342 On the Forms and States of Fluids [1831. 



If the marks in fig. 15 be supposed to represent Fig. 15. 



the tops of the heaps, an idea of the whole will ~"_j-_^-_^ 



be obtained. Three inches along these heaps - 



included eight, but across them it included T_~I_~L' 



fifteen nearly. These numbers are therefore the Z_~~L.~I_T_~I 

 relation of length to breadth. But along the I_~H~IL~7V~ 

 lines of the quadrilateral arrangement three ~ ~" 

 inches included eleven heaps, which, notwithstanding the dif- 

 ference in form, is the same number that was produced by the 

 same plate, with the same depths of water, when the heaps were 

 round ; therefore an equal number of heaps existed in the 

 same area in both cases ; and the departure from perfect rect- 

 angular arrangement, and also the ratio of 1 : 2, is probably 

 due to some slight influence of the sides of the plate. 



85. When mercury covered with a film of very dilute nitric 

 acid is vibrated (77), the rectangular arrangement is constantly 

 obtained. When vibrated under dilute ink (78), it is still more 

 beautifully seen and distinguished. The tin plate sustaining 

 the mercury was square, and when the whole surface was 

 covered with crispations, the lines of the rectangular arrange- 

 ment were always at angles of 45 to its edges. 



86. When sand is sprinkled uniformly over a plate on which 

 large water crispations are produced, i. e. four, five or six in 

 the inch, it gives some very important indications. It imme- 

 diately becomes arranged under the water, and with a little 

 method may be made to yield very regular forms, 



It is always removed from under the heaps, 

 passing to the parts between them, and fre- 

 quently producing therefore the accompanying 

 form, fig. 16, of great regularity. As the sand 

 figure remains when the vibration has ceased, 

 it allows of the determination of position, the 

 measurement of intervals, &c. very conveniently. 



87. Very often the lines of sand are not continuous, but 

 separated with extreme regularity into portions, as represented 

 fig. 17. The portions of these lines were sometimes, with 

 little sand on the plate, very small, fig. 18; and when more 

 sand was present they were thickened occasionally, fig. 19; 

 then assuming the appearance of heaps arranged in straight 

 lines at angles of 45 to the lines regulating the position of the 



