CHRYSOBERYLS FROM HADDAM AND BRAZIL. 1 17 



son,* vvlio quotes Klaproth's analysis, and states that he ex- 

 amined the niiiieiul some years a2;o, Imt haviiiu; aecidentally 

 lost his results, he was unable to publish tiiem. He ot)- 

 serves, liovvever, that tlie only constituents he lound were 

 alumina, silica, and oxide of iron. When I was about to 

 prepare the comtnunication wiiicli I now have the honour to 

 lay before the Society, a more recent analysis of tiie Chry- 

 soberyl of Brazil, by M. Augustus Arfwedson, was observed, 

 by me, in Tilloch's Philosophical Maa;azine.t He conhrmed 

 the results of Professor Thomson, and considered the che- 

 mical composition of this substance to be — silica, 18.73; and 

 alumina, 81.43, with a trace of oxide of iron. 



The cymophane, from Fladdam, was sent to M. Hauy by 

 the late I)i'. Bruce, in 1810, to have his opinion concerning 

 its nature. t Previous to that period, the mineralogists in 

 the United States supposed it to be Cnri/rulum. The late 

 celebrated crystallographer observes, " La cymo|)hane des 

 Etats Unis a d' abord et»^ prise pour une variete de corin- 

 don. Effectivemcnt elle se rapproche de ce mirieral parsa 

 durete, par sa pesanteur specilique, et me me par le resultat 

 de son Analyse, qui a donne environ 7^ parties d' ahiminc 

 sur 100, avec 18 de silice, ct 6 decliaux.".^ I was anxious 

 to examine the cymophane found at Haddam. especially as 

 M. Hauy does not name the author of the analysis he quotes. 

 The specimen used for my experiments was of a pale green 

 colour. It did not present any of the chatoyant appearance 

 so remarkable in the variety from Brazil, and some speci- 

 mens from Saratoga in New York, where it was lately dis- 

 covered by Dr. Steel. Its s|)ecific gravity. In' two trials, was 

 3. .508 and 3.;'597. It is not magnetic, and t)efore the blow- 

 pipe it is infusible. For a further description of the phy- 

 sical characters of this mineral, I refer to Hauy and Cleavc- 

 land. 



' Thomson's Chemistiv, vol. iii. p. 213. 



t No. for November, 18'23, p. oST. 



1 Annates du Museum d' Hntoirc Naturcllr, tome xviii. p. 37. 



§ Traite de Mincralogic, 2me Edition, vol. ii. p. .lOO. 



