560 Scientific Proceedings, Royal Dublin Society. 



6 cms. in diameter. This communicates below with, a tube, benfc 

 twice at right angles, and brought up to a height of about 50 cm. 

 above the level of the bulb. This tube is open at top — only loosely 

 plugged with cotton wool to keep out dust. In the bulb are twa 

 platinum electrodes of large area (each 2J x 4 cms.), and set with 

 their plane surfaces vertical and about 2 millimetres apart. These 

 communicate with the binding screw h, h by platinum leads fused 

 in the glass. At the upper surface of the bulb are two openings. 

 The one a small tubulure stoppered by a sound tight-fitting cork 

 (a ground-stopper would probably be preferable), the other a 

 tubulure, over which is sealed by strong turner's-cement a metal 

 cap having a plane surface above, and perforated by an orifice^ 

 about 2 mms. in diara.) at o. Below this metal cap there is a. 

 widening in the glass tubulure which is filled with cotton wool 

 before the cap is cemented in position. Over the ground surface- 

 of the cap a second cap s can be screwed. This is open at top. 

 A small disk of platinum foil, rather less in diameter than the 

 ground surface of the cap, is perforated by a very small puncture, 

 and then cemented down over the orifice o. To effect the puncture 

 a very fine steel needle is used, the platinum disk being placed 

 upon a smooth sheet of ebonite or even glass. As will be seen, tlie 

 size of the perforation regulates the degree of sensitiveness of the 

 instrument. When the disk is cemented above the orifice o, the 

 cover s is screwed on, a little loose wool being inserted in its opening,, 

 to protect the perforated plate from dust. 



The second tubulure now serves to admit some clean mercury 

 which is poured in till it rises well into the bulb. On top of this, 

 water, mixed with about 15 per cent, by volume of sulphuric acid, 

 is poured till the bulb is almost filled. The cork is then tightly 

 inserted, and the instrument is ready for calibration. 



What will be the effect of passing a current of given strength 

 through the instrument ? Suppose a current of one ampere. The 

 volume of gas given off per second by this current is 0"1734 cubic 

 cms. if measured at 0° 0. and 760 mms. pressure. But this quantity 

 of gas cannot escape through the orifice in the platinum plate in 

 one second, except the pressure upon the gas is increased. It 

 therefore rapidly accumulates in the bulb, increasing in pressure, 

 and driving the mercury up the vertical index-tube communicating^ 

 with the bulb. This goes on till the increase of the gas has 



