178 
Birds of Celebes; Picidae. 
Biiceros of good flying -powers in marked contrast to its fellows — are absent 
from tke Australian Region ; all three, also, are absent from the Sangi Islands, 
a circumstance which speaks for their recent volcanic upheaval. As in the case 
of the Bucerotidae^ Celebes has two Woodpeckers, one rather large, the other 
a small one, which, moreover, find their nearest allies in the Philippines with 
Sooloo. The Bucerotidae greatly enforce Mr. Wallace’s view of the former 
land- connection of Borneo, Sumatra and Malacca and of the earlier separation 
of Java from them, and a certain amount of similar proof is brought by the 
Woodpeckers. As Hargitt shows in his Catalogue the same species of the 
following Himalayan or tropical genera are peculiar to Malacca, Sumatra and 
Borneo, whereas Java has its own peculiar species of them, viz. 
Malacca, Sumatra and Borneo 
Chrysophlegma (2 sp.)’) 
Oauropicoides 
Lepocestes 
Miglyptes (2 sp.) 
Hemicercus 
Java 
Clw'ysophlegma (2 sp.) 
absent 
absent 
Miglyptes (1 sp.) 
Hemicercus. 
On the other hand several other Himalayan genera extending into the Sunda 
Islands are represented by identical species in Java, Borneo, Sumatra and Ma- 
lacca, viz. Tiga^ Chrgsocolaptes.^ liemilophus (ranging to the Himalayas in one 
species), Thriponasc (also in the Philippines) and Igngigjicus (most species of which 
are local, but one, I. auritus Eyt., has the above range extending further to 
Cochin China). Micropternus^ an Indo-Chinese form connects the Malay Penin- 
sula, Sumatra and Java hy a subspecies of the Indo-Siamese M. phaeoceps., while 
Borneo has its own peculiar species. The wide-spread Dendrocopus of EurofDe, 
Asia, N. America and southern S. America also has a species in Malacca, Su- 
matra and Java, but is wanting in Borneo. The two last-named genera are, 
however, more likely than the others to have extended their range by migration, 
because they belong to a more northern area — especially Dendrocopus. Of 
particular interest is the reoccurrence of the South and Central American genus 
Picumnus in the Oriental Region from the Himalayas and China south to Su- 
matra. 
To account for the distribution of the M^oodpeckers Wallace suggests a 
Central Asiatic origin, whence they spread to South America by way of North 
America, and developed into great diversity of form under the favourable con- 
ditions of the Neotropical forests (Geogr. Distr. II, 303). Tristram has urged 
a circumpolar origin. Marshall speaks for their evolution in the New World 
from the Passerine branch, and their gradual spread into the eastern Palaearctic 
Region, whence, as in North America, they were driven sonthwards by the 
glacial period, to work their way in many cases north again as the cold mode- 
rated. This view contains, perhaps, the fullest explanation. 
1) Sumatra and Malacca liave each a peculiar species. 
