128 Prof. Hare's improved Eudiometrical Apparatus. 



water of a pneumatic cistern, the air being allowed to escape, 

 and the valve then to shut itself under 

 the water, on lifting the vessel it comes 

 up full of the liquid, and will remain 

 so, if the lower orifice be ever so little 

 below the surface of the water in the 

 cistern. Thus situated, it may be filled 

 with hydrogen, proceeding by a tube, 

 from a self-regulating reservoir. If the 

 apex A be then placed under any ves- 

 sel, inverted duly in the usual way, the 

 gas will pass into it as soon as the valve 

 is lifted. 



Volumes of atmospheric air are taken 

 by the same instrument, simply by low- 

 ering it into the liquid of the cistern, 

 placing the apex under the vessel into 

 which it is to be transferred, and lifting 

 the valve: or preferably by filling it 

 with water, and emptying it in some 

 place out of doors where the atmo- 

 sphere may be supposed sufficiently 

 pure, and afterwards transferring the air 

 thus obtained, as above described, by 

 opening the valve while the apex is within the vessel, in which 

 the mixture is to be made. In this case, while carrying the 

 volumeter forth and back, the orifice must be closed. This 

 object is best effected by a piece of sheet metal, or pane of 

 glass. 



It is necessary that the water, the atmosphere, and the gases 

 should be at the same temperature during this process. 



III. Sliding Rod Gas Measure. 

 The construction of this instrument differs from that of my 

 sliding rod eudiometers, in having a valve which is opened and 

 shut by a spring and lever, acting upon a rod passing through 

 a collar of leathers. By means of this valve, any gas, drawn 

 into the receiver, is included so as to be free from the possi- 

 bility of loss, during its transfer from one vessel to another. 

 This instrument is much larger than the eudiometers for ex- 

 plosion, being intended to make mixtures of gas, in those cases 

 where one is to be to the other, in a proportion which cannot 

 be conveniently obtained by taking more or less volumes of 

 the one than the other, by means of the volumeters: as for 

 instance, suppose it were an object jto analyse the air accord- 



