214 ANNALS OF SCOTTISH NATURAL HISTORY 



or from Crieff and Comrie up the Earn, and so on into the 

 watershed of Dochart and Tay. 



We do not now go into Moray or Dee on this opportunity ; 

 but it is well to relate here that all our replies to inquiries 

 from any localities north of the Grampians are strongly 

 negative locally, and collectively so as regards summer 

 visitants. 



Now we have traced the advance along the lines of 

 least resistance (or most favourable routes) for the birds, it 

 may be well to speak shortly of the autumn and winter 

 influxes, or irruptions, and annual visitations from over-sea, 

 which have taken place at very regular intervals, beginning 

 about 1 86 1, and recurring, with more or less migration 

 annually, in smaller quantities, about every seven to eight 

 to twelve or fifteen years — e.g. 1861, 1868-9, 1886-7, 

 1898-9, and so on. We have broadly indicated these by 

 dates in red ink on the eastern divisions of Scotland from 

 Shetland southwards. Now, by far the largest bulk of the 

 regular migrations are recorded from localities south of the 

 Grampians ; and only when the septennial (or at longer 

 intervals ?) irruptions — caused by congestion at continental 

 centres — take place with prevailing south-east winds, do the 

 birds on flight lose control of their powers. And then the 

 farther isles in the north, even as far as Faroe, receive 

 unwonted visits, and some numbers of the wanderers. 

 How many perish far out at sea — i.e. what proportion utterly 

 fail to " catch up " to land-shelter first of our mainland 

 west coast, and then of our outer isles — must in great 

 measure remain a mystery, until simultaneous observations 

 be carried out by capable observers at many stations. 



The stations I would select (were I able to be in all 

 at once) I have often indicated before, — Isle of May, Pent- 

 land Skerries, Fair Isle, Flannen Isles, Monach Isles, and 

 perhaps north-west Mayo in Ireland. Then, Barra, Tiree, 

 St. Kilda, Skerryvore, Ross of Mull, north coast of Ireland 

 perhaps at Rathlin Island } and Isle of Man. 



These recommendations are selected from experience 

 gained by the returned schedules of the British Association 

 Committee's correspondents, and much other experience in the 

 field, during the larger part of a lifetime devoted to out-door 



