324 NATATORES. 



quit our shores, or of their breeding places, we are not so 

 well informed. 



The history of most of the race is, indeed, imperfect and 

 unsatisfactory. Many which have been described as breed- 

 ing in the arctic regions, have not been seen there by those 

 who have had the best opportunities of observing. But, as 

 was formerly said of the pine-forests in the case of the 

 crested tit, the marshes of the artic regions, and indeed of 

 places without the arctic circle, are not very easily explored. 

 From the few of whose habits we do know something, it may 

 be inferred, that all the inland-breeding species keep them- 

 selves very close in the breeding season. The species which 

 have more of the sea character, collect at their favourite 

 places on the shores to breed, and are found in considerable 

 numbers in the same locality; but if we may judge from the 

 example of the common wild duck, or mallard, those which 

 breed inland, disperse themselves over the country, find 

 their summer food among the roots of the same aquatic 

 plants in the cover of which they have their nests, and thus 

 pass the summer unknown, not only in the fastnesses of the 

 more extensive and inaccessible marshy pools, but on the 

 banks of rivulets when these afford cover, without being 

 much seen. I have known a mallard to be taken by an 

 angler's hook, on the sedgy bank of a small stream, in the 

 breeding season ; though, to ordinary observation, the birds 

 had all quitted the district for several weeks. But we have 

 to do with them only in so far as they are British birds, 

 open to common observation and that is as winter visi- 

 tants, appearing and disappearing in their seasons, but of 

 whose retreats, while they are absent, we know not very 

 much. Thus we are unable to generalize them in anything 

 like a satisfactory manner, but must take those traits of 

 each genus that are best known, and leave their history to 

 future observation. 



