length, until ultimately it would practically all 
but cease, unless the exhaustion of the tube were permanently 
and most powerfully maintained. Both ends of the despatch- 
tube in our diagram will be seen to be closed by valves y v'. 
This may rather astonish our readers, who may naturally ask 
how air is to be blown through or drawn out of an hermeti- 
cally closed pipe. Well, it is only seemingly closed. There 
is a grating at G, and another grating at H. “ But why shut 
up the pipe at both ends with valves, and open it again with 
gratings ? ” The why 33 will presently be clearly enough 
perceived. Fig. 2 shows the fan-wheel (fig. 3) inclosed in its 
case. Fig. 4 is the same in section. When the fan- wheel (w) 
is whirled round, the air in the interspaces ( a c a) between 
the fans (a a) will fly outwards from the periphery at what- 
ever rate of motion that air has acquired, the instant an 
indraught from the exhaustion-tunnel E is let in between the 
fans at the centre of the fan- wheel (c c c), and an ex-current 
is permitted through the air-vent (Q Q'). Now let us set the 
fan-wheel at work. We have closed the valves T, S, Q, as also 
the valve of the air-chamber R, and the wheel is accumulating 
power in vacuo. We open the near terminal valve of the 
despatch-tube V, and run in the carriage D. We permit an 
out- draught from the fan- wheel case through the open air- 
chamber X, by turning the valve T (to t t). We allow an in- 
draught into the fan-wheel through the exhaust-pipe E, by 
turning the valve S (to s s ). Instantly the exhaustion of the 
despatch-tube begins. The atmosphere is sucked in at its 
open end F, and at the distant grating H, and with it the 
carriage D is drawn along towards G, until, as it passes the 
grating G, it moves a trigger, which closes the valves V, T, 
and S, and opens the valves R (to r r) and Q (to q q) . An 
in-draught of air now sets in through the feed-pipe E from 
the open air-chamber X, and acquiring rapid motion in the 
interspaces of the fan-wheel, is thrown oft* in a perfect gale 
2 i 2 
