Territorialism and ** Sexual Selection'* 57 



common, instead of isolated castles, as it were, is correlated with 

 the adoption of special sites, which are naturally more or less 

 scarce. The weaver bird wants a tree overhanging a deep pool 

 of water (D). The pied starlings a piece of cliff-like, but not too 

 hard, river bank, the jackdaws a cliff or church tower and so on. 



Once after riding all over several adjacent farms I came at 

 dusk to a place where there was a bend of a stream against 

 rising ground ; a small krantz with a shallow cave. As we passed, 

 a little cloud of rock pigeons flew out. It was their only suitable 

 refuge in all the country we had traversed and the reason of 

 the development of strong flight and the homing instinct in such 

 pigeons was impressed on one in a flash. 



It must be noted that our non-territorial examples, though not 

 vividly coloured, have their patterns and markings, some of which 

 aie developed in the breeding season. In the male of the English 

 starling, for instance, the beak becomes of a colour very like the 

 evening primrose. This may be explained by what we may 

 perhaps call "limited territorialism," as such a colour would be 

 easily seen by a trespasser in a dark hole. 



In the same v/ay I should explain the iridescence on the 

 pigeon's neck and his habit of puffing out the crop. His dis- 

 play, in which he coos and wheels and runs round his mate, I 

 should describe as a very natural way of proclaiming her his 

 property amongst a crowd. 



Both territorials and non-territorals use their voices but if we 

 take the thrush or nightingale as the type of the plain-coloured 

 territorial which proclaims itself by song we see at once there is 

 a type of song that we may call "proclamatory," we also observe 

 that the producers of this type of song constantly repair to some 

 point of vantage from which to deliver it, usually at dawn and 

 near sunset (E). 



This presents a contrast to the chatter or chattering songs of 

 jackdaws, starlings and weaver birds. 



Our theory in no way bars the probability that both sound and 

 colour stimulate the joy of life in birds as most things in Nature 

 are put to more than one purpose. The voice in birds is certainly 

 used to express many simple emotions and responsive communica- 



