616 Mr. W. P. Pycraft on the 



presently, this skull also presents much in common with 

 forms not included in the subfamily. Of these, some must 

 undoubtedly be placed within the group ; others are nearly 

 related thereto, and this not on the evidence of the skull 

 alone, other parts of the skeleton tell the same tale. 



The skull of Acanthidositta agrees with a considerable 

 number of the forms at least which make up the subfamily 

 Synallaxince of the 'Catalogue of Birds of the British 

 Museum/ in that the anterior nares are of the Schizo- 

 rhinal type, while the maxillo-palatincs are long and slender, 

 embracing the under surface of the vomer. Further exami- 

 nation may shew that some of the species now included in 

 this group do not share this peculiarity. On the other 

 hand, I have discovered that several species included in 

 allied subfamilies arc Synallaxine in this particular at lca-t, 

 and thereby differ from the forms with which they are 

 associated. 



Among the skulls which present, in the matter of the form 

 of the external nares and of the maxillo-palatine, a very close 

 agreement, there will be found on the other hand slight, but 

 definite, differences in the form of the tympanic cavity. 



In Acanthidositta we have the simplest, and therefore 

 probably the most primitive, condition of this region; where 

 the tympanic wing of the exoccipital is but slightly developed, 

 and does not extend upwards beyond the level of the distal 

 border of the tympanic rim — for the attachment of the 

 tympanum. The postorbital process and the processus 

 articularis squamosi are represented by mere tubercles. In 

 Synallaxis and Siptornis, where the tympanic cavity has 

 shifted downwards and inwards still more than in Acanthi- 

 dositta, so as to open on the ventral instead of the lateral 

 aspect of the skull, the tympanic wing rises above the level 

 of the periphery of the tympanum ; while the processus 

 zi/t/omaticus squamosi has become a broad plate extending 

 upwards to join the postorbital process, but bearing a 

 fenestra between them. This is most marked in Siptornis. 

 In Cinclodes, Pseudocolaptes, and Homorus the postorbital 

 process is more strongly developed, and distinct from the 



