204 Prof. C. V. Raman and Mr. Ashutosh Dey on 



how this mathematical result may be confirmed experi- 

 mentally. The general case of two discontinuities con- 

 sidered by Harnack covers a considerable and interesting 

 variety of forms of vibration, and the method described in 

 the previous paper has now been successfully extended so 

 as to obtain an experimental confirmation of Harnack's 

 results in some of these cases also. 



Experimental Metliod. 



If px and 2 represent two discontinuous changes of 

 velocity travelling on a string of finite length which com- 

 pletely determine its motion, the velocity-diagram of the 

 string must in general consist of three parallel straight lines 

 as shown in fig. 1 (a) or fig. 1 (b). Each of the two outer 



Fi g- !(«)• Fig-. 1 (b). 



lines passes through one of the fixed ends of the string and 

 is separated from the intermediate line by a discontinuity. 



In fig. 1 (a) the discontinuities are of the same sign, and 

 in fig. 1 (6) they are of opposite signs. A velocity-distri- 

 bution similar to that shown in either of the figures would be 

 obtained if a string has initially a uniform angular velocity 

 about a point in its own line, and if in the course of this 

 motion the two points A and B are suddenly fixed, either 

 simultaneously or else successively, at an interval less than that 

 necessary for an impulse to travel from A to B or vice versa. 

 If the point about which the string has initially a uniform 

 angular velocitv lies within AB, we have the case shown in 

 fig. 1 (a). If it lies in BA produced, the velocity-distribu- 

 tion is similar to that shown in fig. 1 (b). The photographic 

 records presented with this paper (PI. III. fig. 2) refer to 

 a case in which the string has initially a uniform angular 

 velocity about a point lying outside the two fixed stops A 

 and B, and the discontinuous changes of velocity which 

 determine the form of vibration are therefore of opposite 

 signs. 



The method by which the records are obtained is similar 

 to that described in the previous paper, but with certain 

 necessary modifications. The tension and initial motion of 

 the string are, as before, secured by a weight attached to its 



