KATURAL HISTOET OE LAKE UEMI. 347 



"World meet. The divides which separate three great systems of 

 rivers — namely, the river-sysitems of the Indian Ocean from those 

 of the Atlantic, the river-systems of the Atlantic from those of 

 the Arctic, and the river-systems of the Arctic from those of the 

 Indian Ocean — meet in the highlands of Armenia and Kurdistan, 

 between the massif of Asia Minor and the Persian plateau. 

 Notwithstanding the altitude of the region, tlie meteorological 

 conditions tend towards those of a desert, and produce features 

 eharucteristic of deserts. On the whole the annual evaporation 

 is so much in excess of the annual precipitation that the lakes 

 do not overflow, and their water is more or less salt. Similarly, 

 in Turkish territory we have Lake Van, in Russian territory 

 the Caspian Sea, and in Persia Lake Urmi. 



Climate. — The climate of the Azerbaijan plateau may be fairly 

 classed among the continental, or " excessive " climates as they 

 liave been termed by Buflbn ; but at the same time it is the 

 Healthiest in Persia, because the summer heat is more tempered 

 than in the other provinces and there is generally a sufficiency of 

 water for man and beast. 



The seasons are extreme, but very variable both as regards 

 duration aud time of year. This variability is largely due to the 

 fact that these uplands of N.W. Per&ia, of 4300 feet altitude and 

 upwards, form part of a tongue of high land which runs from 

 N.W. to S.E., and divides the warm depression of Mesopotamia, 

 in which the palm-tree flourishes, from the deserts of Persia and 

 from the hot South-Caspian provinces of Grhilan and Mazenderan. 

 Thus we find, in close proximity, the snow-mountains of Kurdistan 

 and tropical lands warm enough for the growth of the date-palm. 

 It is a pity that methodical meteorological observations in so 

 interesting a region should be practically non-existent. In 

 default of the more accurate indications of the meteorological 

 conditions, I am compelled to supplement my remarks by such 

 observations on animal and vegetable life as seem to afford 

 information on the climate. 



The winter snows may commence towards the end of November 

 (23rd in 1890), or they may not fall until February 1st (1893), 

 There is often fioe weather between the first snows and those of 

 January and February. The last snows may fall as late as Lady- 

 day. In cold January weather the minimum thermometer some- 

 times falls below 0°Pahr. (12° P. is a common temperature), and 

 the snow does not melt in the middle of the day. As a general 



27* 



