256 Mr. J. H. Patten on a new Air Pump. [Oct. 



are a great source of difficulty, may be made larger and stronger, 

 and the apertures, of course, will be more accurately closed, 

 without at all affecting the degree of exhaustion. The vapour 

 arising from the oil necessarily used in all pneumatic instru- 

 ments, is in this completely excluded from the receiver, and the 

 vacuum in the exhauster being torricellian, that in the receiver 

 will approach as near to it as the elasticity of the air will permit. 

 The glass parts of the instruments can be obtained from any 

 glass house, and the barrel (which would be more elegant of 

 glass) can be made at any steam-engine or gun manufactory, 

 and a clock maker will be competent to construct the brass 

 work. The subjoined sketch, although not drawn by an adept 

 in the art, will, 1 hope, give you an idea of it. It represents a 

 vertical section of a table pump, supposed to be divided directly 

 through the centre, with one half of the wood work, to which it 

 is attached. 



It is a number of months since I first thought of it ; I then 

 had one constructed with a barrel of sheet brass, and the plate of 

 the pump of tinned iron ; it was very coarsely done, and the 

 exhauster was filled with linseed oil, but notwithstanding its 

 rouo'hness, it far exceeded my expectations. I have never yet 

 been able to get an iron barrel, as it cannot be procured here, 

 and numerous avocations have prevented its being obtained 

 elsewhere. 



Figs. 9 and 10 (PI. XXXII) correspond in their lettering. 



In fig. 9, A B, C D, E F, represent a vertical section of the 

 instrument, G is a barrel of cast iron or glass, screwed firmly to 

 the table EF, in it is the solid piston H moved by the rack 

 work I. K is a glass globe resting upon the table CD, of a 

 little less capacity than the barrel G with which it communi- 

 cates by the glass tubes L and M firmly cemented into the piece 

 N and into the bottom of the barrel G. To the top of the globe 

 K is cemented the thick cap O, through which are made two 

 apertures, into one of which is screwed the stop-cock P com- 

 municating with the plate of the pump R ; over the other aper- 

 ture rests the valve S opening into the atmosphere (the construc- 

 tion is seen in fig. 10). In the globe K is a stiff wire ascending 

 into the cock P a short distance, and on it is screwed the valve 

 T ; the other end descends into the tube L, and to it is attached 

 the wooden or cork ball U. We will now suppose the piston H 

 withdrawn, and the barrel G tilled with quicksilver; the tubes 

 L and M being open will be filled to the height of the dotted 

 line. Put the piston carefully in so that no air shall be between 

 it and the mercury. As the piston descends, the mercury rises, 

 and when it reaches the ball U it floats it, and by means of the 

 wire forces the valve T against the aperture that communicates 

 with the receiver R, and as the mercury continues to rise, the 

 air driven before it has no way of escaping but through the 



