28 JOURNAIv OF THE 



massive, g'enerally presenting- a g*neissic lamination. 

 The amphibole constituent has long* been erroneously 

 called "smarag-dite," solel}^ on account of its color. 

 Analyses show over 17 per cent, of alumina, and Pro- 

 fessor Dana calls it edenite in his new "Systemof Min- 

 eralogy." This rock forms dikes in the peridotites at 

 Buck Creek, Clay county. North Carolina, and is asso- 

 ciated with peridotites in many places in this and the 

 neig-hboring- counties of Georg-ia. 



No absolute contacts have been observed between 

 the peridotites or pyroxenites and the enclosing- g-neiss. 

 There is always a border of schistose talc, from two to 

 three feet in thickness, developed between them. Of- 

 ten there is also a variable zone of chlorite or vermicu- 

 lite, or of both tog-ether, intervening- between the talc 

 and the peridotite mass; and in many cases this bears 

 corundum in considerable quantit3\"'^ 



EVIDENCE BEARING ON ORIGIN. 



We have seen that the peridotites occur, as a rule, in 

 lenticular masses and sheets, having- their long'er axes 

 oriented with the lamination of the g'neisses and schists 

 of the country. Some of these lenses are short and 

 blunt, and occasionally the outcrops present a ver}" ir- 

 reg-ular outline and cover areas of several hundred 

 acres, as at Buck Creek, in Clay county, North Caro- 

 lina (Plate II). The smaller masses are in just the 

 form one would expect to find in small intrusions of 

 molten mag-ma into thoroug'hly laminated crystalline 



4- For more detailed description of the rocks and modes of occur- 

 rence aud distribution of both the peridotites and corundum, the 

 reader is referred to ''Corundum and the Basic Mag-nesian Rocks of 

 Western North Carolina," Bulletin ii, N. C. Ccol. Survey, and to 

 "Corundum of the Appalachian Crystalline Belt," Traus. Am. lust. 

 Mining Etigineers, XXV, Atlanta Meeting-, 1895, 



