32 ART. 2. — B. koto: 



reason why the soil is rich in soda and poor in fossil remains in 

 tlie Upper Kybwj-sang formation, What 1 call the marl gold 

 occurring in the said rock is due, it seems to me, to the same 

 cause, and the ore-bringers are in all probability the diabasic rocks 

 which never fail to appear in the formation. 



One more hill-neck, called Mal-clihi '\ had to be crossed, and 

 from it we looked dowai (PI. III. fig. 3) at our destination, Chin- 

 jgu, 18 km from Pan- song. The rocks w^ere the same as those at 

 the ferry. Weathering had been working here deep into the roots 

 of the mountain, producing a thick eluvial cover of red earth. The 

 country would have been long ago base-leveled, or i-ather beveled, 

 were it not for the presence of the hard gray sandstone which 

 is intercalated with marls. 



I must once more lay stress on the influence of the nature 

 of the rocks upon the land-features. From Pan-song hither, I 

 traced the red formation"-' till we came to the ferry; and thence to 

 Chin-Jgu we saw the underlying beds of grag marl and sandstone.^'' 

 The red complex underlies the flinty tuflite and slate, and another 

 complex of the green volcanic tiifis and breccias *', which build up 

 the region between Pan-shng and Fa-san, so that as we go east- 

 wards w^e ascend the geologically younger horizon. On the other 

 hand the non-volcanic red and grag formations extend meridionally 

 northwards for one and a half degrees as far as Sang-jgii along 

 the eastern flank of the Chiri-san range with breadth of 30 km. 

 The whole l)elt presents wdiat the geographers of the Davis school 

 call the mature and old-age topography. For the reason that 

 the marls and sandstones of the west being soft as compared with 

 the greenish-gray volcanics of the eastern half of Kgöng-sang-Do, 



1) W^l ^^ '^) ri. xxxiv. I. Traverse, No. 'i (nil) iu the profile. 

 3) The Biime, Nos. i ami 5 (ms, sdra). 4) The same, No. 2 (sh). 



