122 THE MIGRATIONS OF BIRDS 



offspring to permit the annual toll of death and to 

 leave a pair to begin the next breeding season, with 

 a certain ratio of additional individuals to fill the 

 gaps left among breeding birds that chance to be 

 killed before they have accomplished the act of 

 reproduction. 



It has been noted that many birds of continental 

 areas in the Tropics are sedentary, or that, where 

 there is wandering among them, the distance cov- 

 ered is comparatively slight and is traversed more or 

 less in the shelter of their usual cover. Storms and 

 the crossing of broad waters do not figure in the 

 migratory movements of these, nor are they in dan- 

 ger of becoming lost. As a corollary, we find the rate 

 of reproduction low; from two to four eggs consti- 

 tute a set, with two or three as the usual number. 

 Broods may come at irregular seasons, but appar- 

 ently it is normal for them to have but one breeding 

 period each year. Among exceptions to this rule 

 may be recorded honey-creepers of the genus 

 Coereba^ which in the West Indies rear brood after 

 brood throughout the year, and are in consequence 

 overwhelmingly abundant in comparison with other 

 species. This, however, is unusual among tropical 

 birds. 



The case of the small migrant perching bird of the 

 northern hemisphere is far different. Families of 

 three to six are the rule, with four or five as the aver- 



