222 



Mr. G-. J. Burch. Contributions to a [Apr. 17, 



that of a gold-leaf electroscope, or essential, in the sense that some 

 small current may be necessary to maintain a deflection. Lippmann, 

 in his original thesis, did not mention any direct experimental inves- 

 tigation of this problem, though his whole mathematical argument is 

 in accordance with the supposition, that no current is required to 

 maintain a deflection. I therefore began my practical examination of 

 the properties of the capillary electrometer in January, 1887, with this 

 experiment. I made a quick-acting electrometer, carefully insulating 

 its parts with freshly cut soft rubber, and joined it up in series with 

 a very sensitive galvanometer in the derived circuit of a potentio- 

 meter. Both instruments were furnished with short-circuiting keys- 

 On opening the electrometer key, there was a large excursion of 

 the meniscus. As soon as it had come to rest in its new position 

 under the action of the electromotive force, the galvanometer short- 

 circuiting key was opened. There was no deflection. But any thing 

 that caused the meniscus to move — whether a slight change in the 

 pressure on the mercury, or a small movement of the rider of the 

 potentiometer, produced a large or even violent, but always tem- 

 porary, deflection of the galvanometer.* 



I found afterwards that Fleischl had obtained similar results in 

 1879. 



The converse of this experiment is specifically referred to hy Lipp- 

 mann. If a charge is communicated to an electrometer by rubbing 

 the terminals with the finger, and the circuit is left open, the mercury 

 returns to its zero point very much more slowly than if the circuit is 

 closed. And I found that by careful attention to the insulation the 

 capillary electrometer could be made to hold a charge quite as long as 

 a gold-leaf electroscope. Thus another instrument made a few weeks 

 later was left charged with the circuit open, and " no change in the 

 deflection could be detected after 35 minutes." 



The most remarkable experiment of this kind was made unin- 

 tentionally. In June, 1898, I made a new projection electrometer for 

 the work on the electromotive properties of nerve in which Professor 

 Gotch and I are engaged at Oxford. This instrument is the most rapid 

 and sensitive that I have ever employed, and it is still in use. It is 

 a modification of that figured in my book, f The capillary, which 

 points vertically downwards, is contained in a glass trough about J mm. 

 wide and 25 mm. long, formed by grinding away the front of a 

 thick-walled tube so as to lay open the bore from end to end. This 

 trough is closed in front by a piece of glass, and its lower end clips into 



* A galvanometer Ly Elliott, of 25,500 ohms resistance. The current given by 

 one Daniell cell through a resistance of 63,900 megohms produces a deflection of 

 one scale-division, according to the maker's certificate. 



f " The Capillary Electrometer in Theory and Practice," reprinted from the 

 ' Electrician,' 1896. 



