14 M. G. Wiedemann on Torsion. 



beaker during the temporary torsion of the wire. The tem- 

 porary torsions Tz and Ta were then : — 



I. 



II. III. 



IV. 



v. 



VI. 



T* 2167 



2179(2188) 2191 



2194 



2194 



2194-5 



Ta 2181 



2194 2200 



2198 



2200 



2203 



Hereupon the wire was allowed to unturn by lifting the twist- 

 ing-weights ; its permanent torsion was immediately 559, and 

 after a longer time 552. If water was again let into and out 

 of the beaker, the permanent torsions Yz and Pa were : — 





I. 



II. 



III. 



IV. 



p* 



542 



542 



541 



541 



Pa 



549 



549 



549 



549 



In the temporary torsion, therefore, the first loadings and 

 also the unloadings produce first of all an increase ; in the 

 permanent torsion the first loading effects a diminution. If 

 this effect has run out, in the present experiments a decrease 

 apparently takes place in both cases at the loading, and an 

 increase at the unloading, to the slight amount of 6-9 scale- 

 divisions. In order to ascertain if this trifling effect might 

 not result from the construction of the apparatus notwith- 

 standing the utmost care in its arrangement, the wire was 

 again unloaded. The positions of the mirror before and after 

 the loading corresponded to 584*7 and 579*2 scale-divisions 

 respectively. The difference amounts to +5*5 in favour of 

 the unloading. If, on the other hand, the steel ring was 

 screwed beneath, in which the point that was fastened under the 

 wire and sustained centrically by it was suspended, the position 

 remained perfectly unaltered at 587. If the wire was then tem- 

 porarily twisted up to 2212 scale-divisions, and then slowly 

 carried over to the permanent torsion- position 494, and after 

 a longer time to 488, and then alternately loaded and unloaded 

 beneath ; I obtained the following : — 





I. 



II. 



III. 



IV. 



V. 



VI. 



Pz 



482-5 



480-8 



480-1 



479 



478-8 



478-8 



Pfl 



481-8 



480-2 



480 



478-8 



478-8 



478-8 



Here, therefore, after the first effects have run out, which 

 at first produce at each change of tension a diminution of the 

 permanent torsion, the latter remains perfectly invariable. 

 The above-observed small final variations of Yz and Pa, as 

 well as of Tz and Ta, are consequently to be attributed to 

 slight rollings of the steel point in the ring, which must always 

 occur if at the action of the twisting-weights it is pressed to 

 one side of the steel ring, rotated on it, and then at the change 



