IS BlLLKTIX OK THE 



Will hold, that will he of some advantage. IMie map showing the distribution 

 of the Prothonotary Warbler has been discussed. A similar map, Fig. 2, 

 of the migraut records of Kirtlondi. has been prepared for purposes of com- 

 parison. The approximate localities are indicated by the arrows. As it is 

 well known that valleys are highways of migration for many species, as in 

 the case of the Prothonotary Warbler. 1 have therefore connected the 

 arrows by the broken lines, to indicate approximatel}' the supposed routes. 

 Dr. Stejnegcr has suggested that the limited number of birds favors the 

 idea of a single narrowly limited route. This is especially likely to be the 

 case south of the mouth of the Ohio river. Above that place the route 

 apparently l)ranches, but it is highly probable that a narrow route is fol- 

 lowed. Our lack of data from South Carolina to St. Louis leaves a great 

 gap in the route. As the bird breeds among conifers it may be influenced 

 by such tree's in its migration, in which case the pine barrens of the Coastal 

 Plain and the cypress swamps of the Mississippi valley, to the mouth of 

 the Ohio, may be shown some preference. I do not understand the South 

 Carolina records: perhaps they are stragglers from the general Gulf Coast 

 pnie barren route to the Mississippi river. Further data is necessary to 

 settle this point. At least it would be worth while to search the pine 

 barrens of the (iulf Coast during the last of April or the first week of 

 ^[ay for the bird. So much for a working hypothesis. From the foregoing 

 remarks it will be seen that the estimated route is somewhat as follows : 

 The birds spend their winters in the Bahamas, reach the coast of the United 

 States about the latter half of April (West Jupiter. Florida, April 19. 27: 

 St. Helena Island. South Carolina, April 27. May :>), and pass west via 

 the Pine Barrens to the Mississippi, up which they ascend and reach the 

 vicinity of the Ohio valley about the first week in May (St. Louis, May S: 

 Cnicinnati. first week in May: Wabash. Indiana. May 4 and 7) and reach 

 the latitude of northern Ohio and southern Michigan about the second or 

 third week in May (Morgan Park. Illinois. May 21; Glen Ellyn, ^Liy 7: 

 Rockford. May 2.') : Lake Koshkonong. Wisconsin. May 24 ; Battle Creek. 

 Michigan. May 11: Kalamazoo, May 15;* Ann Arbor. May 14. 1.'). 10. is; 

 Cleveland. Ohio. May 4. 12, 13, 15 and June (?) ; Oberlin, May 11) and are 

 l)reeding in Oscoda and Crawford Counties, Michigan, early in June ^Mncic-.- 

 nac, May 21 : Luzerne. June 15.) The Minneapolis (May 13) and Toronto 

 CMay 16) captures. As the routes do not converge, it suggests that the 

 breeding area may be extensive. What is the destination of the birds along 

 the south Shore of Lake Erie? 



There remains to be considered another factor Avhich has apparently 

 influenced the location of the migration routes. This is a factor whose 

 influence could not have been estimated but for the careful and detailed 

 studiov nf the glacial geologists, particularly the work of Mr. Frank Leverett 



*Mr. F. TT. Cliapin writes me as follows concerning the capture of his Kalamazoo 

 specimen: "This Kirtland W'arhler I secured within the city limits, early in the morn- 

 ing of May 1.5. 188."). The specimen I have marked as a female. It was shot at very 

 short range and consequently was somewhat mutilated. It was in a small pine tree in 

 company with other warblers. I remember that the motions of the bird were rather 

 slow and not so active as most warblers. It was very tame." 



