WATER. 



prevent the water returning into the bellows after it has 

 been driven out by clofing the bellows. This kind of 

 pump has been frequently propofed, and the advantages of 

 difpenfing with barrels and piftons loudly infilled upon ; but 

 the refiftance of the leather in folding, and the lofs of water 

 by leakage^ and above all the want of durability, will always 

 prevent the adoption of fuch pumps. 



The Pump with a Diaphragm of Leather, •which does tiot 

 Jllde In the Barrel This is very nearly allied to the bellows- 

 pump. The belt form for conilrufting it is fully defcribed 

 in our article Ship'j Pump, where the invention is attributed 

 to Benjamin Martin ; but we find the fame thing was long 

 before applied by Meflrs. Goflet and De la Deuille, in 

 France. (See Belidor's Arch. Hydraulique, vol. ii. p. 120. ) 

 This is a good pump, but is not durable, becaufe the con- 

 ftant ftrain on the leather will caufe it to crack. 



Sucking-Pump, which gives out a continual Stream — Mr. 

 Smeaton applied the following fimple and etfeftual expe- 

 dient to make a fingle fucking-pump deliver the water 

 equally in the defcent of the bucket as in its afcent. Tiie 

 pump-rod was enlarged, by furrounding it with a cylinder of 

 wood at the part where it rofe above the furface of the 

 water contained in the cillern at the top of the pump. This 

 cyhnder of wood was of fuch diameter, that its feftion 

 was equal to half the area of the pump-barrel at the 

 place where the bucket worked. When the bucket was 

 drawn up, and raifed water into the ciftern at top of the 

 pump, the vv'ood cylinder, which was attached to the pump- 

 rod, rofe up out of the water in the ciitern, and thereby 

 made place in the ciftern for one-half of the water which was 

 brought up by the bucket, and in confequence only one-half 

 of the water ran out at the fpout of the ciftern ; but when the 

 bucket moved downwards, in order to fetch another ftroke, 

 this cylinder of wood difplaced from the ciftern half as much 

 water as the pump brought up in the former inftance, and 

 confequently an equal quantity of water was given out at 

 the fpout in either cafe. 



If the pump is worked by the force of a man working a 

 fimple lever, then he will make the down-ftroke of the 

 bucket in Icfs time than the up-ftroke, and in this cafe the 

 area of the cylinder fnould be made lefs than half the area of 

 the barrel of the pump. It muft be obferved, that this 

 contrivance is only a remedy for the unequal efflux of water 

 from the fucking-pump, and that the power required to 

 work the pump is ftiU left unequal in the up-ftroke and 

 down-ftroke, becaufe it is only in raifing up the bucket that 

 the water is drawn from the well below ; and that water 

 which runs out at the fpout when the bucket dcfcends, is 

 drawn from the ciftern at the top of the pump, and not from 

 the well. When the pump is worked by a man with a 

 lever, this inequality of the refiftance is advantageous, be- 

 caufe a man can exert his force moft conveniently when he 

 deprefles the end of the lever to draw up the bucket ; alfo, 

 in a fingle-afting fteam-engine, the principal power is 

 exerted to draw up the bucket. 



In machines worked by wind, water, or horfes, the moving 

 force is uniform, and the refiftance muft, by fome means, be 

 made uniform alio, or the machine will move by fuddcn ftarts. 

 A fufficitnt weight maybe applied to the oppofitecr.d of the 

 lever to counterbalance one-lialf of the force neceffary to draw 

 up the bucket ; this weight will tend to diminilh the force of 

 drawing up the bucket, and v.-}ien the bucket defcends, and 

 the machine would otherwile have nothing to do, it will 

 have to raife up the weight ready to aid it in the fucceed- 

 ing llroke. Or a fly-wheel may be applied : but a ftill 

 better method is to employ two pumps to aft alternatelv, by 

 which means the refiftance is continual, and the effiux of 



water alfo. When two fucking-pumps are employed, they 

 may be combined together, by making them both draw 

 from a common fu6tion-pipe, and both may be made to lift 

 the water into the fame ciftern. Or two or three force- 

 pumps may be combined together, as is defcribed in the 

 article Pump, in order to produce a continuous ftream. 



Alr-Vejfel for equalizing the Df charge of IV ate r from 

 Pumps. — This is the moft perfedl contrivance for effeding 

 that purpofe. It is a clofe vefTel of any figure, which will 

 contain air, and is made to communicate with the pipe 

 which conveys the water away from the pump. This com- 

 munication muft be made at the lower part of the air-veflel, 

 fo that the water will have free ingrefs and egrefs from it. 

 The air in this veftel will be comprefled into a fmaller fpace, 

 in proportion to the column of water which the pump has 

 to raife ; and by its clafticity endeavouring continually to re- 

 gain its former fpace, it will aft as a fpring to equalize all 

 fudden motions of the water through the pipe ; for in any 

 pump which afts by a ban-el and pifton, the water will be 

 propelled by ftarts ; and even if two or three barrels are 

 combined together fo as to produce a continual efflux of 

 water, fuch efflux will not be perfeftly equal during all the 

 periods of the motion. 



The evil of this may appear trifling and fo it would be 

 merely with refpeft to the difcharge of the water ; but it muft 

 be confidered that the mafs of water contained in a long , 

 pipe is very great, and that it requires a very confiderable 

 force to put this mafs in motion with that velocity with 

 which it muft flow through the pipe. Now if the 

 operation of a pump is by ftarts, the mafs of water in 

 the main-pipe will remain at reft, preffing on the valve 

 during the time that the pifton is withdrawn from the bot- 

 tom of the working barrel. In this cale, the force necef- 

 fary to put the water in motion muft be expended at every 

 ftroke, becaufe if the column comes to reft only for an 

 inftant, it muft be put in motion again bofore the operation 

 can be refumed : this is a heavy additional load upon the 

 firft mover, and has another more ferious evil in ftraining the 

 pipe and all parts of the machinenr ; becaufe the column of 

 water in the pipe, after it flops, runs back for a fmall 

 fpace until the valve (huts ; and it makes juft as great a con- 

 cufDon or fhock when its motion is fuddenly (lopped by the 

 fliutting of the valve, as any other fohd body would do 

 which was of the fame weight, and moved with the fame 

 velocity. In large ftcam-engines, the fliock occafioned by 

 the fliutting of the valve is exceedingly violent, unlefs an 

 air-veflel is applied. In that cafe, if the pump urges the 

 water with a fudden motion, the air in the vefle! will yield, and 

 admit the water into the veflTel in far lefs time than the whole 

 column of water could be urged into motion ; but as tlie air 

 will become comprefled by more force than the column of 

 water in the pipe, the elafticity of the air will force the 

 water from the veflel and up the pipe with a regular mo- 

 tion, and this wiU continue until the air has regained fo 

 much fpace that its elafticity is only juft fufficient to ba- 

 lance the column of water in the pipe. 



The air-veflel fliould be placed as near the pump as pof- 

 fible, that it may produce an equable motion of the water in 

 the whole length of pipe. The air-veflel is of confiderable 

 advantage when a column of water of great length is to be 

 raifed by a fingle-afting pump. If the pifton of the pump 

 at one end of the pipe is put at once into motion, even with 

 a moderate velocity, the ftrain on the pipe would be very 

 great before the column of water could be put in motion. 

 But the air-veflel tends to make the motion along the main- 

 pipe lefs defultory, and therefore diminifties thofe ftrains 

 which would really take place in the pipe. It afts 

 2 l^ke 



