1 66 THE MIGRATION OF BIRDS. 



a natural division unless it contains a good round 

 number of species. If the species are few, that is 

 no reason for lumping them into one or two genera, 

 at the cost of concealing some of their most inter- 

 esting features. Thus the Shore Larks fairly claim 

 distinction from all other Larks, under the generic 

 name of Otocoris ; the Grosbeaks from all other 

 Finches under that of Pinicola ; the Eider Ducks 

 under that of Somateria ; the Little Auk under that 

 of Mergulus ; the Gulls under Pagopliila, Glaucus, 

 and Rliodostethia. In my opinion, if a genus 

 illustrates or implies any important fact, either 

 geographical or biological, it ought to be retained, 

 even if the species it contains are few. 



It might also be remarked of these nomadic 

 migrants, that none of them are represented by 

 closely allied forms in the Southern Hemisphere, 

 and many are exclusively Arctic, a fact of great 

 significance, indicating very restricted migratory 

 movements through all Avian time, and probably 

 confined close to the limits of glaciation. The 

 geographical distribution of the Shore Larks [Oto- 

 coris) or the Waxwings [^mpelis) illustrate this. I 

 will select the former genus because it seems the 

 most anomalous, and yet after all it is not abnormal 

 in any respect. All the Shore Larks are nomadic 

 migrants, or actually sedentary, and number six 

 more or less clearly defined species or races, all 

 obviously very closely allied. Probably they formed 

 one circumpolar species previous to the Post-Pliocene 

 Glacial Epoch. Driven south by the advancing 



