J. E. LITTLEBOT — THE JIIGUATIOX OF BIRDS. 35 



by Palraen, a continental ornithologist, in a work entitled ' Die 

 Zugstrassen tier Yo2,cl.' Miiirauts which cross the ^[editerrancan 

 in their northern fli<T;hts are reported always to do so in certain 

 defined routes. At Gibraltar the distance is short, and the passaj^e 

 easily accomplished ; farther to the east the islands of Sardinia and 

 Corsica present a ready higliway towards France and northern 

 Italy ; while the islands of Sicily and ]\Ialta become handy stepping- 

 stones between Tripoli on tlie north African coast and the mainland 

 of the Italian Peninsula. The overland route afterwards pursued 

 is not so readily determined, but it seems probable that the course of 

 rivers, coast-lines, the position of lakes, mountain-ranges, and the 

 gt'Ucral outlines of the district ti'aversed, become beacons by which 

 migration is directed. At Ust-ussa in Siberia the river Petchora 

 describes a great curve ; after running nearly due north it reverses 

 its course and flows for about 150 miles in a south-westerly direc- 

 tion before again turning northward to discharge itself into the 

 Arctic Ocean. An immense distance would be saved by crossing 

 the prairie from Ust-ussa to Alexievka; but notwithstanding this 

 inducement, the river-course is invariably followed by migrants. 

 Mr. Seebohm watched their flight most carefully and he believes 

 that the shorter course is never adopted. Migration from the south- 

 eastern portion of Europe appears to proceed principally from the 

 Elack Sea, along the course of the Volga and other rivers, next 

 across the short interval which intervenes between these rivers and 

 the sources of the Dwina, the Petchora, and the Ob, and thence to 

 the boundless breeding-grounds which border the White Sea and 

 the Arctic Ocean. 



Migration between our own shores and the northern homes of 

 our winter visitants is, in all probability, mainly directed by coast- 

 lines. The reports from lighthouses, to which I have alluded, 

 establish the fact that the great stream of autumnal migration 

 always strikes our shores from an easterly or south-easterly direc- 

 tion, and it is readily accounted for by this hypothesis. Migrants 

 follow the coasts of the Baltic southward, until a westerly trend 

 becomes necessary ; Denmark is then crossed ; the island of Heli- 

 goland lies directly in their onward course, and the coast-line of 

 Holland, with its small outlying islands, conducts them within an 

 easy distance of England. Only a detachment of the mighty stream 

 of migrants which skims along the shores of Holland and Belgium 

 finds its way to England, and it is believed that some of these, 

 after reaching our eastern counties, pursue the coast-line southward 

 and again cross to Belgium or France. The committee of the 

 British Association reports as follows : * — " The great east to west 

 stream of migration is mainly composed of some few well-known 

 species which regularly come to us in the autumn, the great body 

 undoubtedly remaining to winter. Placed in order of rotation, 

 according to their numerical superiority or otherwise, we find the 

 skylark, starling, hooded- crow, and rook, the song-thrush, blackbird, 



* ' Eeport Brit. Assoc, for 1881,' p. 190. 



