On the cryjlalline Forms of Cerundum. 9 



pearance of fuch planes ; but, on examination, it was clearly accidental. The autriority of 

 the Abbe Hauy in cryftallography, is fo great, that the exifteiice of fiich modification ought 

 not to be denied without further examination ; though I cannot, in this inftance, adopt it: he 

 derives this variety, which he calls fubpyramidal, from a decreafe of three rows of molecules, 

 at the angles of the bafe of the two pyramids of the primitive rhomboid; and hefeems to attribute 

 the fame formation to the pyramidal variety with double pyramid, which he fuppofes may exift. 



The primitive cryftals, and the firft and fecond modifications of corundum, are from the 

 peninfula of India. The third modification, or the pyramidal variety is from China ; nothing 

 approaching this form being among the fpecimens which Mr. Greville received from the 

 peninfula of India. 



The preceding obfervations, and particularly the laft-mentioned modification of corundum, 

 compared with the beft defcription of the fapphire, fuggeft the further examination of the degree 

 of connedlion, if not of identity, of thefe oriental ftones. 



In both, the hexaedral pyramids are ufually incomplete in their apex, and they vary in acute- 

 nefs. I have ftatcd the degree in which the folid angles of the pyramid (taken as complete) 

 vary, in corundum, to be from 20" to 40''. 



Rome de L'Ifle ftates, that the fapphire varies from 20" to 30". The Abbe Hauy (Journal 

 de Phyfique, Aug. 1793) mentions two varieties of the fapphire, one meafuring at the folid 

 angle of the pyramid 40° 6', the other 57° 24'. I never faw a fapphire with fo obtufe an 

 angle as the laft, but many whofe angle at the top, if the pyramid had been complete, would 

 have been the fame as that of the corundum. Befides the analogy between the cryftals of 

 corundum and the fapphire, by the union of two hexaedral pyramids at their bafe, it alfo exifts 

 by the meafure of their angles ; and both fubftances are fubjeil to the fame irregularity, 

 fometimes appearing as a fingle hexaedral pyramid, and fometimes as an hexaedral prifm : 

 iporeover, the fapphire fometimes has on its folid angles, alternately, the fame triangular planes, 

 (^g* 5')> ^^^ ^"o the prominent triangles on the planes of the extremities (fig. 10.), which 

 often appear in the cryftals of corundum. The Abbe Hauy, in the Journal de Phyfique, 

 Auguft 1793, names this variety, orientale entieagofie which is reprefented in the annexed plate 

 (fig. 18.), and fays, that the fmall triangular planes make with the terminal planes an angle 

 of 122°. 18'; and, in the defcription of the fame triangular planes in the corundum (fig. 16.), 

 it appears that thefe planes are the remains of the planes of the primitive rhomboid, and form, 

 with the terminal planes, an angle of 122° 34'. 



Perhaps the rhomboidal cryftal, which Rome de L'Ifle had given as one of the forms of the 

 fapphire, (hould be reftored to it. He had examined it at M. Jacquemin's, jeweller to the 

 crown (Criftallographie, i edit. p. 221), and he fupprefled it in his fecond edirion, but often 

 exprefled to me his regret in having made the alteration. I have before me a letter from that 

 celebrated naturalift, dated September, 1784,* in which he inclofed, for my opinion, a copy of 

 a letter he had received from Mr. Werner, with models of fome cryftals ; among them 

 two called by him rubies ; one a rhomboid, of which the angles of the fummit are fubftituted 

 by planes (fig. 19.), the other is precifely the fame as fig. 3, 4, and 5, of the annexed plate. 



* A letter to the fame effeft was written to Mr. la Metherie, and publiflicd in the Journal de Phyfique, 

 May, 1787. 



Vol. III.— Aprii 1799. C The 



