382 Dr. N. Ernest Dorsey on the Surface- Tension of 



The magnitude of the error introduced in this way depends 

 upon the amplitude of the waves, and this is probably the 

 reason why the results obtained on different days were not 

 comparable ; for I frequently changed the amplitude of my 

 fork so as to alter the distinctness of the field in the telescope. 

 Many weeks were spent in locating this error ; and after 

 finding it no reading was taken nearer the fork than four 

 centimetres. Under these conditions the measured wave- 

 length does not depend upon the number of waves measured, 

 nor upon the distance of the waves from the fork, nor upon 

 the amplitude of the fork (so long as it is small). All these 

 points were tested experimentally. I was very seldom troubled 

 with reflected waves. 



The mirrors M x and M 2 must be good plane metal mirrors. 

 Owing to the double reflexions ordinary glass-front mirrors 

 cannot be used, since in my arrangement the images overlap 

 and distort one another. The silver surface obtained by 

 depositing silver by Brashear's process* on good thin plate 

 German mirror-glass was used in this experiment. 



Periods of Forks. 



The fork F 2 is a large one with a mirror and counter-weight 

 screwed into the ends of its prongs, and belongs to a set of 

 Konig forks used to show Lissajous's figures. The mirror 

 was replaced by a second counter-weight, and the plate of 

 glass which generated the waves was fastened by screwing 

 down the weight until the plate was clamped tightly between 

 it and the end of the prong of the fork. This allowed the 

 plate to be removed and cleaned without danger of wetting 

 the fork. Sliding on the prongs were light brass weights 

 that could be clamped wherever desired. 



F 2 was also provided with sliding weights ; and after 

 putting on the strips of copper (S 2 ) I tuned it (by means of a 

 vibration-microscope) to unison with F 2 while the weights on 

 the latter were near the centre of its prongs, and while the 

 plate of glass was dipping into water. 



By means of the smoked glass and pendulum method the 

 frequency of the vibration-microscope, which was an octave 

 above F 1; was determined. It made 125*8 vibrations per 

 second. Then as a check the vibration-microscope was again 

 compared with Fj, and it was found that the ratio of their 

 frequencies was still two to one. The vibration-microscope 

 rather than Fj was compared with the pendulum because it 

 was more readily handled than the larger fork. 



* Astrophysical Journal, i. p. 252 (1805). 



