304 Theory of the Pneumatic Paradox. 



the orifice with those which are moving towards it," becomes con- 

 tracted, and theu again enlarged ; so that at the distance from the 

 orifice equal to its semidiameter, where the greatest contraction 

 takes place, the diameter of the stream is about eight tenths, and 

 of course its area about two thirds of that of the orifice. Sir 

 Isaac Newton gave the contracted part of the jet the name of 

 vena contracta. The same phenomenon occurs when a cylin- 

 drical tube is adapted to the orifice. It will be perceived, there- 

 fore, that in the experiments of Venturi just described, the water 

 did not entirely fill the tube at the part where the glass tube was 

 inserted. Now flowing liquids in contact with air in a state of 

 rest, carry along with them a portion of the contiguous air, and 

 this efl"ect taking place within tlie tube at the place of the vena 

 contracta, the air in the glass tube becomes rarefied, and its elas- 

 ticity being thereby diminished, is no longer sufficient to resist 

 the atmospheric pressure upon the colored water in the vessel 

 T, which consequently rises. Venturi, well aware, without 

 doubt, though I do not recollect that he states the fact, that the 

 ascent of the colored water was dependent upon the vena con- 

 tracta, describes no attempt to obtain a similar result by inserting 

 the glass tube at any other part of the tube KLV, and he was feir 

 from applying to the whole tube, like some recent writers, a con- 

 clusion true of only a small portion of it. 



The only, direct evidence, as it respects Hquids, of the proposi- 

 tion of Mr. Spencer, that has come to my knowledge, is a single 

 experiment of Bossut, described in his work entitled " Traite 

 Theorique et Experimental d' Hydrody- 

 namique," the first edition of which was 

 published in the -year 1771. The follow- 

 ing translation from the edition of 1796, 

 comprises Bossut's description of the ex- 

 periment, and his remarks thereupon. 

 Let the cylindrical horizontal tube EN 

 be adapted to the reservoir ABCD ; and 

 let us suppose this reservoir to be kept 

 constantly full to the height AB, and the 

 water to flow freely in the tube without 

 meeting with any resistance. It is certain 

 that, if we except the pressure which re- 

 sults from the weight of the column of water EN, the tube expe- 



