20 



BULLETIN 185, U. S. DEPAETMENT OF AGEICULTUBE. 



central California, where it becomes common before the first anivals are usually 

 noted in Texas or Florida. The route the species takes from Brazil to California is 

 one of the yet unsolved migration puzzles. 



Since the above was written much additional information has been 

 obtained on the movements of this species, and now it is possible to 



T 



{^"^ 



M* 



BREEDING RANGE 

 ^^^m WINTER HOME 



- — — — Migration rouTes of Atlantic wintering bircis 

 —...— Known port of mi^ ration route of FhclFic winTerinjy birds 

 XXXX Supposed mii^rdfion route of Pacific winte ring birds 



Fig. 8. — Distribution of the wliite-winged scoter {Oidemia deglandi). Birds wintering on the Atlantic 

 coast follow an elliptical migration route, passing eastward in fall to the Labrador coast and returning 

 the following spring by a route from Long Island Sound northwestward directly to the breeding grounds. 

 Birds wintering on the Pacific coast are kno'mi to migrate spring and fall along the coast between Wash- 

 ington and southern Alaska, but where they cross the mountains can only be surmised, for they have 

 never been noted any^vhere on the 500-mile strip between the Pacific Ocean and the Mackenzie River, 

 and apparently this part of the migration route is accomplished at a single flight. (See p. 21). 



solve this migration puzzle. It is now known that the cliff swallows 

 go around the Gulf of Mexico instead of across it. The isochronal 

 lines on the migration map show that the birds advance along the 

 Pacific coast of Mexico faster than along the Gulf side, so that on 

 March 20, when the van has not quite reached the lower Rio Grande 

 of Texas, it is already far north in California. 



