273 Mr. D. A. Bannerman on an Ornithological 



3. The size is smaller. In 12 specimens the wing- 

 xneasurements vaiy from 200-220 mm.^ and in 13 

 examples of C. I. livia from the British Islands and 

 southern Europe the wings measure from 217- 

 232 mm. 

 (When alive the birds strike one throughout as heing a 



distinctly smaller race.) 

 As already mentioned, there is no trace of hybridism in 

 birds from the Canary Islands, and they appear to be a 

 perfectly definable race, of which the characters noted are 

 always constant. C. /. canariensis is in my opinion more 

 closely allied to C. I. schimperi than to C. I. livia. 



Columha I. canariensis differs from C /. schimperi from 

 north Africa, south Arabia, and Nubia in having 



1. The general colour throughout decidedly darker. 



2. The lower back light grey instead of slate-grey. 

 In size the two subspecies are similar. 



Seven birds from the above localities have wing measure- 

 ments of 198-221 mm. 



Specimens from Palestine, Columha I. palesiince Zedlitz, 

 have the light upper parts of C I. schimperi but have the 

 lower back white, and in size they are apparently slightly 

 smaller. 



Mr. Stuart Baker (' Indian Pigeons and Doves,' p. 131) 

 believes that the Indian Blue Rock Pigeon {Columha livia 

 intermedia) extends through Persia, Arabia, Egypt and 

 northern Africa as far west as Tunis, and he does not 

 recognise C. I. schimperi Bonap. or C. I. palestince Zedlitz. 

 In these conclusions I do not agree with him, as I certainly 

 consider the light north African form is a perfectly distinct 

 subspecies. I have only a few examples from Palestine, but 

 these appear to bear out the characters assigned to this 

 subspecies. 



Rock-Doves are found in all the eastern islands of the 

 group ; they were generally met with in the region of the 

 coast but in much smaller numbers than in Gran Canaria, 

 where thev abound. 



