RIGIDITY OF ROCKS AM» HYSTEEESI8 FUNCTION. 7 



readings taken : and so on till a definite amounl of couple was 

 reached. 



8. Then, the weighl was transposed in the opposite way so that the 

 couple diminished gradually and ultimately became oppositely 

 directed. In this way, a series of observations was made to 

 complete the cycle many times. 



9. From the difference of the deviations of the two imaeres, the 

 amount of twist due to the corresponding weight difference was 

 calculated. One mm. of scale division =1.845 x 10~ 4 radian. 

 One gram of mass =6.71'2xl0 3 c.g.s. unit of couple. 



Thus, it will be easily seen that the result of observation was at 

 no time affected by any external disturbance such as the yielding of the 

 scale-support, or minute displacements of the telescope, or flexure of 

 the floor on which the tw 7 isting apparatus rested. One instance is 

 given in PI. III. for the case of a piece of clay slate. As might be 

 expected, the irregularity of the course of the curve A is enormous, 

 while the course of the curve B is quite free from it. 



Here it is necessary to explain why, in the third operation above 

 mentioned, I put a superfluous weight }> M on the pans giving no 

 couple. When the tw T o weights, whose respective masses are nij and 

 m 2 , are put on the pans Qj and Q 2 respectively, the resulting couple is 

 given by 



G = [{ Wi — >»:.,}+ «rf;»! + m-S\ r g- 

 the upper or the lower of the double sign being selected according as 

 {?»] — m.,} is increasing or decreasing ; where o is a numerical coeffi- 

 cient depending on a dissipation of couple due to the imperfectness 

 of the arrangement ; e.g. friction of the pulley, imperfect flexibility 

 and extensibility of the strings. For future reference, m l — m. : ,=M and 

 m 1 +m s =M will be called the Effective Mass and the Resisting Mass 

 respectively. Thus 



