VOL. LXXXVIII.] PHILOSOPHICAL TRANSACTIONS. 369 



could preserve that smoothness and natural brilliancy which are common to all sub- 

 stances that freely assume a crystalline form. Like the crystals of feldspar which 

 we meet with in the porphyroid granites, the corundum crystals have been en- 

 veloped, at the time of their crystallization, by the substance of the rock which 

 was forming, at the same time with themselves, in an imperfect and confused 

 crystalline mass; and the corundum crystal, before it had acquired its perfect 

 solidity, necessarily received on its surface the impression of the different particles 

 of the rock which enveloped them ; this naturally renders the surface rough and 

 dull. Crystals of feldspar found in the granitic porphyroid rocks, exhibit the 

 same kind of appearance, from the same cause. 



The corundum crystals are in general opaque, or at least they have only an im- 

 perfect transparency at the edges: when broken into thin fragments, the pieces are 

 semi-transparent: when held between the eye and the light, and examined with a 

 powerful lens, it will be perceived that their interior 'texture is rendered dull by an 

 infinite number of small flaws crossing each other, much resembling the medullary 

 part of wood, when viewed in the same manner. The degree of transparency of 

 the small interstices which are between these flaws, is further evidence that this 

 texture of small flaws occasions opacity, which augments in proportion to the 

 thickness of the fragments. This kind of internal structure has also a very strong 

 analogy with that of feldspar in granite and porphyry. The endeavour to split 

 these crystals, in a direction either perpendicular or parallel to their axes, meets 

 with a very considerable resistance: they may indeed be broken in these directions; 

 but the rugged and irregular surface of the broken parts, clearly proves that the 

 direction in which the crystalline laminae have been deposited on each other, has 

 not been followed. The regular hexaedral prism of these crystals, cannot there- 

 fore be considered as the form of the nucleus of the crystal; and consequently is 

 not the primitive form of the crystals of this substance. 



If, in order to discover the direction of the crystalline laminae, a variety of 

 crystals be examined, some will hardly fail to be met with, which, on their solid 

 angles, formed by the junction of the sides of the prism with the planes of the 

 extremities, present small isosceles triangles. These are sometimes greater and 

 sometimes smaller, and form solid angles of 122° 34', with the extreme planes of 

 the crystal. They are in some instances real faces of the crystal; but most fre- 

 quently they evidently are the effect of some violence on that part. The smooth- 

 ness and brilliancy of these small faces, in the latter case, show that a piece has 

 been detached in the natural direction of the crystalline laminae. It is indeed 

 much less difficult to separate a portion of the crystal at these angles, than at any 

 other part; and, in following the natural direction of the faces, with a little 

 patience and dexterity, all the crystalline laminae may be detached, and progressively 

 increase the size of the triangular face. This operation however cannot be done 

 indiscriminately on all the solid angles of the crystals, but only on the alternate 

 ones at the same extremity, and in a contrary direction to each other. As to the 



vol. xviii. 3 B 



