12 ON THE MIGRATION OF BIRDS. 
existence or of torpidity*. A few, indeed, may occa- 
sionally be seen at the customary time in spring even 
when the weather is frosty, the increments of tempe- 
rature by no means corresponding with the sun’s 
increasing northern declination, but- they generally 
seem to withdraw again. On the 9th of April, 1S21, 
several Sand-Martins were observed at a sandpit in 
the township of Cheetham, but the weather becoming 
cold and stormy they quickly disappeared ; they were, 
however, soon after discovered, in greatly increased 
numbers, at a sheltered bend of the river Irwell, in 
the adjoming township of Broughton. This circum- 
stance proves that if the weather is severe and bois- 
terous when the Summer Birds are first seen in spring, 
they do not retire to their winter retreats, as has been 
supposed, but merely seek sheltered situations where 
they can procure a supply of food. 
Inquiries into the temperature of the supposed 
winter retreats of the Periodical Summer Birds may 
now be looked upon, it is presumed, as quite super- 
fluous, since it is sufficiently apparent from the 
preceding Tables that even that of the atmosphere 
is much higher at the time these birds disappear 
than it is when they appear, the very reverse of 
what ought to be the case if they become torpid, 
* Since writing the above I find that Montagu, in the Sup- 
plement to the ‘ Ornithological Dictionary,’ asserts that he has 
occasionally discovered the Lesser Pettychaps (Sy/via hippolais) 
in the south of Devonshire in mild winters. 
