364 THE STRUCTURE AND LIFE OF BIRDS cuHapP. 
not appearing till the latter half of September. A 
few unmated old birds precede the young ones by two 
or three weeks. It is strange that the young and inex- 
perienced should start, apparently, so much sooner 
than is necessary. Possibly, however, they have it 
strongly hinted to them by their elders that their 
room is wanted, since the supply of food is limited 
and the young second broods are voracious. There isa 
very remarkable exception to the general order of 
departure. The old Cuckoos are in the same position 
as the unmated birds of other species. They have no 
responsibilities to keep them in the north, and they go, 
leaving their young to find their way south as best 
they can. 
These curious phenomena show how intimately 
migration is connected with nesting. Many birds who 
have paired stay late in order to bring up their young. 
Yet every year it happens that some Swallows and 
Housemartins start for the south, leaving their young 
to dieinthe nest. This seems unnatural ; still, parental 
affection keeps them in England longer than they 
would remain but for that. We may imagine a 
struggle within them between the love for their young 
and the migratory impulse, and the latter at last 
becomes overpowering and swamps the former. The 
unmated cock birds, in whom no such struggle 
between opposing motives goes on, have started long 
before. Many who have no young brood to tend 
are delayed by their moulting. It is possible that 
the unpaired migrants, like unpaired pigeons, have 
their moult delayed, and that their early departure 
southward takes place before it begins. This question 
