434 DE W. PEDDIE ON 



below which recovery from fatigue has proceeded to a greater extent than it has for 

 larger angles of distortion. 



o 



Zero Effect of Period of Oscillation. 



In order to determine whether or not the period of oscillation had any influence on 

 the values of n and b, on the date 27.10.97, the large oscillator was replaced by the 

 oscillator of smaller moment of inertia, which was used in the experiments described in 

 the first paper. The results are given in fig. 14. A comparison of the results given 

 in Table V., for the experiment 27.10.97 (2), with the results for previous experiments 

 with the large oscillator, e.g., with the results for the experiment 20.10.97, shows that 

 no change by halving the period. With such speeds of oscillation we must therefore 

 regard the results as independent of "after-action." 



Law of Oscillation. 



We have already found that the period of a complete oscillation is very nearly 

 constant, being slightly greater for large oscillations than for small oscillations. 

 Some additions were made to the apparatus in order to make possible determinations of 

 the times of outward and inward motions over a given range. Fig. 17 shows the 

 details. The torsion head, to which the upper end of the vertical wire is attached, is 

 seen at the top of the diagram. The horizontal lead ring is seen attached to the lower 

 "iid of the wire. A Wimshurst machine is seen on the left side of the wire. A 

 vertical glass tube is seen at one extremity of a diameter of the lead ring. Its lower 

 end is drawn to a fine point, and it is filled with a coloured liquid. A similar tube is 

 placed ;it the other end of the diameter of the ring to secure symmetry in the oscillator. 

 The liquid in the tube is placed, by means of a copper wire, visible in the diagram, in 

 electric connection with the lead ring ; and a copper wire also connects the torsion head 

 (which is insulated by means of blocks of paraffin from the support to which it is 

 clamped) to one pole of the Wimshurst machine. When the machine is worked, the 

 liquid is driven out of the tube in a fine jet. On the right hand side of the diagram, at 

 t lower level than the lead ring, are seen massive iron blocks, between which is clamped 

 •i horizontal steel wire, which is weighted at its outer end in order to give it a suffi- 

 -iently long period of vibration. This wire supports a horizontal sheet of paper, which 

 vibrates with the wire. If this paper be at rest -while a torsional oscillation is given to 

 the vertical wire under test, the jet of liquid will trace a circle on the paper. But if 

 the paper now oscillates on the whole transversely to the motion of the jet, a waved 

 curve will be traced, which crosses the circle at each semi-vibration. The interval of 

 lime between two successive crossings is constant (equal to the period of semi- vibration 

 of the steel wire), and we can thus obtain a comparison of the times of outward and 

 inward motions over a given range. 



