164 History of ' Mechanical Inventions and 



stuffing-box, and also occasioning a necessary friction for the 

 hollow tube. The conical ends of this hollow tube are re- 

 ceived into hollow cones in the bottom plate and top of the cy- 

 linder, (in which there is of course a conical valve and a stuf- 

 fing-box,) and the conical opening in the bottom plate is 

 merely an expansion of the aperture of the tube from the re- 

 ceiver. The length of the hollow tube is a small fraction of 

 an inch less than that of the cylinder. In the figure the pis- 

 ton is rising. The receiver and air in the hollow tube are 

 rushing in below the piston and getting equally rarified, while 

 the air above the piston is getting expelled by the jumping- 

 valve in the cover and no otherwise ; and when the piston is 

 fairly at the top, not a bubble of air of any density ought to 

 remain if the valve be properly made. When the piston de- 

 scends, its first effect is to carry down the hollow tube by a 

 fraction of an inch, till it is prevented from being carried far- 

 ther by pressing on its negative cone in the bottom plate. This 

 movement closes the connection of the region below the piston 

 with the receiver, and the air escapes by the j umping-valve 

 of the piston, and no otherwise ; and when the piston reaches 

 the bottom, not one bubble of air ought to remain lodged any- 

 where. But all the while that the piston has been descending, 

 the air has been rushing in from the receiver through the hol- 

 low tube in the piston, above the piston, and so on. These 

 rods running through the collars of the piston work so well, 

 that in an air-pump which had no advantage besides by the 

 use of them, an ingenious friend of mine, since dead, preferred 

 working his valves in this way ; and I recollect seeing in the 

 Annals of Philosophy, some years ago, an air-pump proposed, 

 in which no fewer than four such solid rods were introduced 

 into two barrels to do the work here effected by one hollow 

 one. I have besides seen one internal one, and one external 

 hollow one proposed, and two internal solid ones, and have 

 imagined all sort of things before arriving at that which is so 

 provokingly simple compared with the others, that I had al- 

 most rather not have discovered it, than have had to accuse 

 myself so much for stupidity in not discovering it long ago. 

 I have thought over it occasionally for several months, and 

 cannot simplify it ; and having shown the design to some in- 



