Vol. 60.] LATER FORMATIONS SURROUNDING THE DARDANELLES. 253 



foci. Iii Iinbros. however, a date later than Oligocene and earlier 

 than Sarmatic can be assigned to the andesite-rlows in the south- 

 eastern portion of the island. 



At Keshan, the interstratification of Lower Tertiary sandstones 

 with andesite is distinctly visible. A section on the south side 

 of the Keshan ridge, which rises about 1000 feet from the plain, 

 shows, from below upward : shale and sandstone, about 40 feet of 

 andesite, 300 feet of sandstone, 30 feet of volcanic rock (probably 

 andesite), 200 feet of sandstone, and a coping of volcanic rock. 

 This section is exposed in a small ravine at Tekekeui, where there 

 are unaltered sandstones overlying the andesite, and within 2 feet 

 of the solid rock ; while under the andesite, similar sandstones and 

 tine shales are equally unaltered, even close to the contact-surface. 

 The andesites and sandstones have the normal dip of the sur- 

 rounding strata, that is, about 10° north-north-eastward. 



In Lemuos, Prof. L. de Launay has observed quartz- andesites, 

 dacites. trachyandesites. and augitic andesites; he con- 

 siders them to be perhaps Miocene, certainly later than the sand- 

 stones which they have dislocated and intersected (0. pp. 209, 219). 



At Hagio-Strati Island, 20 miles south-south-west of Lemnos, 

 I found nothing but homblende-andesite at the sea-level. 

 This island is about 1000 feet high, rugged and trackless : it 

 seemed to me to be a uniform mass of volcanic rock, without any 

 appearance of Tertiary sedimentaries (see Appendix I. p. 277). 



On Imbros I found that the whole south-eastern face of the 

 island, exclusive of the promontory of Megalai Kephalai, which is 

 virtually separated from it by a salt-lagoon enclosed by sand-ridges, 

 consists of grey and dark-red andesite. weathering into rounded 

 lumps, and forming a red soil between them. Following the valley 

 of Melano Potamo north-westward to the watershed, I found an 

 exposure of biotite-augite- andesite, differing considerably 

 in external appearance from the surrounding rocks, and splitting 

 into rectangular vertical prisms which resist weathering; its peno- 

 logical character is also peculiar (see Appendix I, p. 277). The 

 track passes for 3 or 4 miles over andesite. with occasional 

 obsidian-dykes, and well-defined flows of what must have been 

 semi-fluid material, which have curled over and now form large 

 caves. At the watershed, the track skirts a breccia with a matrix 

 of hardened volcanic mud. Fragments of porcellanite are 

 abundant, but the main mass, on both sides of the watershed. is 

 andesite, shown in fig. 3 (p. 254). The higher hills to the south- 

 west are apparently volcanic, and include several isolated domes. 



At the head of the Gulf of Xeros are several small outcrops of 

 volcanic rock, generally hornblende-andesite. on the synclinal 

 fold of the lacustrine sandstone-strata, and in one of these, at 

 Xero Mikro Island, I found schists lying horizontally under volcanic 

 rock. 



At Enos there is a large volcanic mass, forming a detached boss 

 about 1300 feet high, and roughly 6 miles in diameter : Yiquesnel 

 describes it as formed of trachyte and domite (3, p. 326) : and 



