477 
1921.] Bird-Migration by the Marking Method. 
method, and even in the case of purely resident species the 
extent of local movements was considered worthy of study. 
Earlier schemes, perhaps more by chance than by intention, 
had been confined for the most part to a limited number of 
especially interesting species which had well repaid the 
labour and cost of marking : no guide, therefore, existed 
as to the possibilities of a more comprehensive plan of 
campaign. 
It soon became evident, nevertheless, that in many direc¬ 
tions the results were very meagre as compared with the 
number of birds being marked. The scope of the Inquiry 
was accordingly limited to a few species, selected for a 
variety of reasons. Three qualities were thought necessary 
for a species to be classed as a good subject : the birds had 
to be procurable for marking in large numbers, they had to 
afford a good percentage of reappearance records, and their 
migratory movements had to be of interest. Some of the best 
subjects under the first and second of these headings, such as 
certain gulls and game-birds, were unfortunately lacking in 
the third quality, while most of the small passerine birds 
showed an extraordinarily low percentage of reappearance 
records apart from cases where they were re-caught by the 
marker. The species finally selected were the Lapwing, the 
Starling, the Song-Thrush, the Blackbird, the Woodcock, 
the Mallard, and the Herring-Gull. To these were added 
the Redbreast and the Swallow, chiefly because a large stock 
of the appropriate size of ring remained, and also the Lesser 
Black-backed Gull in the expectation, never realised, that a 
large number could be marked. Had the work continued, 
the tendency would have been to concentrate more and 
more on the Lapwing and the Starling. 
The Interpretation of Results. 
In the following sections a few species which have 
afforded interesting and comparatively numerous records 
are discussed separately and at length. The method adopted 
in each case has been based on a system of grouping in 
