308 Theory of the Pneumatic Paradox. 
one line in diameter, and situated two and six inches respectively 
below the ajutage. To the top were soldered two leaden tubes, 
the larger six feet high and three quarters of an inch in diame- 
ter, and surmounted by a funnel; and the smaller somewhat 
shorter, and terminating above ina glass tube. A shelf of tin- 
ned sheet iron five or six inches wide, was soldered to the oppo- 
site sides of the vessel, so as to cross it two inches from the top, 
under the end of the larger tube. The accompany- Fig. 5. 
ing figure exhibits the form, though not the relative 
proportions, of the several parts of the apparatus. 
The whole was supported by frame-work, not seen 
in the figure. The end of the descending tube, and 
also the lateral holes, being stopped, and the vessel 
and tubes being kept full of water to the brim of the 
funnel, which was somewhat more than seven feet 
above the bottom of the vessel, on removing the oiled 
silk from either lateral hole, a horizontal jet issued, 
which reached the distance of seven and a half feet 
before falling to the brick pavement, situated two and 
a half feet below the bottom of the vessel. On un- 
stopping the lower end of the triblet tube, the water 
immediately sunk nine inches in the small glass tube below its 
level in the large tube, showing a diminution of the effective head 
to that amount. Meanwhile the jet continued to issue from the 
lateral holes, but it assumed, owing to the joint action of onward 
and lateral forces, instead of a horizontal direction as before, an 
oblique downward one, which would of course prevent its attain- 
ing the random due to the actual pressure against the surface of 
the tube, and it reached the pavement at the horizontal distance 
of alittle more than three feet. The distance became less by 
diminishing the head of water. Jets of nearly equal force issued 
from the lateral holes when a similar descending tube without an 
ajutage was used. : pee 
This result completes the proof of the proposition I have beet 
endeavoring to establish, that water, flowing through a cylindrical 
tube, whether in a horizontal, inclined, or—if short—in a verll- 
al position, exerts, contrary to the principle laid down by Bossut 
ers, a lateral pressure, varying with its incumbent head 
endent of its weight, against the interior surface of the 
f 
