660 
Birds of Celebes; Peristeridae. 
chary about setting its foot on the mainland of New Guinea. On small islands, 
on the other hand, it has been recorded from the Mergui Islands off Tenasserim, 
it abounds in some of the Nicobars, though, like the Gannets of European 
seas, it has a favourite breeding-island; it does not occur, as Dr. Vorderman 
says, on the mainland of Billiton, but on some small islands near the coast; 
it has not been met with yet on the mainland of Borneo, hut on the small 
islands of Labuan, lega and others, Mantanani, IVIangsi and Sibutu; it is found 
on small islands near IVIacassar (j' but has not been recorded from the neigh- 
bouring coast of Celebes; in the Solomon group Mr. Woodford has noticed, 
especially on the island of Malayta, that the Pigeons leave the mainland of the 
laige islands and resort at night in thousands to roost on the small detached 
islands off the coast”. Ihese curious habits, which we have termed insular 
(ci. Myristicivora hicolor, Carpophaga concinna, Carpophaga jrickeringi), Mr. W^ allace 
(f i) believes to be accounted for in Caloenas by the circumstance that ‘Taeing 
a ground feeder it is subject to the attacks of carnivorous quadrupeds, which 
are not found in the very small islands”. In the Solomons Mr. Woodford 
could only account for it by the supposition that the birds find themselves free 
from the attacks of the large Monitor Lizards, which are less plentiful there than 
on the mainland. Such enemies are most likely more dangerous to the eggs and 
nestlings than to the old birds, which look as if they ought to be able to look 
out for themselves as well as the average mainland dwellers; but, it must be 
confessed that their resorting to small islands to roost shows superior wisdom 
in taking care of themselves. Of course, if all Pigeons were to adopt iusniar 
habits (if only as regards roosting), it requires no very vivid imagination to see 
that the result would not be profitable to them. Looking for the origin of this 
habit in the species which have adopted it, it seems clear that the insular 
species of Myristicivora and Carpophaga have no structural peculiarities to render 
them specially adapted to this mode of life; it appears likely that they were 
species which were getting the worst of the struggle for existence on the main- 
land, and gradually adopted themselves to the small islands, of which they have 
become the peculiar inhabitants. 
Ihe wide distribution of this species is due to its traversing spaces of sea 
flight. Mr. Hume, as mentioned above, saw flocks of it flying from Batty 
Malve out to sea evidently on the way to other islands of the Nicobars. 
Mr. Whitehead writes (31) that it is “very plentiful on the small islands at 
some distance from the coast of Borneo. This Pigeon migrates from island to 
island, and was very common on Pulo Tega in April, where in most months it 
would be difficult to find a bird . Meyer says it is a common species in the 
Sangi Islands, yet no other naturalist has succeeded in finding it there; this is 
probably because it is present there only at certain times. Mr. Wallace (f 1) 
records a case of one of these birds flying as far as a small coral island, a hundred 
miles north of New Guinea, with no intervening land. “After the island had 
been settled a year, and traversed in every direction, the son (of Mr. van 
