in its beari?ig on the Distribution &;c. of Birds. 239 



New Zealand, and South Africa are separated by infinitely- 

 wider divergences than those of any more northern regions; 

 therefore these southern continents must have been more or 

 less completely isolated during long periods, both from the 

 northern continent and from each other. It would appear 

 that, north of the equator at least, animals slowly migrated 

 southward, keeping pace, as it were, with the growth and 

 southward extension of the grand land- masses which appeared 

 above the sea as the globe cooled. I have alluded to the 

 identity of forms round the Polar area. We have, for in- 

 stance, two peculiar forms of the Laridcs — Pagophila and 

 Rhodostethia — girdling the Arctic circle ; so, too, the Turn- 

 stone, perhaps the most widely distributed of birds, and 

 many species of Tringce. 



But when we come further south we find, along with a 

 steadily increasing divergence, cases of closely allied species 

 recurring at intervals widely apart. Both these facts seem 

 in perfect harmony with the hypothesis of three grand prin- 

 cipal lines of migration — Western Europe, Eastern Asia, and 

 Eastern America. 



Take as an example the Woodpecker tribe. Here we 

 have the genus Picoides confined almost exclusively to the 

 most northern habitats of forest trees. We find one species, 

 with varieties barely, if at all, separable, from Norway east- 

 ward to the Rocky Mountains, and others, very closely allied, 

 across the northern belt of America. The next most 

 northerly form of Woodpecker is represented by our Picus 

 major. Of this group, the typical Picus, there may be from 

 fifty to eighty species, according to the differing views taken 

 by systematists. But whether we look to the European, 

 Asiatic, or American forms, we find a close similarity in the 

 most northerly species ; while as we proceed southward the 

 species become more and more distinct, until, before entering 

 the tropics, we have lost the genus altogether. Picus major 

 is identical in Britain and Japan ; the North -American 

 P. pubescens and P. harrisi of the west of the Rocky Moun- 

 tains are not far separate. From each we find further and 

 further divergences till we come to Formosa, Algeria, and 



