MARGARORNIS.—DENDROCINCLA. 171 
Colombian State of Antioquia, in the State of Panama, and in Costa Rica. Specimens 
from nearly all these places are before us, and we find that they agree in every 
respect. 
Compared with M. rubiginosa this species is very distinct, for not only is the general 
plumage very much browner and the spots on the under surface more numerous and 
larger, but the wings are relatively short and much rounded. 
We have no record of this bird beyond Salmon’s note that the iris in life is dark and 
that its food consists of insects. 
Subfam. GLYPHORHY NCHINA*. 
a. Rostrum latum, turdiforme, haud cuneiforme ; cauda brevis, quam ale brevior. 
DENDROCINCLA. 
Dendrocincla, Gray, List Gen. Birds, p. 18 (1840); Ridgway, Pr. U.S. Nat. Mus. x. p. 488 (6 Jan. 
1888) ; Sel. Cat. Birds Brit. Mus. xv. p. 162. 
This genus is a somewhat isolated one, for with a great general resemblance to the 
Dendrocolaptine in the shape of the wings, tail, and feet, and the bill not very unlike 
that of Dendrocolaptes, it has the nostrils completely covered with a membrane, so that 
the nasal openings are curved slits lying along the lower edge of the nasal fossa; this 
membrane, too, is to a great extent covered with short feathers. This peculiar nostril 
is almost exactly like that of Glyphorhynchus, so that Dendrocincla finds perhaps its 
most natural position near that genus; but the form of the bill is very different, and the 
tail much less elongated, the shafts of the central rectrices being not nearly so much 
produced. 
The genus Dendrocincla has been examined recently both by Mr. Ridgway and 
Mr. Sclater. The former admits twelve species, and names three others as unknown 
to him; the latter gives ten species as known to him, and three unknown. ‘This differ- 
ence does not affect the numbers of the Central-American birds, both authors allowing 
four species as inhabitants of our region. 
The extreme northern limit of the range of Dendrocincia reaches the hot forests of 
* Jn framing our introductory remarks on the family Dendrocolaptide (antea, p. 146) we had intended to 
assign the genus Dendrocincla to the Dendrocolaptinz, and to place it, as Mr. Sclater has done, with Dendro- 
colaptes as a broad-billed section of that subfamily. A further examination of the species of Dendrocincla, 
however, shows us that this arrangement is not compatible with our definitions, for Dendrocincla has the 
nostrils completely overhung by a thick membrane, and not open as in the other genera of the subfamily. 
This character brings Dendrocincla in juxtaposition to Glyphorhynchus, notwithstanding the great difference in 
the form of the bill and structure of the tail, which are described elsewhere under the respective genera. 
Our scheme set out on p. 146 will still stand as it is, but the distribution of the genera in the various subfamilies 
must be altered as follows :—The Glyphorhynchine to contain two genera, Glyphorhynchus with one species and 
Dendrocincla with four species; and the Dendrocolaptine six genera, instead of seven, with fifteen species. 
22* 
