Properties of the Two New Fluick in Minerals. J 31 



opening and shutting like a valve. This effect is finely exhi- 

 bited in a cavity of a specimen belonging to Mr Sanderson, 

 represented in Fig. 10 by A B C D E. In ordinary tempe- 

 ratures, about 45°, there is a vacuity of the size V, in the ex- 

 pansible or dotted fluid, and the dense, or shaded fluid, occu- 

 pies the necks b c, d e, DE, and also the extremity F. By ap- 

 plying the heat of the hand to the specimen, the expanding fluid 

 in the branches V C, V D, finds space for itself, by filling up 

 the vacuity, but as there are no vacuities in the portions of ex- 

 panding fluid at A B, B, and E F, they must necessarily force 

 out the dense fluid which confines them. The dense fluid in 

 the neck E D, is thus made to appear at D, and the whole of 

 the dense fluid at b c is driven off to d e, till, accumulating 

 there, it is drawn by attraction to the nearest neck, m n op. 

 Here it first lines the circumference of the hollow neck, from 

 its powerful attraction for topaz ; and, as the lining becomes 

 thicker, it appears as a slight elevation between o and p, and 

 between m and n. These elevations increase till they leap to- 

 gether by their mutual attraction, and form a column of the 

 dense fluid in n p o. The column b c of dense fluid has now 

 disappeared entirely, and the space A B C D is filled with 

 the expanding fluid. The heat of the hand being continued, 

 the expanding fluid A B forces itself through the little cylin- 

 der of dense fluid d e, which resumes its place the moment 

 that a portion of the former has passed. But as the same 

 heat has been expanding the fluid between n p and C, which 

 pushes out part of the dense fluid at m n o p, this dense fluid, 

 and the surplus of what was displaced from b c, moves along 

 the sides of the cavity till it occupies the portion q r, of the 

 branch V D. Sometimes the dense fluid is entirely driven 

 from in n o p, and part of it sent to the extremity C ; though, 

 in general, a very small portion remains at the very neck m o. 



As the specimen cools, the dense fluid quits in o and q r, 

 and is gradually transferred through the neck d e to the neck 

 b c ; every portion of it invariably resuming its primitive po- 

 sition. 



A curious modification of these actions is seen in a cavity 

 of the specimen shown in Fig. 11. The branch b V has al- 

 ways a vacuity V, while the cavity A, connected with it by the 

 filamentous channel o b, has no vacuity. At the ordinary 



