147 



On May the 21st the tld7'd immigration began with the 

 simultaneous arrival of birds in Devon and Essex followed bj 

 others in Kent on the 23rd, in Devon and Hampshire on the 

 24th and 26th and in Devon on the 27th and 28th. Like 

 the last, this immigration could be traced step by step, the 

 more westerly birds leaving North Wales and going north 

 through the Isle of Man. On the east the only point of 

 diflPerence between this and the former immigration seems to 

 have been that a large number of birds passed directly north, 

 reaching Suffolk, Norfolk, and Lincolnshire, and it may be 

 that these birds v^ere emigrants from our east coast making 

 for more northern breeding-quarters. 



The fourth and last immigration, like the first, was con- 

 fined to the western half of the southern seaboard and was 

 noticed first of all at the Eddystone light on the morning 

 of May the 31st. On June the 1st it was noted in Devon 

 and on the 2nd in Dorset and Hampshire. The first arrivals 

 passed straight over Devon and the Bristol Channel into 

 South Wales and reached North Wales on June the 1st, 

 whence they followed their predecessors northwards. The 

 later ones seem to have passed north through Wiltshire and 

 Surrey, but their further progress could not be traced. 



Chronological Summary of the Records. 



April 21. Devon. 



„ 27. Wiltshire, Cambridge and Shropshire. 



,, 28. Devon. 



„ 30. Devon (many), Dorset, Hants, Somerset and 

 SufEolk. 

 May 1. Kent^ Oxford, Somerset (many), Staffordshire, 

 Cardigan and Yorkshire. 



,, 2. Devon lights, Devon, Berks (many), Shrop- 

 shire, Merioneth, Cheshire and Durham. 



„ 3. Hants (few), Surrey, Berks (decrease), Glamorgan, 

 Radnor (few), Derby and Cheshire (several). 



