212 THE MIGRATIONS OF BIRDS 



persal at New York are almost identical, as in that 

 year it was breeding north to the coast of central 

 Maine and south to northeastern North Carolina. 

 This indicates a concentric dispersal in these two 

 directions, though one would naturally suppose that 

 the bird would travel more rapidly toward the south. 

 There is evident wandering in late autumn and 

 winter, so that casual occurrences have been re- 

 corded recently from Wisconsin, Illinois, Tennessee, 

 and Mississippi. It is usually several years before 

 breeding colonies are established in new regions. At 

 Washington, D. C, in November, 1913, a flock of 

 several hundred arrived preceding a storm from the 

 north, and spent all one afternoon in whirling in the 

 wonderful evolutions customary to this species back 

 and forth over the Mall. Individuals were recorded 

 all through the winter, but it was several years be- 

 fore the bird became established commonly as a 

 breeding species. Now it is nesting all through the 

 country in near-by Virginia and Maryland. In the 

 United States the starling does not seem to have 

 established any regular migratory habits, but seems 

 to wander at random. It has been believed that a 

 certain number have been carried southward in 

 migrant flocks of grackles, cowbirds, or red-winged 

 blackbirds, as starlings in autumn frequently roost 

 with these species. This method of dispersal may 

 be probable in some cases, though it cannot be con- 



