i84 
NATURAL HISTORY 
tbe}'^purfue the fun into lower latitudes, as fome fuppofe, in order 
to enjoy a perpetual fummer, why do they not return bleached ? 
Do they not rather perhaps retire to reft for a fcafon, and at that 
■juncture moult and change their feathers, fince ail other birds 
are known to moult foon after the feafon of breeding ? 
Swifts are very anomalous in many particulars, diffenting from 
all their congeners not only in the number of their young, but in 
breeding but once in a fummer; whereas all the other Britijh 
hirundines breed invariably tiv'ice. It is paft all doubt that fwifts can 
breed but once, fince they withdraw in a fhort time after the flight 
of their young, and fome time before their congeners bring out their 
fccond broods. We may here remark, that, as fwifts breed but once 
in a fummer, and only two at a time, and the other hirundines 
izvice, the latter, who lay from four to fix eggs, increafe at an 
average five times as faft as the former. 
But in nothing are fwifts more fingular than in their early retreat. 
They retire, as to the main body of them, by the tenth of Augujl, 
and fometimes a few days fooner : and every ftraggler invariably 
withdraws by the twentieth, while their congeners^ all of them, flay 
till the beginning of Otlober ; many of them all through that 
month, and fome occafionally to the beginning of November. This 
early retreat is myfterious and wonderful, fince that time is often 
the fvveetefl: feafon in the year. But, what is more extraordinary, they 
begin to retire ftill earlier in the moll foutherly parts of Andaliifia, 
where tliey can be no ways influenced by any defed of heat ; or, as 
one might fuppofe, defed of food. Are they regulated in their 
motions with us by a failure of food, or by a propenfity to moult- 
ing, or by a difpofition to reft after fo rapid a life, or by what? 
This is one of thofe incidents in natural hiftory that not only 
baffles our fearches, but almoft eludes our gueffes ! 
Thefc 
