ELASTICITY OF AIR. 



51 



Fig. 6. 



The valve in the top of the barrel is in this case continually under the at- 

 mospheric pressure, and therefore the air confined in the pump can never be 

 driven through it, until it is condensed by the piston, so that its force shall be 

 greater than that of the atmosphere. From the causes already explained, 

 arising from inaccuracy of mechanical construction, some small space must in- 

 evitably remain between the piston and the top of the barrel, even when the 

 piston is drawn upward as far as possible. This small space will contain con- 

 densed air, and the valve at C will cease to act when the air which occupies 

 this space exceeds the atmospheric pressure by a force less than the tension 

 of the valve. 



When the piston is pressed to the bottom, a small space will likewise re- 

 main between the piston and the bottom, which will be occupied by air, but at 

 each ascent of the piston this air expands, and is subject to constant diminu- 

 tion as the working of the pump is continued. The principal source of imper- 

 fection in such an instrument, independently of that which arises from me- 

 chanical inaccuracy of its construction, depends on the tension of the valve in 

 the top, and the pressure of the atmosphere upon it. To diminish this imper- 

 fection, the valve in the top is sometimes made to communicate by a pipe with 

 a small subsidiary exhausting-syringe, by which the pressure of the atmosphere 

 on the valve may be partially withdrawn, so that a less force acting under the 

 valve may open it. 



A perspective view of an air-pump, with all its accompaniments, constructed 

 upon this principle, is exhibited in fig. 7, where the several parts of the ma- 

 chine are marked with the same letters as the corresponding part in the sec- 

 tional diagram, fig. 4. The subsidiary syringe just alluded to, is also repre- 

 sented at G. It is worked by a handle, H. 



EXPERIMENTS WITH THE AIR-PUMP. 



The pressure and elasticity of air are capable of being strikingly illustrated 

 in various ways by experiments with the air-pump. 



If a glass receiver, open at both ends, have a strong bladder tied upon o: 

 end so as to be air-tight, and be placed upon the open end on the plate of a 

 pump, when the air is exhausted from the receiver, the pressure of the extern 

 atmosphere on the bladder will immediately cause its upper surface to b< 



