Properties of the Two New Fluids in Minerals. 1 27 



of the expansible fluid either unite into one, or are subdivided 

 by the interposition of some portion of the dense fluid, which 

 has been expelled from its primitive situation, and drawn into 

 its new position by capillary action. When the specimen 

 cools, the two fluids quit their new position ; and, as if they 

 were endowed with vitality, they invariably resume the same 

 positions which they occupied at the commencement of the ex- 

 periment. 



Another form of the cavities, still more remarkable, occurs 

 in a very fine specimen belonging to Mr Sivright. These 

 cavities resemble a number of parallel cylinders, as shown at 

 AB in Fig. 2 ; but, owing to some cause which it is difficult 

 to conjecture, a number of them have been afterwards turned 

 aside towards C, so as to be open at one of their extremities. 

 From these extremities, which terminate in the surface ACB, 

 the fluids have made their escape, and have left the interior of 

 the cavities lined with a black and transparent powdery residue, 

 which always remains after their evaporation When the ca- 

 vities thus inflected and deprived of their fluids are submitted 

 to the microscope, they exhibit the most extraordinary shapes, 

 some of which are represented in Figs. 3, 4, 5. They have 

 the appearance of having been formed by a turning lathe ; 

 and such is the symmetry and beauty of their outline, that it 

 is not easy to conceive that they are the result of any mecha- 

 nical cause. One of these cavities, which is unconnected with 

 the rest, resembles a finely ornamented sceptre, as shown in 

 Fig. 3 ; but what is more remarkable, the different parts of 

 this figure lie in different planes, so that, when it is seen in a di- 

 rection at right angles to that of symmetry, it appears merely 

 a number of disjointed lines, as in Fig. 6. 



The inflection of the cavities AB into the directions bC, 

 &c. Fig. 2, and the discharge of their fluid contents at the 

 surface ACB, could only have taken place when the whole 

 mass ACB, though crystallized, had not attained its perma- 

 nent induration. This opinion derives great support from 

 the fact, that the lines bC are perpendicular to the axes of 

 the prism, and consequently lie in the planes of most eminent 

 cleavage. The direction, therefore, in which the fluids were 



