302 Mr. Uerries on an improved Air-pump, [Oct. 



of arts in this city, of which I am a student, some alterations 

 have occurred to me as likely to render the instrument more 

 perfect, and as these have been approved of by some persons 

 of science to whom I communicated them, and who have 

 informed me that they believe them to be new, I take the liberty 

 of sending the annexed description and drawing for the purpose 

 ef laying them before the pubHc, through the medium of the 

 Annals, I am, Sir, your obedient servant, 



is Joseph Herries. 



Description of Phil e XXXVII, see^g. 1. 



A A is the barrel of the pump ; B B two thick metalhc plates 

 screwed to the flanges C C C C ; D is a pipe leading from the 

 receiver, and communicating with each end of the barrel, 

 through a small hole bored in the plates B B; E is a solid piston 

 working through the stuffing box F ; G is a rod working air^ 

 tight through the piston in a collar of leathers. On each end 

 of this rod there is a conical valve H and H', ground into the 

 openings of the pipe D, having a small degree of play, so thiit 

 both valves cannot be shut at once. These valves are guided 

 by a continuation of the rod working in the openings of the 

 plates BB, in the side of which there is a small groove for the 

 admission of air, shown by the dotted lines. I and V are coni- 

 cal valves opening outwards, and working in a socket in the 

 screw nuts K K, r being supported by a spiral spring ; from 

 these valves there are openings which communicate with the 

 pipe L. 



Suppose now that the piston is at the bottom of the barrel, 

 and the three valves I V and H' shut. If the piston is drawn 

 upwards, the friction of the leathers on the rod G will carry it 

 along with it, and shut the valve H ; H'' will then be opened, 

 and allow the air from the receiver to rush down the pipe D, 

 and fill up the vacuum formed below the piston, while the air 

 above the piston will be forced out at the valve I, which will 

 shut with its own weight. On moving the piston downward 

 the valve H' will instantly shut, and H be opened ; the air from 

 the receiver will rush in above the piston, and the included air 

 below it will be forced out at the valve V, and escape through 

 the tube L, (where it may be advantageously employed for 

 condensation,) the spring will then shut the valve. Thus by 

 working the pump a continued stream of air will be thrown out 

 from the receiver, until the exhaustion is completed. 



It is obvious, however, . that as the whole pressure of the 

 atmosphere is sustained by the valves I and V, the air con- 

 tained in the barrel will not effect its escape until, by compres- 

 sion with the piston, its density is superior to the external air, 

 and should a small stratum of this air remain in the barrel 



