370 NATURAL HISTORY 



place of their roosting, I took care to wait on them 

 before they retired to rest, and was much pleased to 

 find that, for several evenings together, just at a quarter 

 past five in the afternoon, they all scudded away in 

 great haste towards the south-east, and darted down 

 among the low shrubs above the cottages at the end of 

 the hill. This spot in many respects seems to be well 

 calculated for their winter residence : for in many parts 

 it is as steep as the roof of any house, and therefore 

 secure from the annoyances of water ; and it is moreover 

 clothed with beechen shrubs, which, being stunted and 

 bitten by sheep, make the thickest covert imaginable; 

 and are so entangled as to be impervious to the smallest 

 spaniel : besides, it is the nature of underwood beech 

 never to cast its leaf all the winter ; so that, with the 

 leaves on the ground, and those on the twigs, no shelter 

 can be more complete. I watched them on to the 13th 

 and 14th of October, and found their evening retreat 

 was exact and uniform : but after this they made no 

 regular appearance. Now and then a straggler was 

 seen ; and, on the 22nd of October, I observed two, in 

 the morning, over the village, and with them my re- 

 marks for the season ended. 



From all these circumstances put together, it is more 

 than probable that this lingering flight, at so late a 

 season of the year, never departed from the island 1 . 



1 Mr. White appears to have; a strong bias to believe that martins, &c. 

 remain dormant in this country, having taken up a very erroneous notion 

 of the difficulty of the passage. Mr. Cartwright, during many winters' 

 residence on the coast of Labrador, had repeated opportunities of minut- 

 ing and measuring the flight of birds from point to point in their migration 

 to and from the north, and he asserts from the results of repeated obser- 

 vations that the eider duck at the time of migration flies at the rate of 

 ninety miles an hour, and there is no reason to suppose that the flight of 

 that bird is particularly rapid. A large bird is easily distinguished in 

 the air at the distance of a mile ; and if we consider that in less than a 

 minute when flying from one hill to another it passes out of sight ; we 

 shall easily satisfy ourselves that the calculation does not at all exceed 

 the bounds of probability. The passage of birds across the Mediterranean, 

 the British channel, and the Sound, is by no means long, and, if it be per- 

 formed at a rate at all approaching that at which the wild fowl are ascer- 



