34 THE CROW FAMILY 



A most interesting piece of evidence bearing on the point is 

 supplied, curiously enough, not by rooks but carrion-crows. A 

 German observer noted that a band of eight of these birds were 

 in the habit of frequenting the edge of a pond, which on three 

 sides was screened by a thick growth of osiers, high enough to 

 conceal the approach of an enemy. Outside the osier belt was a 

 post, and on this one of the crows was seen perched not only once 

 or twice, but many times during a period of several summer weeks. 

 When alarmed it uttered a loud note of warning, and its comrades 

 rose at once and flew off in silence, followed by the sentinel. But 

 even in this case it is not certain that the bird was actually engaged 

 in outpost duty. Much closer observation is required before the 

 sceptical can be convinced. Is it the same bird that stands sentinel 

 each time ? If so, why it and not another, and what time does it 

 give to watching and feeding respectively? Or is there division of 

 labour, and, if so, what determines the division ? l 



In the short winter days rooks and daws begin to move towards 

 their sleeping quarters long before dusk. Rooks almost always roost 

 in a clump of trees. Records of their sleeping on the ground are 

 rare. 2 Sometimes our resident birds continue to roost during the 

 winter at the rookery, this term being here used exclusively to mean 

 the nesting trees. Usually they appear to cease making the rookery 

 their dormitory in September, not returning to sleep there till the 

 following spring, and repairing, in the meanwhile, each evening to 

 some neighbouring wood, sometimes a considerable distance away, 

 which may often serve as a common roosting - place for several 

 different flocks. It is instructive to note that this change of sleeping 

 quarters coincides with the date of the general autumnal migrations, 



1 For the account of the crow sentry see the Zoologisches Garten, 1904, p. 377 (Dr. Horning). 

 Accounts of flocks of geese putting out sentinels are given both by Mr. St. John, quoted in 

 L. T. Hobhouse's Mind in Evolution, p. 272, and by Mr. J. G. Millais, in his Wildfoider in Scot- 

 land, p. 01. But the difficulty in unreservedly accepting these statements is that they involve 

 not merely a question of fact, but one of interpretation. It is not impossible to give to what 

 one sees a meaning it may not really possess. Much more evidence than is available is needed 

 to carry conviction. 



2 Field, 1886. 



