3 Prof. Magnus on the Motion of Fluids. 



repeated this experiment in various ways, but have arrived at 

 different results from those described by A'euturi. 



3. It is knoNVU that a jet of water, from the point where it 

 leaves the orifice to the point where its contraction is a maximum, 

 exhibits itself as a solid continuous mass ; a number of success- 

 ive swellings then occur, and the fluid is afterwards divided into 

 smaller detached masses. The most relined test that can be 

 applied to determine whether the air is carried forward by such 

 a stream, is certainly furnished by the flame of a candle ; but 

 however near I brought the flame to the transparent portion of 

 the jet, I was unable to detect the slightest motion. This was 

 first observable when the flame approached the second or thu-d 

 of the swellings alluded to, and became stronger as it approached 

 the point where the water was broken into separate masses. 

 Here a violent motion was imparted to the flame, which, how- 

 ever, was in general soon extinguished by the spray cast upon it. 

 Such is the action when the motion of the jet is steady ; but if 

 it oscillates, which it frequently does, then the flame becomes 

 unsteady, and is carried forward by the water at an anterior 

 part of the jet. 



Were the air in consequence of its adhesion carried along by 

 the water, the action must be strongest where the velocity of the 

 water is gi-eatest. As this, however, is not the case, as, on the 

 contrary, the bending of the flame increases according as the 

 motion of the water becomes unsteady, there is reason to conjec- 

 ture, that if the unsteadiness of the jet were entirely to disappear 

 (which, however, is perhaps impossible), the motion of the flame 

 would vanish likewise. 



4. The other experiment upon which Venturi grounds his notion 

 is the following : — A cylindrical tube AC, Plate I. fig. 1, enters 

 horizontally a cistern BEFD, which is tilled with water to DB. 

 Opposite to this, and a little apart from the end C, is brought a 

 small rectangular canal SMBR, made of tin ; the vipper portion 

 SR is open, and the bottom rests upon the edge of the cistern at 

 B, the canal thus rising from M to B. If water be permitted to 

 flow with some velocity through the tube AC into the cistern, 

 the stream will ascend the canal MB, and will flow out at V. 

 At the same time a motion occurs in the water of the cistern 

 BDEF : this water enters the canal SB, and proceeds forward 

 with the stream AC, so that the surface sinks in a few seconds 

 to M, the lower edge of the canal. 



5. Some time ago, an ingenious modification of this expe- 

 riment was published by Prof. Feilitzsch*. He uses for this 

 purpose a tin vessel, EDFG, fig. 2, divided into two portions by 

 a partition HI. A cylindrical tin tube, ABC, 2^ inches diameter 



* PoggendorfF's Annalen, vol. Ixiii. p. 216. 



