in Topaz, Beryl, and Diamond. 175 



vacuities, which finally unite in one. The refractive power of 

 the expansible fluid varies from 1*1311 to 1*2106, while that of 

 the denser fluid is 1*2946, which is very much less than that of 

 water. From the few experiments which I was able to make on 

 these fluids when taken out of the cavities, it has been inferred 

 that they are hydrocarbons. 



The distribution of these cavities, in the specimens which 

 contain them, is a subject of peculiar interest. They are often 

 found singly, and of different sizes, at different depths in the 

 mineral ; but they most frequently occur in strata, and of such 

 different magnitudes that the two fluids are distinctly seen in 

 the largest, while the rest gradually diminish till they disappear 

 in black points which the microscope can hardly descry. Three 

 or four strata nearly parallel to one another, and with cavities of 

 different sizes, rarely occur. In general the strata lie in planes 

 frequently intersecting one another, and having no connexion 

 with the primitive or secondary planes of the crystal. In some 

 specimens the planes of the strata are curved, and in rare cases 

 the sections of these planes are curves of contrary flexure. 



In 1844 I was led to re-examine several hundred specimens 

 of topaz with a more perfect microscope and a fine polarizing 

 apparatus, with the view of ascertaining the nature and proper- 

 ties of certain crystalline deposits whicli I had noticed, and to 

 which I had referred in my earliest observations*. In these 

 new researches, the results of which were published in two papers 

 in the ' Transactions ' of this Society for 1845, I discovered two 

 new classes of phenomena which had escaped the notice of pre- 

 ceding observers, and which threw much light on the formation 

 of the minerals in which they were exhibited. 



In many specimens of topaz from Brazil and New Holland, I 

 discovered numerous cavities, filled with crystals of various 

 primitive forms, and with different physical properties. These 

 crystals are either fixed or moveable. Some of the fixed crystals 

 are beautifully crystallized, and have their axes of double refrac- 

 tion coincident with those of the specimen which contains them. 

 In some cavities there is only one crystal, in many two, three, 

 and four, and in a great number the crystals actually fill the 

 cavities to such a degree that the circular vacuity in the fluid 

 cannot take its natural shape, and can often be scarcely recog- 

 nized among the jostling crystals. 



Upon the application of heat to these crystals, some of them 

 gradually lost their angles, and melted slowly, till not a trace of 

 them was visible. Others melted with greater difficulty; and 

 some resisted the most powerful heat I could apply. The 



* See Edinburgh Transactions, vol. x. p. 21, note, and plate 1. fig. 10, 

 plate 2. figs. 20, 21 • p. 419, note, and plate 19. fig. 4, 



