0. K. Wead — Intensity of Sound. 233 



tive in producing sound. Yet it is often assumed that the 

 effective part of this fraction is a constant fraction of the total 

 energy, for all amplitudes of vibration. The critic, I think, 

 did this, as I can find no other possible explanation of the 

 mistake he has made ; but such an assumption is pure guess- 

 work. 



In the paper referred to, a considerable part of the work re- 

 ported bore on the determination of the rate of loss of energy ; 

 it was found for example that, with Ut 3 fork at its maximum 

 amplitude, it lost energy at a rate that would, if it had been 

 continued uniform, have exhausted the supply in 4|- sec. ; but 

 when the amplitude was very small, the rate too was reduced, 

 and the corresponding time was 15 sec. 



Stefanini, in a later paper,* criticises the form in which I 

 write the exponent expressing the damping effect, viz : 



/ 7 z'+z"\ 



xt = (a + b 9 \t 



Where z' and z" are the amplitudes at the beginning and 

 end of the time t. He prefers to write it at m ; he finds for 

 several forks, whose vibrations he photographed, that m was 

 tolerably constant for each fork ; for different forks it varied 

 between 0-84 and 0*94. An important difference between the 

 two formulas is that according to Stefanini's, the amplitudes at 

 the end of successive equal times should be in a geometric 

 series, or if the ratio of amplitudes is constant the intervals are 

 constant ; according to my formula in the last case, the inter- 

 vals will increase as the amplitude diminishes. Stefanini used 

 long thin forks intended for producing Lissajous's curves. 

 From one of his tables (Tab. IX, p. 330 for Sol,) we deduce 

 the following: 



Time, arbitrary units 10 20 30 40 50 60 



Interval 10 10 10 10 10 10 



Amplitude on photograph, 48'3 32'3 22*8 16'35 12*2 9*2 6'6 

 Ratio 1-50 1-41 1*39 1*34 1-33 1-39 



My Table II, p. 182, gives for Sol 3 the following ratios equal 2. 



Amplitudes in div, 10-5; 8-4; 6-3; 4-2. 



Intervals, sec. 4-70 5*34 6-31 8-17 



Evidently the new formula does not fit any of the experi- 

 ments exactly ; like mine, it is only empirical, not rational ; 

 every experimenter must determine the formula that best fits 

 his own conditions of working. 



In this second and longer paper the author attempts to de- 

 termine experimentally by forks whether the intensity of sound 



* Beiblatter, xiii, 871. Atti, xxv, 307-400. 



