i893. THE DISTRIBUTION OF BIRDS. 107 



There is another province, which commences, probably, in British 

 Caffraria, certainly about the Berea, near Durban, and extends along 

 the Drakensburg mountains, including the forest regions of the 

 Transvaal, such as the Lydenburg district, and reaches the Zambesi, 

 whence I believe it will be found to stretch across to Benguela and 

 Angola. 



Lastly, I propose to recognise a 



V. — Victorian^ or Camaroonian Sub-Region, 



consisting of the elevated mountains of Central East Africa, and 

 extending from the highlands of the Lake Regions across to the peaks 

 of the Camaroons. This is a natural sub-region and is characterised 

 by a number of forms. The recent collections sent home by Mr. 

 H. H. Johnston, C.B., show that this mountain fauna extends to the 

 Shire Highlands, as is evidenced by the presence of Cryptospiza and 

 the inevitable green XenocichlcE ^ which we now look for at a certain 

 elevation, and which are known from the Shire Mountains, Kilima- 

 njaro, Elgon, and the Camaroons. The occurrence of many peculiar 

 species and genera such as Pcedilorhynchus in the Uganda Highlands, 

 and again in the Camaroon Peaks, shows that there is a close con- 

 nection between the Avifaunge of these elevated mountains, and^ 

 doubtless, Kenia and Ruwenzori will be found to belong to the same 

 system. The highlands of Shoa certainly belong to it, and some of the 

 Abyssinian mountains will also, in all probability, have to be included. 



VL — Lemurian or Mascarene Sub-Region, consisting of 

 Madagascar and Adjacent Islands. 

 E. — Indian Region. 



This follows the usual boundaries laid down by Dr. Sclater and 

 Dr. Wallace. The division of the Indian Peninsula into provinces 

 will probably be determined by the mean annual distribution of rain- 

 fall, as shown by Mr. Hume in his interesting paper (" Stray 

 Feathers," vol. vii., pp. 501, 502). This paper has not received 

 half the attention which it deserves, and many of the problems con- 

 nected with the presence of a Malayan element in the western portion 

 of the peninsula and the mountains of Southern Ceylon become less 

 difficult of comprehension when we find that they belong to countries 

 where the rainfall is 70 inches and upwards. In the Himalayas 

 we have apparently the foci of three separate sub-regions ; there is a 

 strong Palaearctic element illustrated by the number of species of 

 Carpodacus, Pyvrhiila, etc. There are also recognisable two more sub- 

 regions. 



Another Sub-Region would be 



The Himalo-Chinese Sub-Region, somewhat as defined by Mr. 

 H.J. Elwes, and illustrated by such forms as Polyplcctron ; while a third 



1 Not named directly in honour of Her Most Gracious Majesty, but intended to 

 show that Mount Victoria in the Camaroons, and the Victoria Lake country la 

 Equatorial Africa, belong to the same natural sub-region. 



