the Cohesion-Figures of Liquids. 427 



in a faint flame-like circular cloud ; the denser portions accu- 

 mulate at two opposite points of the ring, which is thus, at its 

 thinner portions, bent upwards (No. 2) and then drawn down- 

 wards into graceful festooned lines by the heavier portions, 

 which form separate rings smaller than the parent ring (see 

 No. 3). These rings in like manner descend; the denser 

 portions of colouring matter accumulate at the two opposite 

 points 90° from the points of attachment to the festooned lines; 

 each small ring is in this case also bent upwards, while it drops 

 two other rings, which in their turn go through the same series 

 of changes, until the figure becomes almost too complicated to 

 follow (see No. 4). This is a very common result with a mode- 

 rately weak solution of cochineal ; with a stronger solution, the 

 figure undergoes some beautiful modifications. The heavier 

 portions of the colouring matter collect not at two, but at four, 

 six, or even seven or eight points of the ring* — bending it up- 

 wards in as many curved lines, and letting drop as many rings, 

 each of which becomes the seat of manufacture of two rings; 

 and in this polypus method of division and subdivision the 

 colouring matter is diffused through the solution. 



The colour of the figures may, of course, be altered by adding 

 to the water a few drops of a solution of nitrate of copper, of a 

 salt of tin, of ammonia, &c.f If too much be added, the rings 

 will not go through their changes, but subside to the bottom of 

 the vessel, and remain for hours as rings. If the drops of 

 cochineal be allowed to fall slowly into the water through a 

 small funnel containing a filter, rings only will be formed ; and 

 these being nearly equidistant, and expanding as they fall, call 

 to mind the beautiful rings formed by bubbles of phosphuretted 

 hydrogen bursting in air. I have noticed more than once, that 

 as the drops fell slowly through the funnel, and the solution in 

 the filter became denser, a ring would overtake its predecessor, 

 shoot through it, and expand beneath it without disturbing it. 

 But the best method of exhibiting these figures is that first 

 described — namely, by allowing a single drop to fall into the 

 water, and to go through its various changes before a second 



* See the figures Nos. 1, 2, and 3 of cochineal in water containing a 

 little oxalate of ammonia ; but the effect is not confined to the presence of 

 this salt. 



f In a glass cylinder 1 1 inches high and 3 inches in diameter, into which 

 about half an ounce of solution of ammonia had been poured before filling 

 it up with New River water, figures of great symmetry and beauty were 

 produced sufficiently slow in their development to admit of being studied. 

 In this instance, as in some others to be noticed presently, the curved fes- 

 toons were connected by a very delicate membrane. If the cylinder be 

 kept covered when not in use, the solution will act for a week or more 

 without being renewed. 



