Mr. R. Gill on an Air-pump without Clearance. 259 



level of the oil under the piston to lower, and consequently the 

 valve is left uncovered ; the spring is so arranged that, as soon 

 as this happens, it will have expanded so much as to be incapable 

 of resisting the upward effort caused by the friction of the piston 

 q against the principal piston-rod, and the valve v is therefore 

 carried upwards and opened. The apparatus is now in the posi- 

 tion shown in the figure : the space below the piston was com- 

 pletely full of oil, and the joints through which air might pene- 

 trate are all covered with this liquid ; consequently the space left 

 by the motion of the piston must remain perfectly empty, at 

 least as regards air, as it is possible that even fixed oils might 

 emit vapour of exceedingly low tension. The vacuous space 

 below the piston being now in communication with the receiver 

 through the tube g, becomes filled with air more or less rarefied ; 

 at the same time the air contained in the upper part of the cy- 

 linder being compressed by the ascent of the piston, raises the 

 valve oc, and escapes into the atmosphere through the aperture e. 

 As the pressure above the piston is greater than that below it, a 

 small leakage of oil takes place downwards through the interstice 

 round the little piston-rod z, and dropping upon the shield 

 attached to the valve v, falls into the annular space b b, and 

 there accumulates during the ascent of the piston. The inter- 

 stice for leakage allows only a small quantity of oil to pass ; other- 

 wise the valve v might be overflowed before the piston begins 

 to descend, and consequently before the valve was closed. It 

 will be seen that in every part of the piston's motion its lower 

 edge is immersed in the oil, which prevents any lodgment of 

 air between it and the sides of the cylinder — a circumstance 

 which it is next to impossible to obtain in common air-pumps, 

 as the piston cannot be in perfect contact with the cylinder ; and 

 it is evident that any such lodgment of air must prevent the 

 total exhaustion of the receiver, as it will maintain in the cylinder 

 a certain pressure, below which that in the receiver cannot fall. 



At the commencement of the downward motion of the piston, the 

 valve v is closed immediately by its piston q, and the valve x is 

 closed by its own weight and the atmospheric pressure; the 

 space below the piston becomes smaller and smaller as the pis- 

 ton descends; at the beginning of the upward stroke all *the 

 space below the piston was full of oil, and during the up-stroke 

 more oil has passed down into the annular space; the conse- 

 quence is, that when the piston reaches the bottom of its stroke, 

 the space below being full, and in fact more than full, of oil, it is 

 evident that the air extracted from the receiver must be com- 

 pletely expelled through the apertures dd, together with that 

 small quantity of oil which passed downwards through the in- 

 terstice around the rod z during the ascent of the piston. The 



S2 



