54 ART. 2,— S. YOSHIWAKA : 



continuation of that of Formosa, as is explained hereafter, the strike 

 of the western part thus differs from that of the latter region which is 

 E — W. This is probahly due to a fault running X — S, which direc- 

 tion is also taken by a line of andésite eruption in Ishigaki-jima. The 

 strike of rocks in Yonaguni-jima again shows the existence of a fiiult 

 of the same direction running through its central portion. Previous 

 observers have assumed a great fault line parallel to it on the eastern 

 side of Formosa. These lines of weakness all running; X — S were 

 probably caused by a pressure independent of that which folded the 

 Riukiu Curve, and it is also probable that the former pressure preced- 

 ed the latter. The " Mayon system" of Prof. Koto* will perhaps 

 show some relation to these lines. 



VI. UNINHABITED ISLANDS IN RIUKIU. 



Under this heading will be treated the lîorodino and Pinnacle 

 groups, both lying at a considerable distance from the other islands 

 in the Riukiu Curve. 



The very small islets scattered at a distance of about 155 miles 

 to the south-east of the southern point of Okinawa-jima, and forming 

 the Borodino Group, are named Rasa, South l>orodino (Minami-öagari) 

 and North l^orodino (Kita-ôagari). The first is in lat. 24° 32' 30" ¥ 

 and long. 131° 19' E, the second in lat. 25° 55' N and long. 131° 14' 

 42" E, and the third lies about 6| miles to the north-east of the 

 second. Their length and width measure respectively 2.5 and 1 

 miles, 5 and 3 miles, and 3 and 2 miles. 'They are all composed of 

 raised reefs which end in perpendicular cliffs on the shore. 



Other small islets to the north-east of the Yaeyama subgroup, 



* Kütn, On the Geologic Structure of the Malayan Archipelago— -/oi/r/j. Coll. Sei. Imp. 

 Univ., Tokyo, Vol. XI, Part. II, 1899. 



