•$ ■ On the eryJialUni Forms of Corundum, 



In thefe three varieties, the cryftalline lamina can be feparated, as in the hexagonal prifin, 

 at the three folid alternate angles of each end, but in a contrary direction to each other. The 

 planes which appear when the lamina are detached regularly, form folid angles of 22" 34' 

 with the plaiies of the extremity : this arrangement is analogous to that of the hexaedral prifm, 

 '1 he difference of form arifes from the cryftalline laminz depofited on the planes of the pri- 

 mitive rhomboid, decreafing by more than one row of molecules, on the planes of one of the 

 triedral pyramids of the rhomboid, and by lefs than one row, on the planes of its other pyra- 

 mid. This general obfervation, on the manner in which this primitive cryftal of corundum 

 paffes to the different varieties jufl mentioned, is the only one I have eftablifhed with any de- 

 gree of certainty at prefent. Specimens with perfeft cryflals, whofe angles may be meafured 

 with accuracy, will probably arrive from India, and give further demonftration, as to thefe and 

 other varieties of modifications of corundum. We may conceive, that if in this modification 

 the cryflallization had ceafed before the intire formation of the cryftal, there would have 

 remained fmall ifc/celes triangular planes, on three of the alternate foUd angles, formed by the 

 juncElion of the planes on the ends, with the edges of the truncated pyramid. Thefe ifofceles 

 triangular planes referable thofe we have feen in the firft modification (fig. 4. and 5 ) , and form, 

 in the fame manner, folid angles of 122° 34' with the planes on the ends of the prifm 

 (fig. 16.) 



Finally, if during the formation of the cryftal, in this mpdification, it fliould happen that the 

 laminae depofited on the three planes of the rhomboidal parallelopiped, on the fide where they 

 undergo a greater decreafe, do not undergo th« decreafe of one row of molecules at the acute 

 angle of the fummit, the cryftal will be a real hexaedral pyramid (fig. 17), whofe acute angl« 

 at tfie fummit, meafured on the fides, will be nearly 24°, in one of the varieties ; 40° for the 

 moft obtufe; and 20° for the moft acute variety: the angle of their triangular planes, in the 

 firft hiftance, 13° 41'; in the fecond, 22° 20'; and 11° 28' in the third. I have not feen 

 any perfeft pyramids ; but in many the hexagonal plane terminating the pyramid is fo fmall, 

 that it renders its total fuppreffion probable. 



This decreafe necefTariiy produces a fingle pyramid, as above mentioned ; neverthelefs, there 

 are inftances of cryftals of corundum, belonging to the variety where the terminal planes 

 make^ with the planes of the pyramid, a folid angle of about 100°, in which two pyramids of 

 the fame dimenfions, hav ing their fummit replaced by a fmall hexagonal plane, are placed bafe 

 to bafe. 



I have alfo obferved among the cryflals of the obtufe variety above mentioned, in Mr. Gre- 

 ville's coUeftion, an inftance of the decreafe tailing place by feveral rows, on one three=.fided 

 pyramid of tiie primitive rhomboid, and by fingle rows on the other. Confequently, the 

 cryftal is a fhort regular hexaedral prifm, terminating on one end only by an hexaedral pyra- 

 mid ; the planes of which, as well as of the prifm, are alternately broad and narrow, and almoft 

 perfect, its apex being replaced by a very fmall plane. 



I (halt conclude, by mentioning a variety of corundum, defcribed by the Abbe Hauy, in the 

 Journal des Mines, No. 28; in which the edges of the terminal planes of the hexaedral prifm 

 are replaced by planes which form an angle of 116° 31' with the terminal planes ; but in the 

 numerous coUeition of Mr. Greville I have not feen this variety. One cryftal had an ap- 

 pearance 



