S. Yoshiwara — On the Bonin IslancU. 301 



Andesite and tuif are the chief rocks composing the Ogasawara 

 group. The latter are always inclined in various directions, showing 

 great disturbance after their formation. We find great limestone 

 exposures, resting in quite horizontal layers, in some parts of the 

 Haha-jima and Chichi-jima sub-groups. The high hill of Sekinion- 

 yama in Haha-jima is entirely formed of this rock, which is besides 

 found in separate areas at Oki-mnra and Akaiwa (Sankaku-iwa). 

 In Chichi-jima the same limestone forms the southern point of land 

 called Minami-jaki, as well as Minami-jitna and many neighboui'ing 

 rocks. The limestone is always in elevated places, especially that of 

 Sekimon-yama, which is found at a height of about 800 feet above sea- 

 level, and forms two successive vast terraces along a hill of agglomerate 

 tuff which is about 100 feet higher than the limestone plateau. 

 Here the limestone extends towards the interior for a few miles, and 

 was originally an elevated coral reef. LithotJiamnium, Bryozoa, and 

 corals are predominant in the limestone. Numerous Orbitoides are 

 also found in the same rock at several places. In Minami-jima 

 I found at least three species, one of which is very thin, with the 

 very large diameter of 47 mm. In Oki-raura in Haha-jima, at least 

 three species of small Orbitoides were also collected. A few 

 individuals of one small species have been taken from the limestone 

 of Minami-zaki. All the Orbitoides above mentioned belong to 

 Zepidocyclina ; ^ some of them are identical with those from the 

 Miocene of Java described by Verbeek. Thus the submarine volcanic 

 eruption of the Ogasawara group began in the Eocene epoch, and 

 had already ceased before the Miocene. This old volcanic chain 

 must therefore be separated from the Fuji chain, which is younger 

 Neogene or partly even still younger in origin. For the old chain 

 I propose the name of " Ogasawara chain." No indisputably Eocene 

 volcanoes corresponding to the age of this chain have yet been found 

 in Japan, especially near the Fuji chain. 



The Fuji chain certainly extends as far as Hachijo-jima. On its 

 south there are, in a line with the Fuji zone, the islets of Aoga- 

 shima, Bayonnaise Eocks, Smith's Is., Tori-shima- (Ponafidin Is.), 

 and Lot's Wife Eock, all of which are composed of andesite 

 rock. No land is found between the last islet and the Ogasawara 

 group. When we examine on a map the relative positions of the 

 islands near the Ogasawara group, we find them in two rows. The 

 first row passes from the sub-group of Muko-jima to that of Haha- 

 jima in a line from north to south, at the longitude of about 142° E. 

 The other islets, viz. Eosario-jima on the west of the Ogasawara 



' Localities of Orbitoides in Japan had been coutiued to Nakakosaka iu Province 

 Kotsuke and Kawaguchi in Pro-vince Kai, prior to my joiu-ney to Formosa and Riukiu 

 (Loochu) Islands in 1899. Specimens found iu Nakakosaka have been determined hy 

 Jones as 0. [Discoci/clina) dispansa, Sow., but afterwards as LepidocycVma by Verbeek. 

 I collected numerous specimens from the Kmestone iu Iriomote-jima iu Rmkiu (Jouru. 

 Coll. Sci. Imp. Univ. Tokyo, vol. xvi, pt. 1, art. 2, 1901). By my latest researches 

 I am convinced that all these Orbitoides of the above three localities belong to 

 Lepidocyclina. Besides, Messrs. R. B. Newton and R. Holland found a species of 

 Lepidocyclina {0. Verbecki, Newt. & Holl.) iu a slide of Formosan limestone (Journ. 

 Geol. Soc. T5ky6, vol. \n, No. 81, 1900). 



2 Spelt Tosi-shima on map, Fig. 1, p. '298. 



