303 Theory of the Pneumatic Paradox. 



tected from the impulse of the blast, will be sustained against its 

 gravity by the afflux of air against its under side. 



Mr. Spencer's very ingenious explanation is founded on the 

 principle, to use his own words, " that currents of fluid, whether 

 elastic or non-elastic, exert no force but in the direction in which 

 they move ; the latter is fully proved by forcing air or water 

 through a cylindrical tube ; if holes be made through the sides of 

 the tube, none of the fluid will escape." A late writer, in a Lon- 

 don scientific journal, expresses the same principle thus : " It is a 

 well known property of fluids that they transmit their pressure 

 equally in all directions ; but this law applies only to fluids in a 

 state of rest. When they are in motion, they are subject to the 

 laws which regulate the motions of solids, and do not transmit 

 any lateral pressure, except where they meet obstacles to their 

 onward motion." Hence it is inferred, that the currents which 

 radiate from the centre of the disks, exert no pressure against their 

 internal surfaces, and that, as the impact of the blast is not suffi- 

 cient to overcome the exterior atmospheric pressure, the movable 

 disk consequently adheres. 



This inference seems necessarily to follow, if we admit the cor- 

 rectness of the alledged principle of hydraulics from which it is 

 deduced. The only proofs of the principle, as applied to liquids, 

 I have seen cited — not by Mr. Spencer, who cites none — but by 

 others, are the experiments of Bossut and Venturi, both philoso- 

 phers preeminently distinguished for their experimental researches 

 and discoveries in hydrodynamics. It may however be asserted 

 without fear of contradiction, that whoever carefully reads Ven- 

 turi's original work " on the Lateral Communication of Motion 

 in Fluids," cannot fail to perceive not only that he nowhere ad- 

 vances such a general proposition, but that it is irreconcilable 

 with his reasonings in several parts of the work. He describes 

 but two or three experiments with cylindrical tubes that lend 

 any semblance of support to it, and it is hardly necessary to say, 

 that the very novel and interesting results he obtained with tubes 

 of a difiierent form, or with long, descending, cylindrical tubes, 

 are so inseparably connected with the peculiar form, or the posi- 

 tion of the tubes employed, as to furnish no evidence of the gene- 

 ral principle alledged by Mr. Spencer. 



