106 M. G. Wiedemann on Torsion. 



considerably in proportion to the increments of the tempo- 

 rary, the decrease of the oscillation-amplitudes during the 

 same period must, as has been observed, with greater elonga- 

 tions be relatively more considerable. If before the oscilla- 

 tions the wire (as, for example, a drawn wire) already pos- 

 sesses a permanent torsion, this also will be lessened by the 

 action of agitation during the oscillations ; a corresponding 

 displacement of the zero-position accompanies the diminution 

 of the elongations; and the latter can be greater in the direction 

 of the displacement than in the contrary direction*. If after 

 the accommodation the wire be again permanently deformed 

 (for instance, more twisted), it is self-evident that the wire in 

 swinging makes a new period of accommodation, with en- 

 hanced diminution of the elongations by the action of agita- 

 tion, &c.| 



24. Influence of the Load. — As in the above-mentioned 

 torsion-experiments (§§ 12 and 13) in relatively soft wires 

 the static displacements are initially greater with a heavier 

 than with a lighter load, so also must the decrements atten- 

 dant upon oscillations be greater in the former case, but, like 

 those, gradually approach the magnitude which is observed 

 with lighter loading. This is proved by experiments which I 

 made as long ago as the year 1860. 



A soft silver wire of about 0'5 millim. thickness and 75 

 centims. length was suspended in a double-walled vertical 

 brass tube, through the annular space of which aqueous 

 vapour could be conducted. The wire, like those used above 

 (§ 2 et seqq.), carried a brass clamp provided with a mirror, 

 which in like manner supported on a hook a cylindrical brass 

 box 14 centims. in height and 6 centims. wide, into which 

 closely-fitting leaden weights (B) could be put. The ampli- 

 tudes of the torsion-oscillations of the wire, previously an- 

 nealed and then set oscillating, were observed by means of a 

 telescope and scale. The following Table contains under T 

 the excursions in both directions directly observed on the scale, 

 under E the total elongations, under N the mean of each pair 

 of excursions in the two directions, under v the percentage of 

 diminution from the first to the tenth elongation. The num- 

 bers placed before the observations are the ordinal numbers 

 of successive oscillations. In series II., under Z are given 

 the times which had elapsed between the observations. The 

 wire was kept incessantly oscillating during the whole series. 



* Compare, for instance, Neesen, Pogg. Ann. cliii. p. 54 (1874). 

 t Compare also P. M. Schmidt, Wied. Ann. ii. p. 63 (1877). 



