148 von Ihebing, The Dendrocolaptidoe. [April 



Corresponding with the two groups above indicated we find 

 modifications in the structure of the skull. In both series the 

 strongly modified forms have the frontal bone exceedingly large 

 and the nasal foramen relatively small — the extreme reduction 

 being reached in Campylorhynchus and allied genera. By this 

 means the basal bridge between the nostril and frontal bone be- 

 comes extraordinarily large and strong, an adaptation corre- 

 sponding to the increased demand in these birds for strength and 

 resistance at the base of the beak. While the precursors of Pico- 

 laptes seem to be extinct the line of evolution originating from Sit- 

 tasomus is nearly uninterrupted. The skull of Sittasomus differs 

 but little from that of Automolus and to this the skull of Sclerurus, 

 seems closely related. 



With regard to skull structure the Synallaxinse may be con- 

 sidered as a more or less uniform group in which the genera Thripo- 

 phaga and Phacellodomits are somewhat differentiated by the 

 strongly convex base of the beak, prolonged posteriorly in two 

 divergent ridges, surrounding a deep pit. 



A peculiarity of the species of SynaUaxis and Septomis is the 

 large, deep median furrow of the frontal bone with a corresponding 

 projecting ridge on the inner side of the skull. There is also a deep 

 pit at the posterior end of the intermaxillary near the anterior end 

 of the frontal. We meet with the same conformation in Lochmias 

 nematura where the lateral parts of the frontal bone are extraordi- 

 narily convex and separated by a deep median furrow. Cinclodes 

 presents the same condition while Upucerthia differs somewhat in 

 the more projecting nasals which surround the posterior part of the 

 intermaxillary. The skull of Upucerthia resembles that of Thripo- 

 phaga and Phacellodomus while Lochmias agrees with SynaUaxis. 



Of the subfamilies of the Dendrocolaptidse proposed by Sclater 

 the least natural one seems to be the Furnariinse. 



There are in general no great differences between the skulls of 

 Furnarius and SynaUaxis. In the former, however, the frontal 

 fontanelle, so well marked in SynaUaxis is absent, while the frontal 

 bone in SynaUaxis and allied genera is much narrower than in 

 Furnarius. Anumbius agrees in cranial characters with Synal- 

 laxis; and Pseudoseisura with Phacellodomus. If, therefore, we 

 place Lochmias in the Synallaxinre on the basis of skull structure 



