‘ON THE MIGRATION OF BIRDS. 19 
Hirondelles et les Martinets muent une. fois l’année 
en février, par conséquent dans le temps de leur 
séjour dans les climats chauds de l'Afrique et de 
l’Asie; un fait d’ailleurs qui prouve incontestablement 
contre la prétendue torpeur ou sommeil hivernal de 
ces oiseaux. Les observations de M. Natterer ont été 
faites sur des Hirondelles élevées en cage, dont un 
petit nombre a vécu huit et neuf ans en domesticité.”’ 
Old Cuckoos leave us late in June or early in July, 
when the temperature is approaching the maximum 
for the year; and Swifts retire about the middle of 
August, when the temperature, though receding from 
the maximum, is still very high. To what cause, 
then, shall we attribute the early retreat of these 
birds? Certainly not to a deficiency of food, as young 
Cuckoos are frequently found to remain upwards of 
two months after the old birds have left ; and Swifts 
are occasionally seen long after the great body of their 
congeners has withdrawn* ; and yet these individuals 
* In the year 1815 I saw « Swift in the township of Crump- 
sall, on the 20th of October, and the same bird was seen again 
on the 25th, which is more than two months beyond the time at 
which this species usually departs, and nearly a fortnight after 
the last Swallows and House-Martins had left us; and in the 
year 1818 I saw one at Chester on the 18th, 19th, and 20th of 
October. I had opportunities of observing both these birds at- 
tentively for a length of time, and I remarked that they always 
seemed to be in the active pursuit of their prey. White, in his 
‘Nat. Hist. Sel.’ p. 264, mentions an instance of a Swift being 
induced, by attachment to its young, to remain till the 27th of 
c2 
