XXII THE PROBLEM OF INSTINCT 503 



as a whole, in some way that we have hitherto been unable 

 to observe." The chief obstacle to this view consists in 

 the well-known observations of Herr Gatke at Heligoland, 

 that, during the autumn migration, in the case of the 

 great majority of species, the young birds migrate earliest 

 and alone, the adults following considerably later. But 

 admitting, as every one must, the accuracy of Herr Gatke's 

 observations, does the conclusion necessarily follow ? He 

 himself assures us that the birds which rest on or pass 

 within sight of Heligoland only form a fraction of the 

 whole of the migrating hordes, most of them travelling by 

 night at great altitudes, and very few passing within sight 

 of the island, and of these few only, perhaps, one in ten 

 thousand stopping to rest. The fact that young birds of 

 many species are the first to visit Heligoland every year 

 without exception, may possibly be explained by the fact 

 that, while the older birds which lead the way travel high 

 and go on without stopping, a large number of the young 

 fly lower, and being either fatigued by the long un- 

 accustomed flight or attracted by the sight of the land, 

 descend to this elevated and fertile islet for rest and food. 

 The late Mr. Seebohm, whose extensive journeys to Siberia 

 and in various parts of Europe for the purpose of collect- 

 ing and studying birds, rendered him an authority on this 

 subject, gives the early migration of young birds on the 

 authority of the Heligoland observers, but does not 

 support it by any observation of his own in the northern 

 regions from which so many of the migrants come. In 

 America, although some writers state that young birds 

 migrate first in autumn, Mr. C. Hart Merriam, of the 

 Department of Agriculture, tells us that this notion is 

 " contrary to the experience of most leading American 

 Ornithologists and to the information collected by the 

 Committee on Migration of the American Ornithologists' 

 Union."^ But if we reject the conclusion based upon the 

 Heligoland facts as not necessarily following from them, 

 we shall find that there is not much difliculty in forming 

 a theory which accounts for the main phenomena, and the 



1 Report on Bird Migration in (he Mi-idssipl)i Valley, by W. W. 

 Cooke, p. 13. Footnote. 



