INTRODUCTION 15 



them have been shot and the bands recovered. The results show that this group of 

 birds is almost sedentary, but the interesting point is that it is not absolutely so, for 

 one or two per cent of them have turned up, and apparently as breeding birds, in 

 places as far off as East Prussia. So we see another reason why local breeding groups 

 do not become differentiated into subspecies, for they are not absolutely local and a 

 certain leakage or interchange with other groups appears to take place regularly. 



I have said that migration back to the breeding area is not always over the same 

 course by which the birds departed. Most of our North American ducks breed in the 

 great area from the Dakotas, through Manitoba, Saskatchewan, and Alberta, to the 

 Athabasca and Mackenzie Valleys. From here they radiate in many directions. 

 One great group of Mallards, Teal, Widgeon, and diving ducks strikes east, and, part 

 of them collecting at certain spots in the Great Lakes region, eventually arrive on 

 the Atlantic Coast, mostly south and west of New England. Some of these birds, 

 particularly the Mallard, gradually withdraw from the sounds of North Carolina 

 during the winter, so that after the middle of January they are almost absent, and 

 these birds undoubtedly go north by a more inland route. In the Mississippi Valley 

 there is a greater movement north in the spring than south in the autumn, while on 

 the North Atlantic coast certain species are practically absent in the spring in the 

 same places where they are common in the early autumn. We must picture, there- 

 fore, a great circular journey, which no doubt was first developed through necessity, 

 for the States of the Mississippi Valley are often so dry in October that only a limited 

 number of ducks can stop there, while in March and April the great rivers are in full 

 flood and provide optimum conditions for all water-loving birds. These circular 

 journeys are not confined to ducks. They are well seen in the Golden Plover and in 

 the Eskimo Curlew, although of course these waders continue much farther south 

 than do our ducks. Certain warblers also have a different route in the spring. 



It is true in a general way that the Anatidce are less bound by a hard-and-fast time- 

 sense during migration than many of the passerine birds, and yet they are more 

 accurate in their dates of arrival than is commonly supposed, for the mere fact of 

 their non-appearance at a given point does not mean that the species is tardy. It is 

 quite as likely that they failed to stop because water conditions were not suitable, or 

 because meteorological conditions were so favorable that the flight kept to sea, or 

 made the journey without breaks. Arrival of the Lesser Scaup in eastern Massachu- 

 setts is very regular. In my own records October 6th was earliest, and October 19th 

 the latest, in twenty years. The average is October 12th, and as a matter of fact, in 

 spite of small numbers, these ducks arrived on that date for six of the twenty years. 



The latitude at which a species winters is, of course, dependent to a great extent 

 upon the severity of the weather. On our Atlantic Coast, the great sounds where 

 diving ducks commonly winter may be completely closed with ice, and when this 

 happens there is a secondary flight, perhaps for only a short distance, but possibly 



