Motion of a Current and ike Pr^ffure ^the Air. » fjj^ 



refervoir nbove the fmaller aperture of the conical tube being fix inches and a half, and 

 the water which ran through that tube completely filling its wider extremity ; the crooked 

 end of the tube E was introduced into the fmaller aperture of the conical tube, while the 

 bottom of C was two inches below the furface of the water in the refervoir j part of the 

 water immediately rofe fourteen inches in C and filled D ; the tubes C and E were then 

 lifted out of the water, and D was emptied by taking out the Hoppers. The experiment 

 was frequently repeated. 



Experiment 2. The veflel D was made fpherical, two inches diameter, the fame tubes 

 C and E adapted as before. The two tubes F and G were added to this, and clofed by the 

 finger of the operator when water was raifed. The tube E being lengthened a little, was 

 placed within the fmall aperture of B, when the charge of the refervoir was feven inches. 

 The bottom of the tube C was juft below the furface of the water. In a very fhort time 

 part of the water rofe near fixteen inches through C, and filled D i, the tubes jT and g were 

 then opened, and this water ran out. 



From feveral experiments, I think water will not rife more than fixteen. inches under 

 fimilar cirtumftances. 



Upon the admiffion of air into D, the dream of the jet diminifhed fo much, that it did 

 not half fill the wider aperture, and it was long after the tubes F and G were clofed before it 

 could recover its full force. This (hews the neceflity of a cock or valve to clqfe E, t;o 

 prevent the diminution of the jet, when air is admitted into D, if we wi(h to make the 

 engine raife much water, and empty itfelf at ftated times. For no water will rife fo long 

 as the conical tube is not full. If a cock be perfe£Hy air tight, it will anfwer the purpofe 

 very well. 



Fig. 2 points out a method of conftru£ting a valve for this purpofe : the veflel D, in this 

 figure, is cylindrical, the fmall tube g opens very near the top, under that letter, and the 

 dotted line « pafles through the middle of F. The tube C and E are explained before, m 

 is a valve which, by its weight, clofes the upper aperture of E, when the lever L opens the 

 tubes F and g-, and which immediately before it clofes thcfe pipes it raifes again: the com- 

 munication may be made by a piece of ftrong wire fixed into m, and which, palling 

 through f, will reft upon the lower valve or lever L. This wire is reprefented by the 

 dotted line n. 



It perhaps might be advantageous to have a valve in C to fupport the column of water 

 in that tube alfo. 



The power of the engine feems to be very confiderable when CQnflru<3:ed upon the 

 fimple plan mentioned in the experiments ; it feems very defirable to determine the beft 

 form of the conical tube, and the method to throw mod force upon the rarifying tube, fo 

 as to produce the greateft effcft under a given preflure of water in the refervoir. Citizen 

 Venturi has determined the form of a tube to produce the greateft expenditure*, and it 

 feems likely that it may alfo be the beft form for this purpofe. 



• Philof. Journal, II. 174. This figure is adapted to a certain definite pre{ru;:e.only,.and rauft be varied 

 for other depths. — N, 



The 



