NATUEAL HISTORY OF LAKE URMI. 351 



in height, with a stem 2-3 feet in diameter near its base. It 

 is, however, equally possible that the trees were planted when 

 the islands were inhabited and joined to the mainland. Pormerly 

 the larger islands seem to have been fairly thickly wooded, 

 but the fuel-collector has not only cut down the trees with an 

 unsparing axe, but now the sailors of the ships earn a living by 

 eradicating all trace of their existence by laboriously digging- 

 out their very roots. Walnut-wood is in great request for the 

 best joinery and cabinet-making. 



The banks of the watercourses are recognizable from afar on 

 account of the poplars and willows planted along them. The 

 willows are generally pollarded at regular intervals of time. 

 The long straight poplar poles are used as rafters for the roofs 

 of houses and balconies. Split laths are laid transversely from 

 rafter to rafter, matting is laid on them, then hay, and finally a 

 thick layer of mud and chopped straw. 



G-eology of TIrmi Basin. — The Azerbaijan lake-basin occupies an 

 area of some 20,000 square miles ; its greatest depression is more 

 than 4000 feet above sea-level. The mountains on its periphery- 

 separate its water-system from the circumjacent basins of the 

 Tigris, the Aras, and the Kizil TJzun, the latter two of which 

 flow into the Caspian Sea. Several peaks rise above 10,000 feet, 

 and the volcanic cone of the Savalan reaches 15,000 feet. Out- 

 crops of volcanic rock occur in many localities, and probably 

 had their origin during the vast upheaval of land which occurred 

 after the Miocene age. 



The geological record of Lake Urmi is still but very partially 

 read. The oldest rocks are Palaeozoic and probably of Car- 

 boniferous Limestone nge, because the genus Enclothyra is to 

 be detected among their Foraminifera (p. 452). The Jurassic 

 age is represented by a solitary Ammonite in my collections 

 (p. 418) ; but Eodler, von Borne, and others have noted several 

 localities near the west coast of the Lakeof Urmi where Jurassic 

 rocks are to be found. Eor the palaeontology of the later 

 deposits, I would especially refer the reader to Mr. Bullen 

 Newton's comprehensive retrospect on p. 430. 



The Carboniferous rocks of the Lake itself are overlaid by a 

 great thickness of Miocene chalk, an important section of which 

 •js referable to the Helvetian period, and which contains the 

 remains of organisms identical with those of the Lower Miocene 

 (pp. 430-452). The North Persian Tertiary rocks are divisible 



