Birds. 6959 



extended, occasionally almost equally so with the limits of the rafige 

 of its existence ; also that the climate of every part of the districts 

 passed through in migration are not either uniform nor uniformly con- 

 stant in each year, and hence that it may occur that a bird in its 

 northerly progress, from some cause or other late in its migration, — as 

 for instance, a weakling left behind at the annual starling point, 

 whence all the stronger birds pushed boldly forwards for the North, — 

 may, on finding the instinct of nidification too strong upon it, build 

 its nest in the first spot which copies its proper summer region suffi- 

 ciently near to fall within the extreme limits of its standard of exist- 

 ence, or even in a district which, whilst falling within the range of its 

 existence, is yet excluded from this standard. 



That summer and winter migrations take place almost always, if not 

 always in a line due north and south, is an argument in favour of this 

 explanation. I cannot call to mind a case in which nidificatory 

 migration is markedly east and west over an extended area. In some 

 few cases, it is true, the line of migration is not bounded by strictly 

 parallel longitudinal lines east and west, but this might have been 

 premised, since neither can areas similar in their climates be bounded 

 by parallel lines, but in tracing such, northwards or southwards, we 

 meet with divergent outlying sub-areas which copy the climates north- 

 ward or southward of them rather than those adjacent to them in 

 a direct line. Such exceptions are, however, rarer than might have 

 been expected. 



A few instances selected from among the birds of the British Isles 

 (as these have been best studied) and of these choosing species, which 

 occurring north in Great Britain, might have been expected to occur 

 in Ireland also, but do not, will suffice to illustrate this. 



Nightingale {Sylvia luscinia) occurs in Cumberland (Carlisle) 

 Lancashire, Yorkshire, Lincolnshire, Norfolk, Suffolk, Essex, Middle- 

 sex, Kent, Somerset, Hants, Sussex, and East Devon. It is common 

 only in the easternmost of these shores, being rare in Cumberland, 

 Lancashire, Somerset and East Devon, unknown in West Devon and 

 Cornwall ; apparently so in Scotland, certainly in Wales. North of 

 Britain it occurs in Sweden, Russia and Siberia ; and south, in the 

 eastern parts of France, Germany and Spain, wintering in North 

 Africa, Egypt and Syria; unknown in the Channel Islands and 

 Brittany. Now, if these points be connected on the map, it will be 

 be seen that the regular line of migration is to the eastwards of a line 

 which cuts off Ireland, Wales, Scotland and part of England. 



Red-backed Shrike {La?iius collaris) occurs in Cumberland, 



