148 HABITS OF PURPLE MARTINS IN ARIZONA 



of the case obscure the whole subject. It may be roundly 

 asserted that we have yet to discover where the great mass of 

 Martins bred each year in the United States stay in winter. 

 Our birds are known to come over our border very early in the 

 spring or in February, and gradually spread over the country, 

 reaching the highest latitudes by the middle or latter part of 

 May. Such early appearance subjects them to painful vicissi- 

 tudes of the weather, large numbers having been known to 

 perish in sudden storms or cold snaps. The return movement 

 is less regular ; it begins early, in August as a rule, but is not 

 finished for a month or more. The breeding range of the 

 species coincides with the whole distribution in North America 

 wherever suitable nesting- places can be found, and the bird is 

 moreover resident in some portions. 



But I need not enter upon the full history of so very familiar 

 a bird. Its natural nesting places are hollows of trees and 

 rocks ; the martin-box, or its equivalent, is an innovation to 

 which these progressive birds take very kindly, and hardly 

 any other way of nesting is known in populous districts, where 

 the Martin rivals the Barn and Eave Swallow in domesticity, 

 and surpasses the White-bellied by a long interval. In the 

 West, the case is diflferent, or was when I studied Martins in 

 the then wilderness of Arizona. At Fort Whipple, these birds 

 were very abundant summer residents of that pine-girdled 

 locality, arriving early in April, and taking leave late in Sep- 

 tember. They lived in colonies, and had their being in what 

 might be called martin-houses, made by very industrious and 

 skilful carpinteros — the Woodpeckers. Many a towering pine 

 was dead at the top, like a senile ornithologist whose body had 

 outlived his head, and stood the picture of grotesque despair, 

 with a load of useless lumber in the upper story. Hither came 

 the Woodpeckers — especially Lewis's and the formicivorus — to 

 build baby-houses, like children trooping into the garret. Ko 

 more eligible martin houses could be found than such honey- 

 combed shafts, and the birds knew it. Many such picturesque 

 establishments were fully tenanted by numerous pairs of Mar- 

 tins, who did not seem to confine themselves to deserted Wood- 

 peckers' holes. When once a settlement was effected, the 

 Woodpeckers, who might be inclined to occupy some of their 

 own premises, had notice to quit, and they generally governed 

 themselves accordingly. Peace at any price seemed to be their 

 shrewd conclusion ; for though they might manage to live, as 



