494 Elevation of Water ly its lateral Ailion. 



But before it quite reaches the tops of the tubes O P Q, it will begin to defcend in a full 

 ftream through the lo weft rarefying tube E, which is the fliorteft : the lateral action now ceafing 

 to rarefy the air, the water will rife no higher, but will remain at the fame elevation above 

 each of the cifterns, until the external valves (hall open. After this the valve m will clofe 

 the end of its tube, and prevent the diminution of the jet ; the air from without will enter 

 the loweft veflel D, through the tube G ; part of it will afcend from this veflel through 

 the empty branches O P Qj)f the rarefying tube into the veffels above ; and at the fame 

 time the raifed water will flow out of all the veflels into different cifterns. 



If all the veflels are connefted together by their air pipes, inftead of each being con- 

 nefted to the loweft, the fame equality of aftion will take place. In both cafes, as the 

 elevation of the water in each of the tubes muft correfpond with the degree of rarefadtion 

 in the air before it can begin to afcend, it is evident that the tubes will all be full at the 

 fame time, although they may contain columns of unequal height, when the air begins to 

 rarefy : the imperfeftion of one or two of the valves to the water tubes can therefore be 

 attended with little inconvenience. 



Jf the air pipes were feparately conne£led with the conical tube, fome of the veflels 

 might be filled with water before others : this inequahty, indeed, might be removed by 

 bringing all the tubes into one ; but in either cafe, if the external valves were to remain 

 too long clofe, the raifed water would defcend through the air pipes, and wafte rapidly 

 from feveral cifterns at once. The arrangement that has been recommended will be free 

 from this inconvenience, however long the valves may remain ftiut. 



By proportioning the widths of the water tubes and the capacities of the veflels upon 

 their tops, it will be eafy to raife a little more water into the firft ciftern, than will be 

 delivered into the fecond ; and a little more into the fecond, than will be carried into the 

 third ; and fo on, in order to prevent any of the lower cifterns from becoming empty of 

 raifed water. The wafte thus incurred will be very inconfiderable. If, however, more 

 water (hould be raifed into any of the higher cifterns than thofc below can long afford, it 

 it will be proper to make part of it defcend through tubes for the purpofe, and replenifli 

 the lower ones. 



As the jet will bear the admifllon of fome air without diminution, it is not neceflary 

 that the valve m fliould perfedlly clofe the end of the air tube : if it did, it would not be 

 eafy to open. When a cock is ufed inftead of this valve, it muft be fixed in the tube at 

 fome depth below the furface of the water in the refervoir, and a weight hung upon an 

 arm or lever proceeding from it, to keep open the paflage through the tube. This weight 

 will alfo ferve to counterbalance the empty veflel which . contains the fyphon, and opens 

 the valves. A cock can be attended with little inconvenience in this fituation : if, after 

 long continued ufe, a little water fliould be admitted, it will caufe no perceptible diminu- 

 tion of power. 



A fecond and third feries of elevations may be caufed by conneding more air pipes with 

 the conical tube ; but as the introdu£lion of fo many pipes into the tliroat would diminifli 



its 



