526 



Mr. A. J. Ellis. 



[June 17, 



machine, and Professor Mayer, of the Stevens Institute, Hoboken, New 

 Jersey, U.S., kindly measured the first 5 forks by his electrographic 

 method, both with the greatest care and precaution. The three sets 

 of measurements agreed to less than 1 beat in 10 seconds, and more 

 often less than 1 beat in 20 seconds, when reduced to the same tempera- 

 ture. Thus the value of the tonometrical measurement by beats only, 

 and the possibility of counting a tonometer sufficiently, was fully estab- 

 lished. Koenig's measurements of his own forks reduced to 59° F., 

 and of the actual Diapason Normal at the Conservatoire, Paris, in- 

 tended to be used at the same temperature, also agree with mine 

 within less than the same limits. By these forks I have counted 75 

 standard forks of Messrs. Valantine and Carr, music smiths, 76, Milton 

 Street, Sheffield, who are thus* in a position to make small copies, 

 probably not more than half a vibration wrong at most, at a cheap 

 rate. 



When forks are counted without a resonance jar, they should not 

 be applied to a sounding board, or held one to one ear and one to the 

 other, but should both be held about six inches from the same ear, 

 and their strengths should be equalised by holding the weaker fork 

 closer to the ear than the stronger. 



When the forks are screwed on and off a sounding board or resonance 

 box, there is great danger of wrenching the prongs, unless they are 

 held below the bend, but I have constantly seen this precaution 

 neglected. A wrench immediately affects the pitch and duration of 

 sound of a fork, and renders it comparatively worthless. Such cases 

 have come within my observation. The next enemy to be guarded 

 against is rust. Forks should be kept dry, and occasionally oiled 

 with gun-lock oil. Rust towards the tip affects the fork much less 

 than rust at the bend. My observations and experiments show that 

 errors from rust can scarcely exceed a flattening of 1 vibration in 250, 

 and are generally very much less. But as the amount is uncertain, 

 rust spoils a fork for accurate tonometrical purposes. 



My observations on reed tones are confined to those in Appunn's 

 tonometers at the South Kensington Museum, the Museum of King's 

 College, London, and Lord Rayleigh's, where there are copies, all of 

 which I have counted. They consist of oblong boxes containing the 

 reeds placed side by side. The wind pumped into a large reservoir, is 

 driven from it by a spring, and received into this box, the heavy lid of 

 which, separated from the body by the usual bellows-folds of leather, 

 presses on the wind and drives it on the reeds with a very constant 

 pressure. Below each reed is a pallet which, when pulled out by a 

 valve to its full extent, allows the reed to sound at its highest pitch to 

 which it is tuned. If the pallet be slid in somewhat, the pitch can be 

 flattened by as much as two vibrations without much affecting the 

 quality of tone, and even as much as three vibrations with considerable 



