138 Bulletin American Museum of Natural History. [Vol. XXXIV, 
on the oil-gland prove erroneous or only partly true in the case of the Buc- 
conide, Picidse, Capitonide, and Momotide. _ 
Although the Puff-birds (like the Jacamars) are usually credited with a 
nude oil-gland Gadow states that in some of them itis feathered. On exam- 
ination of every genus of these two families I find that without exception the 
oil-gland is perfectly bare.* 
Both in the Woodpeckers and the Barbets the oil-gland normally bears 
a tuft of down at its tip, and I am not aware that any exception to the pres- 
ence of this tuft has been recorded. In each of these families, however, there 
prove to be several genera in which the oil-gland is nude. ) 
Of the Woodpeckers all the genera with the exception of Sapheopipo 
and Trichopicus have been examined in this connection. In the majority 
the tuft of the oil-gland is of small size; fairly large however in the Cele, 
Chrysophlegma mentalis and C. flavinucha, Megapicos pollens, Sphyrapicus 
and Hypopicus. ‘The genera in which this tuft of down is absent are the 
mutually allied forms, T2ga, Brachypternus, Gauropicoides and Gecinulus, 
but it is present in the related genera Micropternus and Meiglyptes.2 In 
Chrysocolaptes the tuft varies from vestigial to absent even within specific 
limits. In the Barbets (Capitonidee) we find a small, dense tuft in the four 
American genera and the seven genera of green Asiatic Barbets. In the very 
distinct brown Calorhamphus of the same region the tuft is thin and sparse. 
It is in the same condition in the African forms, Xylobucco and (probably) 
Trachylemus, but is wholly wanting in Heliobucco, and the natural group of 
tooth-billed Barbets comprising the genera Pogonorhynchus, Erythrobucco, 
Melanobucco, Lybvus and Tricholema. The following genera (all Ethiopian) 
1T am strongly convinced that a primary division of the antiopelmous Picarians should 
be made into two groups, one containing the Bucconide and Galbulide, the other the Indi- 
catoride, Ramphastide, Capitonide, Yungide, and Picide. This course was suggested by 
Stejneger (Standard Natural History) but Ridgway in his late treatment of the group, which 
he considers a suborder, Picarise, divided it ol once into four n&. Gaaenayennean Pici, Capitones, 
Ramphastides and Galbulee. 
In view of the many radical differences between the Galbule and the remaining super- 
familes, as compared with the few and comparatively unimportant differences between the 
Capitones and Ramphastides, or even between these and the Pici, it seems inevitable that the 
Galbule should be given higher rank. 
Diagnostic characters of the latter are as follows: Nude oil-gland, well-developed czeca, 
two carotids, non-oscinine wing-coverts, only moderately. deep temporal fosse, thoracic 
hemapophyses with lateral ventral enlargements, furcula with hypocleidum, ectepicondylar 
process of humerus absent. The metacarpus and distal end of metatarsus are both very 
different from the highly characteristic forms of the other families, and the plumage alto- 
gether lacks the bright red so universal in the Barbets, Toucans and Woodpeckers. If the 
antiopelmous forms are regarded as an order the two primary groups can be ranked as sub- 
orders; if, however, we begin with a suborder some less definite term, as section, must be used. 
2 On account of the bristly feathers covering the nostrils Gauropicoides and Gecinulus are 
widely separated by Hargitt (British Museum Catalogue) from the other genera above men- 
tioned. They agree with Tiga, however, in the absence of the hallux:and the large size of the 
sixth rectrix, as well as in the nude oil-gland, and in my opinion the six genera enumerated 
form a natural group. 
