44 



Scientific Proceedings, Royal Dublin Society. 



towards the right as the magnetic field round the wire is increased. Tliis is 

 not likely to be due to any errors in the observations, as tbe observations were 

 taken repeatedly with the same results ; this flexure in the curves did not occur 

 when the initial amplitude of oscillation was about 7°, as is shown in the curves 

 of fig. 2. 



The curves obtained in the same way when the initial amplitudes were 

 70°, 50°, and 30° respectively, all showed the same kind of flexure, but to a 

 diminished extent, so that when the initial amplitude is small (7J^°), the 

 flexure in the curves seems to be smoothed out. 



The influence of the magnitude of the initial amplitude on the rate of 

 subsidence is perhaps better shown by reckoning the final amplitude attained 

 after a certain definite number of oscillations have been made, starting in each 

 ease from a different initial amplitude. The results of some observations of 

 this kind are given in Table VI., and are shown as curves in fig. 6. 



Table VI. 



The number at the right-hand end of each curve in fig. 6 indicates the 

 initial amplitude from which the oscillations were reckoned. By comparing 

 the curves in fig. 6 with those in fig. 3, which were obtained with a small initial 

 amplitude of 7^°, they are seen to be of the same type ; but as the initial 

 amplitude is diminished the longitudinal magnetic field in which the greatest 

 damping of the oscillations takes place is increased ; thus, with a starting 

 amplitude of 90°, the greatest damping takes place when the wire is in a 



