72 
Birds of Celebes; Falconidae. 
to be an individual variety of the variable <S. limnaetus\ Prof. W. Blasius finds 
that bird in Mindanao (J. f. O. 1890, 144). It is possible that all these forms 
are reversions to the two original forms of Celebes. 
To these reasons may be added, (D) that both Pernes and Spizaeti lay only 
one or two eggs at a sitting and, thus, they are not likely to get crowded out 
of a large island, but might exist there with the conditions of reproduction and 
destruction at a balance for any required length of time. 
In how' far these remarks are well, or ill, grounded can, of course, only be 
ascertained by a much more thorough knowledge and study of the two genera, 
Perms and Spizaetus, than has been jrossible to us; but the case serves to show 
— and we believe it is only one in very many — that, where some striking 
similarity between two members of distinct genera occurs, this may be looked 
upon as an ancestral character retained, rather than as something acquired by 
the natural selection of varieties of one species, which obtained some advantage 
by resembling the other. Whoever has examined young and old specimens of 
these birds will find it hard to listen to the supposition, that the parallelism 
between the two forms is a fortuitous coincidence, that they are of independent 
development, but in our present state of ignorance any one who will is per- 
fectly free to make this suggestion and hold to it. 
The label on a specimen in the Dresden Museum, stating that the stomach 
contained wasps and their larvae, shows that P. celebensis, like the other species 
of the genus, feeds mainly upon hymenoptera, its peculiarly feathered face 
protecting it against their stings. The allied P. ptilonorhynchus is found as a 
rule to lay two eggs, but one egg at a sitting is not uncommon (Oates, Hume’s 
Nests and Eggs, 1890, III, 181). 
23. PERNIS SP. 
In the Novitates Zoologicae 1896, 177, Mr. Hartert describes a Honey-buzzard 
obtained by Mr. Everett in Saleyer in November, 1895, but wbicli he was unable to identify 
with certainty. “A large bird (female), remiges moulting, whig 440 mm. The whole under- 
side is buff or ocbraceous buff, some of the feathers (older ones) paler, others (the new ones) 
darker and brighter. The throat is surrounded by an irregular black band, the feathers of 
the lower throat and upper breast have narrow deep brown shaft-lines, but all the breast, 
abdomen, flanks, scapulars [axillaries?] and under wing-coverts are uniform Mthout a trace 
of bars or bands. Upper side dark brown as in most Pei'nes, not differing from many 
specimens of P. ptilomrhynclms.” 
Mr. Hartert believes the specimen to belong to the latter species, and not to P. cde- 
hensis. It is manifestly immature, but from the descrijition evidently differs from the young 
P. celebensis by its larger size, and the absence of bars on the flanks, thighs and under tail- 
coverts. P. ptilonorhynchus (Temm.) is a highly variable species, as is pointed out in the 
preceding article, but it may always be distinguished from P. celebensis by its large size, and 
the adult — usually at all events — has a long nuchal crest. 
