384 Mr Dunn'*s Descriptimi of an Improved Air-Pump. 



the valves VV^ When it can no longer effect this, the exhaus- 

 tion must cease, and,, consequently, a near approximation to a 

 vacuum cannot be obtained by means of this construction. 



The best method hitherto proposed for effecting a more per- 

 fect exhaustion, is that of Cuthbertson, which proceeds upon 

 the principle of opening the valves independently of the elasti- 

 city of the air ; and, accordingly, he substitutes in the place of 

 the bladder-valves VV, Fig. 1., the metallic ones VV Fig. 2. *, 

 having the wires WW^ attached to them, which slip stiffly in 

 stuffing-boxes in the piston-rods. On raising either piston, the 

 valve V or V is opened by the friction of its wire in the stuffing- 

 box, and is shut by its depression ; in the former case leaving a 

 free communication betwixt the part of the barrel under the pis- 

 ton, and in the latter case cutting it off. In the pistons he also 

 places metallic valves PP', to be opened by the descent of the 

 piston-rod, and to be shut by its ascent, the valves in the pistons 

 thus opening and continuing open, while those at the bottom of 

 the barrels are shut, and vice versa. Now, as the piston-valves 

 are opening while those at the bottom of the barrels are shut- 

 ting, Cuthbertson found it necessary (though this is not required 

 in the common pump) to exclude the external air from the 

 barrels. 



For this purpose he put air-tight covers CC over the barrels, 

 and made the piston-rods move in the air-tight stuffing boxes 

 BB', and placed metallic valves MM' in the covers for the egress 

 of the air, to be opened either by its elasticity or by the pistons 

 striking against the projecting points pp of these valves. 

 . It is also necessary to prevent the return of the air into the 

 pump during the shutting of these valves, which is done by ha- 

 ving them immersed in oil. 



This construction is certainly, in theory, as near perfection as 

 we are likely to reach by any form of pump, but it is as certainly 

 very complex, and, consequently, very expensive, and liable to 

 go out of repair, — an objection of which those who have been 

 engaged in making these pumps best know the force. 



Believing the only useful part of Cuthbertson's invention to be 

 the contrivance for opening the valves at the bottoms of the bar- 

 rels mechanically, I was of opinion a pump would perform near- 

 " In figures 2d and 3d only one of the barrels are represented. 



