Corundum Stone from Asia. 4,4$ 
H a rhomboidal parallelopiped, truncated at each of its obtuse 
<( angles, by an equilateral triangular plane. 
“ You will have a correct idea of the other crystal, if you 
“ suppose the crystal represented in Plate IV. fig. 87. truncated 
“ at each of the summits of its pyramids, by an equila f eral tri- 
“ angular plane, as in the preceding modification, but deeper, 
fc and in so great a degree, that the three rhombic planes of each 
“ pyramid disappear, with the exception of three isosceles tri- 
“ angles ; this modification differs from the first, only by the 
“ hexaedral prism, and the peeper truncature at the summits of 
“ the pyramids.” 
It is therefore clear, that if the primitive rhomboid of Corun- 
dum decreased only at the superior angles of its laminae, it would 
exhibit exactly the first of these varieties of Mr. Werner's 
ruby, as in the annexed fig. 19. 
As to the second variety of Mr. Werner’s ruby, it is equally 
clear, if in fig. 87, referred to by Rome' de l’Isle, (represented 
by the annexed fig. 20.) no more of the pyramid was left than 
the three small triangles b, a, c, there would be precisely one 
of the forms of Corundum before described, to which the an- 
nexed figure 5 belongs. 
It may perhaps be objected, that the laminae appear to be 
parallel to the terminal planes, in the sapphire, and inclined, in 
the Corundum. There are crystals of Corundum, in which, very 
frequently, the laminae appear parallel to the terminal plane ; 
I was at first, and for some time, deceived by that appearance. 
In other Corundum crystals, the laminae appear to be parallel to 
the prismatic planes; and, to conclude the instances of analogy, 
the superposition of rhomboidal laminae is sometimes observable 
$ L 2 
