628 APPENDIX 



distributed species, whose ranges here meet and overlap. The 

 saiga antelope {Saiga tartarica) of the Russian and Siberian 

 steppes, for instance, extends across Dzungaria as far as Long. 92° 

 East, but no farther ; the wild-ass {Equus hemionus) also ranges 

 as far eastwards, its place being taken farther east by Prjevalsky's 

 wild-horse. On the 90° meridium of Longitude, Prjevalsky found 

 that the ranges of these two overlapped. 



The southern border-range — the Tian Shan, together with 

 the Ala-tau and Barlik groups, constitute an isolated area in the 

 midst of the Central Asian zone, while in the east the Altai Mongo- 

 lian foims another subdivision which needs careful demarcation. 



A certain amount has been written and much conjectured, as 

 to the probability of the "overlapping" of the Altai and Tian 

 Shan fauna and flora in the ranges of Barkul and Karlik Tagh. 

 The traveller Grum-Grjimailo considered that he had estab- 

 lished a sure proof that the Barkul and Karlik Tagh Mountains 

 should be distinguished from the Tian Shan and included in the 

 Altai system (see Geog. Journ., vol. xiii, 1891) ; but I think that 

 his evidence goes to prove that these ranges form a neutral 

 ground between the two main mountain-systems. Orographically 

 they belong to the Tian Shan group, but, from the point of view 

 of their flora and fauna, they show affinities relating to both 

 systems. 



The Altai and Tian Shan mountain sj^stems approach each 

 other to within a hundred miles, the intervening space being 

 occupied by high steppe, which forms the first rise from the 

 Dzungarian lowlands to the Mongolian plateau. This eastern 

 extension of the Tian Shan chain, which approaches so closely 

 to the Altai Range and which extends far into Mongolia before 

 disappearing in the Gobi Desert, is composed of the ranges 

 called the Bogdo-ola, Barkul, and Karlik Tagh, and, still farther 

 east, the Ati Bogdo. It is somewhere amongst these mountains 

 that the true dividing-line exists between the two subdivisions. 



The western portion of this line has been already deter- 

 mined as lying between the Barhk-Urkashar and the Tarbagatai 

 groups ; thence it runs across the plains towards the south-east ; 

 on reaching the Baitik Bogdo the line divides into two, which 

 enclose a neutral zone before meeting again in the Western Gobi. 



