CHEAP AND flMPLE AIR PUMP. 25J 



piston does; and as the piston continues to descend, the 

 valve d ^viil be opene^l, by the pressure of the wire// against 

 the >}oitorn of the barrel, and the air in the receiver will ex- 

 pand freely into the barrel, ♦^^hrough the passages g ^, g g^ 

 as rep'-efeievited in tig. 2. This air wil! also be expelled by 

 the ascent of the piston ; and th\s operation maybe repeated 

 as often as is thoui;ht necei-siiry. As the valves will not be 

 exnosed to much frictiom they will probably keep in order a 

 long time; and if the pump is well made, there seems to be 

 no impediment tc its producing a nerfeot exhaustion*. But 

 whatever care the workmen may take, it is probable, that 

 a rairnUe quantity of air will remain, and that in a state of 

 considerable condensation (under the collar of leathers) 

 between the pitjton-rod, and the top of the barrel; and also 

 between the sides of the barrel, and the upper part of the 

 piston : but this is a fault, to which every pump, I believe, is 

 liable. It niay also be objected, that, when the valve d is 

 opened, ty the pressure of the wire//*, against the bottom of 

 the barrel, the air, by rushing suddenly out of the receiver, 

 will occasion oscillations of the mercury in the gage: these 

 oscillations, however, must be very trifling, not exceeding 

 ^ of an inch, as the valve d would open, were the pressure 

 greater than that of \ of an inch of mercury. All these 

 imperfections however, 1 hope, are completely obviated, by 

 the following construction. — 



At figs. 3 and 4, A B, as before, represent the barrel, C C 2d constmc- 

 the piston-rod, D the piston ; a a the collar of leathers, b h ^'°^** 

 the cup to hold oil, c the expelling valve of brass, d the 

 valve in the piston of brass also, and g g, gg the passages 



• It is obviously impossible, that the whole of the air in the receiver 

 should be taken out in this way, so as to form a perfect exhaustion, pro- 

 vided we admit the workmanship carried to the highest degree of accu- 

 racy, so as not to have the slightest imperfection. Suppose the barrel 

 and receiver to be of equal capaciMes. The first time the piston is drawn 

 up, the air in the receiver will expand to double its former volume, or 

 be of half 'heden-ity it was before. After the second stroke of the 

 pist«n, it will in the same manner be of half i4ie density it was after 

 the first. And thus the rarefaction will go on in a series decreasing in 

 geometrical progression, which can never reach 0, as the last term must 

 fcc half thf value of that preceding, C. 



by 



