416 Philippine Journal of Science 1920 
from Caluya, from Sibay, and from Libagao;* all of them are common 
lowland species of wide distribution in the Philippines.] 
(4) Luzon, Catanduanes, Marinduque, and Lubang. [Polillo belongs 
with this group, although it has three endemic species; its plants show 
some peculiar features.] 
(5) Samar, Leyte, [Biliran,] and Bohol. * * * [The last collection 
made in Bohol included many species not previously known from that 
island; these fully confirm the belief that Samar, Leyte, and Bohol form 
a distinct group.” A small collection made by me in Biliran®™ shows that 
this island is merely a fragment of Leyte; these two islands are so near 
each other that no other result could have been expected.] 
(6) Mindanao and the islands immediately adjacent to it form a group 
by themselves. 
(7) ash Sa be separated from Mindanao on account of the fact 
that it has a number of representative forms of species characteristic of 
Mindanao. [Orthotomus mearnsi and Loriculus dohertyi, of Basilan, de- 
scribed since Worcester wrote the above, furnish further evidence of the 
difference between Basilan and Mindanao. ] 
(8) Bongao, Tawi Tawi, Lapoc, and Sulu form a well-marked natural 
group, to which Sibutu must probably be added. 
(9) Tablas, Romblon, and Sibuyan show no evidence of having been 
connected with any of the neighboring larger islands. They have a number 
of peculiar species of birds, and Tablas and Romblon should probably be 
classed together. [Banton shows a slight relationship to Romblon. Cresta 
de Gallo, south of Sibuyan, is little more than a sand bar.] 
(10) Cebu cannot be regarded as one of the central Philippine group 
but must be classed by itself. 
(11) The Batanes Islands [=the Babuyanes] have a strong Formosan — 
element among their birds. It remains to be seen whether the Babuyanes 
Islands [= the Batanes] must be grouped with them or must be considered 
as detached fragments of northern Luzon. [This paragraph is badly 
mixed. At the time that it was written we had a small collection from 
Fuga and a large collection from Calayan, both in the Babuyanes, but 
nothing was known of the birds of the Batanes. Since then we have secured 
collections from Camiguin, in the Babuyanes, and from Batan, in the Bata- 
nes; we also have a few, mostly unimportant, specimens from Dalupiri, 
Babuyan Claro, Y’Ami, and Sabtan. Although Camiguin, Calayan, and 
Batan, and perhaps some of the other islands of these two groups, show 
differences from each other, they all differ so much from Luzon that they 
form a group quite as distinct as some of the others considered above. ] 
(12) Siquijor, Cagayancillo, and Cresta de Gallo are islands of recent 
origin and their bird faune have been derived from neighboring islands. 
[Siquijor has three endemic species and lacks some of the central-island 
species, but it would be grouped with the central islands if with any. The 
fauna of Cagayancillo is more puzzling. Cinnyris aurora, a characteristic 
Palawan species, is as abundant there as it is on Cuyo; but Dicrurops%s, 
which is common on Cuyo, is not found on Cagayancillo. Zosterops rich- 
mondi, endemic to the latter island, represents a genus that is unknown 
* McGregor, R. C., Philip. Journ. Sci. 1 (1906) 698. 
™ See McGregor, R. C., Philip. Journ. Sci. § A 2 (1907) 315-333. 
* Bureau of Science collection; results unpublished. 
