228 [Dec. 12, 



6 inches long and 4 inches wide, with which I have been able to 

 produce a continuous sound, has been one ; and of Smee's cells three 

 (28), except under certain special conditions (36) : the most suit- 

 able number of the former is two, and of the latter from five to ten, 

 according to their size ; if this amount of power was much exceeded, 

 the action became violent, then quiescent, and the sounds ceased. 



28. With a ring of mercury ^th of an inch wide and 1-Z- inch 

 diameter, the smallest number of the Smee's cells (4), 1st, im- 

 mersed 8J inches, with which the sounds could be obtained was 

 two ; the pitch of the sound was rather high and very feeble ; it 

 lasted only about three seconds, and was not attended by any visible 

 vibrations : 2nd, immersed 5 inches, was three ; the tone was high 

 and sound feeble, for a few seconds with visible crispations, and con- 

 tinued after that more feeble and without visible vibrations : and 

 3rd, immersed ^ an inch, was four ; the sound was steady, tone 

 moderately high, and crispations y^th of an inch wide. And the 

 largest number that could be successfully used, 1st, immersed 8J 

 inches, was eighty and occasionally nine ; the crispations were then 



- very narrow and the action violent : and 2nd, immersed 5 inches, 

 was ten, crispations very narrow and action violent. The largest sur- 

 face of the ten Smee's elements that would produce definite crispations 

 and sounds was about 7-g- inches deep by 7 inches wide ; and the 

 smallest surface of three elements that would effect it was about 1 or 

 2 inches deep by 7^- inches wide, and then only faint sounds were 

 produced, lasting about two seconds, and unattended by visible vibra- 

 tions. It is worthy of remark, that a small number of elements of 

 large surface always produced small crispations and high sounds. 



Influence of Quantity of the Current. 



29. The width of the vibrations and the pitch of the sound are 

 closely connected with, though not solely dependent upon (40), the 

 quantity of electricity which passes into a given amount of mercury 

 surface in a given time. With a surface of given size, the width of 

 the crispations was invariably decreased, and the pitch of the sound 

 raised, by either increasing the number of the plates (their depth 

 of immersion remaining the same) or their depth of immersion. To 

 ascertain the most suitable quantity of electricity, a voltameter was 

 included in the circuit with the phonetic liquid ; the negative ring of 



