150 BRITISH BIKDS. 1vol. xv. 



figure can safely be used for comparative purposes unless 

 it is based on a total of many hundreds of marked birds. 



The unreliability of very isolated and thus possibly abnormal 

 records need not be laboured, but it may be worth while 

 drawing attention to the danger of laying stress on negative 

 evidence except in very clear cases ; the absence of records 

 from a particular area is significant only when it can be shown 

 that the species tends to yield a good proportion of records 

 under circumstances such as are prevalent there. This 

 point was found to be important in connection with the 

 retrapping of marked birds at the places of marking. Very 

 high figures are obtainable by this means ; without counting 

 any individual bird more than once, the Blue-Titmouse 

 yielded, in the Aberdeen inquiry, respectively 90.2, 53.3 

 and 42.2 per cent, of recoveries at three places where the 

 markers carried out sj-stematic retrapping, while a much 

 larger number marked at other places afforded only i.i per 

 cent, in all. 



Examples of the difficulties encountered in dealing with 

 the results of bird-marking need not be multiplied. Enough 

 has probably been said to point the moral that the method 

 is likely to be most fruitful where special attention is directed 

 towards definite problems anel towards those opportunities 

 of studying them which seem most likely to yield reasonably 

 numerous records of a kind readily capable of sound inter- 

 pretation. The writer hopes, accordingly, that the criticisms 

 which he has been led to make of his own methods may be 

 of some service to those who have been working on parallel 

 lines but who are still in a position to plan further develop- 

 ments of their schemes of investigation. 



[We shall welcome any further views on this subject and 

 would especially like to have the opinion of those who 

 "ring." — Eds.] 



