] 90 MR. T. H. HOLLAND ON ROCK-SPECIMENS FROM KOREA. 



seen in interlocked granules, and, on a smaller scale, smashed up 

 M-ith felspar into a polarizinoj mosaic. Bands of these minerals 

 alternate with bands of biotite and a finely granulated, micro- 

 crystalline material, in which it is impossible to individualize the 

 constituents, the whole forming the " Mortelstructur " of Tome- 

 bohm *. It may be observed that in the quartzes the bands of liquid 

 and gaseous inclusions run parallel to one another, and nearly 

 at right angles to the direction of foliation. From these rocks I 

 have separated numerous black scales, which, on chemical and other 

 examination, prove to be graphite. Dr. Gottsche also records the 

 occurrence of graphite in schists from various localities f; whilst 

 Mr. Carles mentions a "bed of graphite" occurring north of 

 Yongheung J ; and Capt. Hall noticed that some specimens of 

 schist, with a N.W. and S.E. strike, on a small island near Hutton 

 I., in lat. 36° 10' N., and long. 126° 13' E., " appear to contain 

 plumbago '^ §. 



Garnet occurs in eyes of the mineral and in isolated grains, which 

 I have been able to separate from the crushed rock by means of a 

 solution of borotungstate of cadmium, and test separately. Pluor 

 occurs in sufficient quantity to be determined in. some cases with 

 the blowpipe. 



In a specimen obtained about 10 li south-east of Tang-ji, blue 

 crystals presenting the characters of dichroite were abundant. On 

 testing the crushed rock in a solution of borotungstate of cadmium, 

 these crystals were found to sink in a liquid having a gravity of 

 2*55, and floated easily in a liquid having the gravity of quartz 

 (2 65). The blue grains were just buoyed up at a density of 2*59, 

 and on separation were found to possess all the optical characters 

 of dichroite. Having succeeded in making a chemical analysis of 

 the grains thus separated, I obtained the following result : — 



Silica 49-12 



Alumina 32-25 



Ferric Oxide 5-65 



Manganese Oxide tr. 



Lime tr. 



Magnesia 11*96 



Loss on ignition '32 



99-30 



Comparing these results with analyses of dichroite, there remains 

 no doubt as to the identity of the mineral from the Korean 



gneiss : — 



* G-eologiska Forening. Forhandlingar, 1881, vol. t. p. 233. 



1 Op. cit. p. 803. 



X Proc. Roy. Greogr. Soc. vol. viii. (1886) p. 305. 



§ Op. cit. Appendix, p. cxxiv. 



