112 MICHIGAN SURVEY, 1908. 



barometer, and clear sky. (Cf. Smith, '07, p. 223.) It therefore seems 

 evident that low temperature and high barometric pressure, with the 

 prevailing northwest winds and clear sky which accompany them fur- 

 nishes the most favorable conditions for the bulk of the fall migration. 

 It will be noticed vmder the head of Migration Eoutes that a few birds 

 prefer northeast instead of a northwest wind. The conditions which 

 would be favorable for the migration of these birds would occur after 

 the passage of a high and while the approaching low was still some dis- 

 tance off. 



It is desirable that similar observations be carried on at other favor- 

 able localities in order to further test these conclusions, and determine 

 whether they are of general application to the fall migration. 



VI. ROUTES OP MIGRATION. 



From the observations made during the falls of 1904 and 1905, it 

 seems that Isle Royale lies directly in the path of a very strong mi- 

 gratory movement. In the fall there was a great massing of bird 

 life. For some unknown reason the path of densest movement was 

 very narrow, at least appearances pointed to such a condition. This 

 apparent narrowness of the route through the island was strikingly 

 shown on September 9 when a trip was made across it from Washing- 

 ton Harbor to Siskowit Bay. About 15 miles Avere traversed, em- 

 bracing every environment from clearings to high hardwood forests 

 and damp cedar swamps. Nearly a day was spent hunting over the 

 clearing and adjacent forest near the head of the bay, but scarcely 

 any migrants were observed. A few Black-throated and Tennessee 

 Warblers and a few sparrows were seen, while an occasional Sparrow 

 or Sharp-shinned Hawk was met. This was not due to a lack of food, 

 as grasshoppers and other insects were very plentiful. At Washington 

 Harbor the reverse was the case; here on September 9 and 10 I saw 

 many migrants, the majority of which were not seen at Siskowit at 

 all. These observations at the harbor were made in the morning before 

 leaving and in the late afternoon of the following day when I re- 

 turned from the bay. While at Siskowit scarcely a bird was heard 

 passing over, although at the harbor they could be heard throughout 

 the night. The path apparently extended lengthwise of Isle Eoyale 

 with Washington Harbor and the region lying between it and the 

 north shore of the island as its diameter. 



In a recent paper, Taverner ('05) makes the statement that perhaps 

 a migration route lies between Isle Royale and Keweenaw Point. From 

 the observations made on the island, I am led to believe that such a 

 route does exist and also one lying much to the west of this point, 

 perhaps to the Apostle Islands and the mainland lying Southwest of 

 them. These conclusions were drawn from a consideration of the fol- 

 lowing facts. The route taken by the majority of the migrants, both 

 those which passed slowly across the island and those observed flying 

 overhead, whether by night or day, lay nearly southwest. During the 

 latter part of August and parts of September, the nights were un- 

 usually bright, so that migrating flocks could often be seen high in the 

 air even when not crossing the face of the moon. The cries of mi- 

 grating birds, heard mostly oft cloudy nights, usually came from a 



