on Visible Vibration. 



199 



and the radius of convexity of the surface varying with the 

 quantity of the fluid. 



90. By proceeding in this way many varieties of figures 

 are produced, of which, perhaps, one or two more will 

 suffice for our present purpose. 



Fig. 4. 



Fig. 5. 



The same remarks which were made respecting the forma- 

 tion of the former figures will apply to these likewise. 



91. To produce symmetrical figures, an essential point is 

 to keep the mercurial surface clean and free from impurities, 

 as the slightest particle of dirt will, in some measure, check 

 the continuous formation of the undulatory curves. 



92. But even this defect is the source of many new 

 curves interlacing among, what we may term, the primary 

 curves, and producing modifications in the figures, which, 

 though not symmetrical, are still pleasing. Thus, when a 

 small shot, or a particle of dirt is on the surface of the 

 mercury, it becomes, when vibration has commenced, a 

 focus of a system of concentric ellipses, the other focus of 

 which is the centre of the mercury. These elliptic curves 

 spread out from the prolate axis to a considerable distance, 

 and produce very complicated curves by interference with 

 the revolving star and the undulae. 



93. There will be as many of these systems of ellipses as 

 there are impurities (if of an appreciable size), and, as the 

 size of these ellipses depends on the distance of the particle 

 from the centre of the mercury, they cannot join in a har- 

 monious whole; but still, as the prolate axis of these 



