208 Canon Tristram on the Tolar Origin of L\fe. 



explain the distribution of species, and to account for the 

 phenomena of migration. We have assumed that there were 

 but three principal lines of migration southward. Why, it 

 may be asked, should we assume these special lines ? Why 

 may not the exiles have departed over land in any direction ? 

 That the fourth coast-line, down the east of North America, 

 was unimportant — if followed at all — appears from the fact 

 that the central area of North America was at that period 

 the basin of a vast shallow sea, with, perhaps, a few islets ; 

 and that south of New England the land continuity must 

 have been broken. And as to the departing overland in any 

 direction, our hypothesis assumes the heredity of habits, and 

 we know that for the most part migrants, where possible, hug 

 the coast-line or follow the bases of mountain-ranges. As- 

 suming, too, the reluctance of the first exiles to move, this 

 being also founded on the heredity we now observe, they 

 would be shut out by the frost from the interior lands, and 

 would congregate on the river-banks and along the shores 

 for food, as many of our native non-migrants do now — e-g-, 

 the Rook — before they finally took flight for the south. On 

 all these questions connected with distribution and migra- 

 tion the various papers of Mr. Seebohm in ' The Ibis ' 

 and elsewhere have been most instructive, and he has 

 dealt with the subject at much greater length in his new 

 work on the Waders, of which he has kindly allowed me to 

 see the advanced sheets. I am aware that Mr. Seebohm 

 invokes glacial epochs and alternations of climate extending 

 over enormous eras uniformly in the circumpolar area to 

 explain the present distribution of bird-life, and that he 

 would also attribute to some — as to the Hinmdinidce — a South 

 Polar origin. I am not prepared to contest his theory of 

 past geologic history, but only to submit that they are not 

 necessary for our solution. 



I have in my former paper taken the distribution of the 

 Picidte and of Pica as illustrations. Before adducing other 

 groups, I may, perhaps, be allowed to lay down certain 

 generalized deductions on migration and distribution : — 



