230 Effects of Twist on Thermo-electric Qualities of Iron. 



if) The third and last result I have to communicate refers 

 to the effect of agitation in the form of brisk tapping, which 

 is known to be quite marked in the case of longitudinal 

 tension. The agitation seems to be less effective in the present 

 than in that case; yet is quite sensible. The two pairs of 

 curves, figs. 5 a, -5 b, and figs. 6 a, 6 b, will give an idea of 



Fig. 5. 



^%, 



(a) 



(») 



Fig. 6. 



the effect. One of each pair refers to a series of observations 

 made as usual, and the other to the immediately succeeding 

 one in which there was interposed at some easily distin- 

 guishable point the operation of tapping (namely, with a 

 wooden rod on the stretched rope by which the leaden weights 

 were suspended). The former pair has the range of twist 

 within the critical value of twist for the reversal, while in the 

 latter it lies beyond it. The effect of agitation seems in both 

 cases to be opposite on the direct and back branches, on one 

 side as well as the other of the axis of zero twist; so that agi- 

 tation has in figs. 5 the tendency to make the branches approach 

 towards each other, and in figs. 6 to separate from each 

 other. Hence it was expected that it might be possible by 

 continual tapping to pass from the second type of the curves 

 (i. e. after the reversal) to the first (z. e. previous to it). This 

 seems to have been but partially fulfilled in a few experiments 



