1831.] on Vibrating Elastic Surfaces. 343 



water-heaps which formed them, and just double in number to 

 the latter. At other times, the sand, instead of being deficient 

 at the intersecting angle, would accumulate there only, fig. 20 ; 

 Fig. 17. Fig. 18. Fig. 19. 



\ / \ / \ 



.v 



\ / \ / \ / * * <k f \ * 



/ \ / \ / \ *%*%#* 



N / S / \ / v * * * v < 



/ \ / \ X' \ ***<%*> 



and at other times would accumulate there principally, but still 

 show the original form by a few connecting particles, fig. 21. 



88. When the heaps were of the form described (84), the 

 sand was still washed from under them ; it did not however 

 assume lines parallel to the rectangular arrangement of the 

 heaps, but was arranged as in fig. 22. 



Fig. 20. Fig. 21. Fig. 22. 



89. When only the circular linear heaps (83) were produced, 

 the sand assumed similar circular forms, concentric and alter- 

 nating with the water elevations. 



90. On strewing a little lycopodium over the water for the 

 purpose of gaining information relative to what occurred at the 

 surface during the crispation, it moved about over the fluid in 

 every possible direction, whilst the crispations existed of the 

 utmost steadiness beneath. The same thing occurred with 

 pieces of cork on very large crispations (98). But when much 

 lycopodium was put on, so that the particles retained each 

 other in a steady position, then it formed lines* parallel to the 

 arrangement of the heaps, the powder being displaced from 

 the parts over the heaps, and taking up an arrangement per- 

 pendicularly over the sand beneath. As the lycopodium forms 

 float on the water they are easily disturbed, and in no respect 

 approach as to beauty and utility to the forms produced by 

 the sand ; but lycopodium may be used with smaller crispations 



than sand. 



* Wheatstone. 



