60 
Introduction; Variation. 
of Celebes and the Philippines, and not Celebes and the Philippines from them; 
their volcanic or coral character (see de Hollander, Land- en Volkenkunde 
Ned. Cost. Ind. 1884, II, 234) and the absence of heavy, ill-flying birds and of 
peculiar generic types speak for the recent upheaval and colonisation of these 
islands^). It might happen that a species or two on the mainlands subsequently 
became smaller, but there is no reason to assum^that the twenty species which 
remained in Celebes and the Philippines all became small hereafter, while those 
which had peopled the islands all maintained the original sizes of the species. 
On the other hand there is reason to anticipate that the individuals of these 
twenty species which had emigrated to the new islands would undergo alteration 
of some kind or other, for the conditions of existence are not precisely the 
same there. It appears, therefore, quite safe to assume that Celebesian and 
Philippine birds develop as a rule into larger races in Sangi and 
Talaut^). 
As is usual in such cases it is not difficult to find more than one explan- 
ation why these things should be so, but not easy to decide which explanation 
is the true one. The most plausible suggestion is that the dangers are fewer 
and the struggle for partners perhaps more severe in Sangi and Talaut than on 
the mainland. 'Hawks and Falcons which prey upon birds seem to be very 
rare in these islands; so far not one bird-eater, strictly speaking, has been killed 
in them, for Tachyspizias soloensis, the most dangerous to small birds, is not 
only a migrant but feeds to a great extent on insects. Moreover there are no 
monkeys in Sangi and Talaut, and other enemies, which could be dangerous to 
breeding birds, their nestlings and eggs, are rarer than in Celebes and the Phi- 
lippines. The chief competition therefore that goes on would appear to be 
among the birds themselves, and the largest and strongest will be more likely 
to secure nesting quarters and partners than the smaller and weaker^). 
Decrease in size in Sula. — Two Birds -of- prey and two Pigeons display a 
slight reduction in dimensions in the Sula Islands, viz. Spilornis rufipectus^ Acci- 
piter sulaensis^ Turacoena manadensis and Macropygia alhicapilla. Sula seems to 
resemble Celebes in its Birds-of-prey. 
Differences in size in the North and South Peninsulas of Celebes. — As a rule 
the birds agree in their dimensions in these districts, but where there is a differ- 
ence it seems to be in the direction of an increase in size in South Celebes. 
Pachycephala meridionalis and Stoparola meridionalis are much larger than the 
1) Prof. Hickson noticed evidences of recent slight elevation in Talaut (Nat. in N. Celebes 1859, 151, 157). 
2) The case affords a good illustration of our postulate, that colonists become more changed than stayers- 
at-home (see, p. 162). 
3) It may be added that many ornithologists are of opinion that the males of most species of birds are 
more numerous than the females, and Dr. Platen certainly collected many more males than females in the 
Sangi Islands (see W. Blasius, Ornis 1888, pp. 527 — 646); but it may well be that the superior plumage of 
the males leads to their being shot and skinned more frequently, and it is preferable not to introduce this 
doubtful element into the argument. 
