156 MR. W. p. PYCRAFX ON THE OSTEOLOGY [Feb. 20, 



8ub-order TRACHEOPHONEiE. 

 Fam. 1. Furnariiclse. 



Sub-fam. 1. Furnariinse. 

 ,, 2. Scleruriiife. 

 „ 3. Synallaxina?. 



,, 4. Phyliclorhina?. 



Fam. 2. Pteroptochidse. 

 „ 3. Dendrocolaptidaj. 

 ,, 4. Conopophagidee. 

 „ 5. Forniicariidse. 



My amendments to this scheme amount to this — I propose ^ to 

 alter the balance and composition of his families 2-5, arranging 

 them as follows : — 



Sub-order Tracheophone^. 

 Fam. 1. Formicariidfe. 

 ,, 2. Dendrocolaptida^ 

 ,, 3. Furnariidfe. 



Sub-fam. 1. Furnariinse. 

 ,, 2. Sclerurinai. 

 ,, 3. Synallaxina?. 



„ 4. Margarornithinse. 



,, 5. Phylidorhinte. 



Fam. 4. Conopophagidje. 



Sub-fam. 5. Conopophagina?. 

 „ 6. Pteroptochinte. 



„ 7. Hylactinre. 



Fam. 5. Xenicidse. 



In this Sub-order all but the Xenicidse have a tracheal syrinx ; 

 and this is remail^able foi' the presence of a latei-al cartilaginous 

 pillar set on to the bronchial ring by a broad base. In some 

 genera this pillar is extremely well developed ; in others it is but 

 small, and may be wanting as in Conojyojihaga, though this genus 

 has hitherto been described as possessing this ^^ processus vocalis." 

 As to the development of this process in the Dendrocolaptidse — 

 corresponding to the Dendrocolaptinse of Sharpe's ' Hand-list,' 

 which includes about eleven genera — nothingseems to beknown. In 

 the references to the syrinx of the " Dendrocolaptidte " which have 

 fi'om time to time been made, this covering title has included both 

 Furnariine and Synallaxine types. These references indeed, in 

 nearly all cases, appear to be based on Mliller's dissections (6), who 

 does not seem to have examined any strictly Dendrocolaptine types 

 in this connection. 



All the Tracheophoneee are holorhinal except the Furnai'iiuie, 

 which are schizorhinal. 



ConojJophaga, as Forbes insisted long ago (2), has nothing to do 

 with the Furnariine, but seems to approximate towards the 

 Formicariine types. While Sharpe regards it as entitled to rank 

 as a Family by itself, it seems to me that we shall be nearer the 



