^8 Nenx) and ufejul Application of tht Syblmi. 



Two fmall demi-cylindrlcal tubes when con(lru£led were joined together in fuch a 

 manner, that one piece ferved for the flat fides of both. They were foldered into the fid» 

 of a fyphon at the middle of the turn, or that part which the water pafles through at its 

 higheft elevation." The flat partition between them was thin, and projefted a little into 

 the infide of the fyphon, but their fcmi-circCilar fides had no projeftion at all. When the 

 branches of the fyphon were held horizontally, the pofition of the two fmall tubes was 

 vertical, and the plane of the partition was parallel to a line drawn between the branches. 

 The fmall cylinder formed by both tubes was then encircled with thread to a proper thick- 

 nefs, and inferted into the neck of an ounce glafs vial. 



Every part of this apparatus being perfedly air tight, the fyphon was filled with water, 

 by holding it in nearly an horizontal pofition below the furface of water in a capacious 

 veflel : both of its apertures were then clofed by the thumbs of the operator, and the 

 branches were 'fet in a vertical pofition : the end of one branch being immerfed in water, 

 its. aperture , was opened, and the efflux through the tube was permitted immediately 

 after, by opening the aperture of the other branch. 



When the ftream had a defcent in the tube equal to one inch below the furface of the 

 water in the veflel) bubbles of air were eje£led, and in a fliort time the bottle was filled 

 with water to a level with the upper fides of the two fmall tubes} a continued circulation 

 was then viCble. The water entered the bottle through the tube on that fide of the par- 

 tition which was ftruck by the ftream, and palTed out through the tube on the other fide, 

 with a velocity proportionate to the adtive weight of the defcending column, which was 

 encreafed or diminiftied at pleafure, by lowering or raifing the fyphon. When the efFeftive 

 fpace of defcent was lefs than one inch, no air was extradled out of the bottle. 



In this experiment the length of each branch of the fyphon was feven inches : the 

 diameter of the tube was all throughout equal to about half an inch. 



The length of one branch of the fyphon being encreafed to forty- five inches, and that of 

 the other to forty-three, the ounce bottle was changed for one of eight ounces, and the 

 fyphon was filled with water in an inverted vertical pofition. After its two apertures were 

 refpeQively clofed with corks, it was laid horiaontally, and the bottle was detached from 

 it, to let out a little water which had been forced into it, atid alfo with an intent to leflen 

 the compreflion of the air it contained. After replacing the bottle, the fyphon was fet 

 into its proper vertical pofition, <vith its ftiorter branch plunged in water, and the efflux, 

 was permitted by drawing the corks. When the (horter column was equal to forty-one 

 inches and aa half, it was requifite to fettle the fyphon a few inches immediately at- 

 the commencement of its operation, to prevent the longer column from wafting^fo much 

 as to let the {horter preponderate. When the ftream begun to extract the air, the fyphon, 

 was raifed to its former elevation, or even higher, and. water continued to enter the bot- 

 tle, and air to be carried down the defcending column, until the fiiotter branch could no 

 longer receive a fupply of water. The aftivc length of the defcending column was then 

 no longer than two inches and an half. 



It. 



