SAXICOLA. STONECHAT. 287 



long, almost straight, and more pointed ; in the second the 

 wings are shorter and slightly concave ; in the third still 

 shprter and decidedly concave ; in the fourth like those of the 

 first. In the first the tail is of moderate length, in the second 

 and third rather short, and in the fourth rather long. If we 

 assume the length and form of the wing and tail as characteris- 

 tic of genera, it is obviously impossible to include all the above- 

 mentioned birds in one genus. If the form of the bill and the 

 digestive organs alone determine the genus, all these birds 

 ought to be grouped along with the Sylviae, Wagtails, and 

 other genera. But if it be expedient to separate species which 

 differ considerably in form and habits, we may form out of the 

 four species in question three genera. The fourth species, 

 which by most authors is referred to a different genus, is much 

 more allied to the first, than either the second or the third are ; 

 and it will be seen on comparing Motacilla CEnanthe of Lin- 

 naeus with the generic character of Saxicola as given by M. 

 Temminck, that it does not agree with it in respect to the 

 wings. That species, with Turdus leucurus of Gmelin, Mo- 

 tacilla stapazina of the same author, and some others, I con- 

 sider as forming a genus apart from such species as Motacilla 

 Rubetra and M. Rubicola, which have the tarsi shorter, and 

 the wings rounded and concave. 



The Saxicolse thus restricted are generally, as their name 

 implies, inhabitants of rocky or stony places, although some 

 frequent also open pastures. Our only British species, for 

 example, is nearly as frequently found in the one kind of 

 place as in the other ; but unless they find holes or cavities to 

 which they can retreat, they are not fond of exposed feeding- 

 grounds. They nestle among stones, in crevices of rocks, 

 or in holes in the ground, whether made by quadrupeds or 

 not ; move with considerable speed by leaping, and feed on 

 insects and worms. Through the genus Petrocincla, they are 

 allied to the Thrushes ; their nearest kinsfolk on the other 

 hand are the Bushchats and Redstarts ; and with them they 

 are nearly related to the Flycatchers, both in form and habits. 

 The bill of the Saxicolae is almost precisely similar to that 



