302 JOURNAL, BOMBAY NATURAL HISTORY SOCIETY, 1890. 



From the foregoing table it will be seen that we consider that there are three 

 main groups of families, which may be considered as really related more or less inter 

 se. First, the Thrushes or Turdiformes. That the Turdidce and Sylviidce are 

 closely interwoven no one will deny, and we must place the Cinclidm near the 

 Turdidee, though their nest-building habits and their want of rictal bristles lead 

 towards the Wrens. Other families may be ultimately added to this Troglodytine 

 group, such as Mimidce, Myiadestidce, &c. ; but that the three leading forms of 

 Turdine families, viz., Thrushes, Warblers, and Flycatchers, are well characterised 

 seems to be as incontestable as the fact that they are intimately related. 



Another group of birds of which the relationships are obvious seems to us to 

 consist of the Crows, Starlings, and Paradise-birds. Starlings ought not to be 

 divorced far from the Crows, both being Ambulatores, and further connected by 

 Heteralocha, Podoces, Pyrrhocorax, and another genera. Mr. Oates'a family 

 Eulabetidse can lead from the Sturnidce to the Oriolidcs. 



The only other assemblage of birds which can apparently claim united affinities 

 are the Finches, which coalesce with the Tanagers on one hand, and with the 

 Weaver birds on the other, and these are not far removed from the Hangnests, 

 lcteridce. 



All the other families of birds seem to stand more or less isolated. The 

 Hirundinidce are decidedly a family apart, their nearest relatives being apparently 

 Muscicapidcr. through such forms as Artomyias. 



The Ampelidce have no very near relations, as far as we can judge, but Shufeldt 

 places them between the Laniidce and Hirundinidce. 



The Laniidce also stand by themselves as the centre of a cluster of families, which 

 are, however, none of them absolutely connected at the present day. The Shrikes 

 approach the Paridce through Falcunculus, and on the other hand the Vireonidce 

 cannot be far off, though, according to Shufeldt, they show great affinity for the 

 Mniotiltidce. 



The Atamidce have apparently no close relations. Mr. Oates puts them as a 

 sub-family of the Laniidce. 



The Paridce are intermediate between the Laniidce and the Certhiidce, and we 

 should not feel inclined to separate the Nuthaches as a distinct family from the 

 Paridce ; but this is a small matter, and they can be placed as a family between the 

 Creepers and the Titmice by those who consider them worthy of a distinct posi- 

 tion. Dr. Shufeldt inclines to the latter view. 



The Regulidce are a tiny family, but it seems convenient to keep them distinct, 

 near the Paridce, with inclinations towards the Sylviidce. 



The Mniotiltidce occupy the same position in the New World that the Sylviidce 

 do in the old, but beyond this they do not seem to be actually related, and they 

 are more nearly connected with the Certhiidce through Mniotilta. 



The Ccerebidce appear to us to constitute a separate family somewhat inter- 

 mediate between the Certhiidcr and the Tanrjuridce. 



