146 von Ihering, The Dendrocolaptidce. [a^hi 



observations furnish valuable hints on the systematic arrangement 

 of the genera. According to their manner of life these birds form 

 three natural groups. Those allied to Furnarius are inhabitants 

 of the open country and low lands. They construct their nests 

 in the ground with subterranean burrows leading to them, some- 

 times of considerable length. The custom of the Ovenbirds of 

 constructing their nests in trees is evidently a secondary adapta- 

 tion and the material employed in their construction — mud — 

 indicates that their ancestors nested in the ground. 



The second group contains the genus Synallaxis and related 

 forms. They live like many other small birds upon trees and 

 bushes and construct big dome-shaped nests, either of grass, moss 

 and other soft materials or of sticks. 



The birds of the last section comprising the Dendrocolaptinse 

 and part of the Philydorinse, live in the forest like the woodpeckers 

 and nest in holes in trees. 



The eggs of all the members of the family are white or whitish 

 green except in a few genera of Synallaxinse in which they assume 

 a uniform blue-green coloration. 



If we compare the above facts with the classification given by 

 Sclater in the 'Catalogue of Birds of the British Museum' we find 

 a general correspondence and are inclined to adopt his subfamilies 

 with some modification. The removal of the genus Anumbius 

 from the Synallaxinse cannot be approved. The Philydorinse with 

 the exception of a few genera approach the Dendrocolaptinse but 

 are easily distinguished by morphological characters. 



Radically opposed to our views, however, is the classification 

 adopted by Ridgway in his admirable work 'The Birds of North 

 and Middle America,' Vol. V, where the birds under consideration 

 are distributed in two distinct families, — Furnariidse and Dendro- 

 colaptidse. The reason for distinguishing two families is stated to 

 rest chiefly on differences in the structure of the skull. I have 

 studied the skulls of a great number of genera and shall explain 

 the results of my researches. 



In accordance with Garrod, Beddard and other authors, Ridgway 

 places the genera with a holorhinal skull in the family Dendro- 

 colaptidse, and those with a schizorhinal skull in the Furnariidse. 

 In the latter group the osseous nostril reaches the posterior end of 



