VOL. XXXII.] PHILOSOPHICAL TRANSACTIONS 95 I 



forcing pipe a 1 a2: but when once the forcing pipe, which here was 46 feet 

 high, is full, before any mercury can enter into the receiver, and force any 

 water out at the top of the pipe al, the mercury between the piston and barrel 

 must rise up to qq, near 3-^ feet above the bottom of the receiver; and as it 

 continues to rise up to pp, the water is thrown out with a velocity proportion- 

 able to the height that the mercury is raised above the 14th part of the height 

 of the water. Now though the friction of solids is here avoided, it is plain 

 that the mercury must move from mm to qq without raising any water, and 

 that it can only force in going from qq to pp, and only suck in falling from oo 

 to mm : and unless the piston is stopped a little while when at lowest, the water 

 will not have time to run out: so likewise the piston must be stopped when 

 at highest, that the receiver may have time to fill. 



Mr. Haskins likewise proposed another way, represented in fig. 2; where 

 the same letters represent the same parts, only here the barrel Is moveable by 

 the two chains e1 e2; and instead of a solid piston, the hollow cylinder cl cc 

 is fixed, and the mercury moving up and down in the lower part of it, sucks 

 and forces the water through the elbow. The figure represents the engine 

 sucking, by means of the mercury hanging from oo to mm. In order to force, 

 before any water can be driven out, the mercury in the inner cylinder must 

 descend from oo to mm, and rise up to pp between that cylinder and the barrel; 

 so that here also a great deal of time is lost; besides the great quantity of mercury 

 used, which is very expensive; because as much mercury is moved every stroke 

 as the water raised. 



These difficulties very much puzzled Mr. Haskins, and quite discouraged 

 some other persons who had got the secret of the invention, and were setting 

 up against him. But when I had considered the matter a little, though I had 

 not time to contrive a machine for it, I told him, that a little mercury might 

 be made to raise a great quantity of water, and there should not be such a loss 

 of time as in his engines ; but that I would have him find it out, before I 

 assisted him farther. In a little time he found out the contrivance represented 

 in fig. 4 ; and afterwards that of fig. 3 ; which last was what I had thought of: 

 and both these were also found out by the late Mr. William Vreem, who was 

 an excellent mechanic. 



In fig. 3, the barrel is moved as in fig. 2, but the plug ddd taking up a great 

 deal of space, there is occasion for no more mercury than what will make a 

 concave cylinder or shell up to pp, between the barrel d1 d2, and the hanging 

 cylinder cl C2 cc, when the stroke is made for forcing; and a concave cylinder 

 between the plug and cl c2 cc, when the suction is made. 



In fig. 4, the barrel with a third cylinder dddd, instead of the plug of fig. 3, 



