562 Scientific Proceedings, Royal Dublin Society. 



of the electrolyte is very slow. It is made good by the addition 

 of a little water now and again. 



If heavy currents are to be run through the instrument, it will be 

 essential to make the electrodes large. Otherwise there is consider- 

 able evolution of ozone. In this connexion the use of phosphoric 

 acid in place of sulphuric acid might be worth trying. The evolu- 

 tion of ozone is objectionable, in that it diminishes the volume of 

 the issuing gas. 



It is very essential that water-dust be kept from clogging up 

 the orifice. The chamber beneath the orifice filled with cotton 

 wool should not be too small. It might be well to place the orifice 

 at the extremity of a long fine tube twisted into a spiral of several 

 turns, or other means devised to trap water-dust ; for observation 

 shows that if the orifice gets wet the readings become too high, 

 falling again when the orifice is touched with bibulous paper. 



None of these points which I have noticed in my experience 

 of the instruments need, I think, stand in the way of its application 

 to many purposes. It possesses the merit of not depending upon 

 the constancy of magnets. It might be cheaply and strongly con- 

 structed, and may, of course, be protected from the risks of currents 

 too heavy for it by a fusible cut-out. By lessening the size of the 

 orifice great sensitiveness may be imparted to the instrument, or 

 the same end attained by replacing the mercury by a liquid of less 

 density. Again we are not confined to the use of a single orifice. 

 Two or more may be provided communicating by taps or screw- 

 valves with the interior of the bulb. Each of these orifices may 

 correspond to separate values upon the scale so as to extend it to 

 any desirable degree. Promptness is attained by using an index- 

 tube of fine bore, and leaving but little free space above the electro- 

 lyte. 



Experiments upon possible variations of the readings with 

 temperature are desirable. Atmospheric changes of temperature 

 will probably entail only small errors in the scale readings. 



I regret that I am unable to give results bearing upon these 

 points here ; but perhaps some reader of these remarks may consider 

 the invention to be of sufiicient promise to induce him to go more 

 fully into its merits and demerits, and assign more definitely its 

 value as a means of measurement. 



