206 NORTHERN ZOOLOGY. 



We have already, while discussing- the relations of the Thamnophilince with the 

 Myotherince, adverted to that singular property, belonging to typical groups in 

 contiguous circles, of uniting by collateral affinity, without destroying those 

 circular affinities by which their own respective circles are described. We have 

 had proofs of this change of analogies into affinities in the families of La?iiadce and 

 MerulidcE ; and we see it, if possible, more strongly exemplified in the group now 

 under consideration. The genus Sylvicola describes its own circle by the types 

 here enumerated, without the intervention of Setophaga, which is the next group, 

 and of equal rank. Yet the typical Sylvicolce are so beautifully connected to the 

 typical Setophagce, without the intervention of the aberrant forms in either group, 

 that it is impossible to imagine a single deficiency in the links which connect them. 

 We call the attention of ornithologists to this most remarkable fact, which may be 

 gathered from the invaluable pages of Wilson, detailing their manners, and by a 

 minute examination of the birds which we shall subsequently notice. The genus 



SETOPHAGA, Sw., 

 as being almost exclusively a North American group, may be here slightly men- 

 tioned. As this is the Fissirostral type of the Variance, we can feel no surprise at 

 its having been blended with the genuine Fly-catchers, from which it can only be 

 distinguished by the different structure of its wings and feet : it is represented, 

 among the genuine Mnscicapince, by the sub-genus Rhipidura, H. and V. The 

 species yet discovered are not many, and these differ in so remarkable a manner, 

 that the principle of their variation can never be explained or rightly understood, 

 without reference to the numerous affinities which this group is evidently intended 

 to reconcile. On one side it is to connect Sylvicola with Trichas, on another it 

 is to unite the Parlance with the Sylviance, and on a third it passes into Petro'ica; 

 and thus effects the union of the Parlance and the Saxicolince. It should likewise 

 contain species which would shew an approximation to Accentor, and thus indicate 

 the union of the three aberrant genera of its own sub-family. Now all these 

 complicated relations may be traced in the species we already know ; and yet 

 there is one form, — the aberrant type leading to the sub-genus Dumicola, — which 

 is yet undiscovered ; and this, theoretically, might be supposed to be the very 

 one which was most essential to establish the affinity of the two genera. 



We must pass over the three remaining groups of this sub-family, namely, 

 Trichas, Accentor, and Parus, and close our remarks with a few cursory observa- 

 tions on the 



