THE PETROLOGY OF SOUTH GEORGIA. 829 



latter can be shown to be continuous with that of the felspar crystal from which 

 the mineral is derived, and is, in a sense, inherited from the felspar. In the same 

 way the scapolite appears in some cases to retain the biaxial character of the felspar. 

 The mineral, perhaps, should be regarded as intermediate between felspar and 

 scapolite. It presents some points of resemblance to the scapolite of Oedegarden, 

 Bamle, Norway, which has also been derived from felspar. According to Judd, 

 this scapolite also frequently retains the twinning of the felspar from which it is 

 derived, and shows great variation in the strength of double refraction. In the 

 South Georgia rocks it is probable that many grains represent incomplete stages in 

 the process of the conversion of felspar to scapolite, and that their properties may 

 likewise be intermediate.* 



The process frequently begins with a heightening of the polarisation-tints of a 

 crystal of felspar, a heightening which is too general to be attributed to a local 

 thickening of the slide. The mineral then usually breaks up into more or less 

 rounded grains having many of the distinctive characters of scapolite. The altera- 

 tion rarely seems to have any relation to cleavage or other fissures, or to the exterior 

 boundary — or at least, such connection can rarely be traced. The final stage is a 

 confused aggregate of rounded or rudely prismatic crystals of scapolite occupying 

 the area formerly occupied by a felspar crystal. 



The alteration is very sporadic in character. Thoroughly scapolitised felspars 

 may be adjacent to crystals totally unaltered. The small felspar crystals of the 

 groundmass have also undergone sporadic scapolitisation, and it is a remarkable fact 

 that the radiolarian tests occurring in these rocks have frequently been changed to 

 scapolite. The presence of scapolite was demonstrated chemically by digesting the 

 rock-powder with dilute nitric acid and a little hydrofluoric acid, and testing the 

 filtrate with silver nitrate solution. A considerable precipitate showed the presence 

 of chlorine. t The rock was previously searched for apatite, the only other mineral 

 containing chlorine likely to be present, but none was found. 



Though carefully sought for, no indications were seen as to the manner in which 

 the accessions of sodium chloride and lime necessary to convert the felspar into 

 scapolite were conveyed into the rock. There is no trace, for example, of the 

 cavities filled with sodium chloride solution, which were found by Judd in the 

 felspars of the Oedegarden rock. Judd considers the chemical reactions necessary 

 to form scapolite from felspar are facilitated, if not effected, by the stresses suffered 

 by the rock during dynamo-metamorphism. The South Georgian scapolite tuffs 

 have also withstood considerable pressure, resulting in an imperfect cleavage. 



The tuffs are interbedded with, and pass into, very fine banded shaly material. 

 In thin section this is seen to consist of excessively fine volcanic material, mixed 



* Mineralogical Magazine, viii, 1889, 186-98. 



t I am indebted to A. Fleck, B.Sc, of the Radiometric Laboratory, Glasgow University, for carrying out 



this test. 



