Packard.] 18 [May 7, 



forms not of tropical origin, are the factors needed to work out the 

 problem. That something of this sort has taken place in marine 

 animals we know to be the fact. Certain forms now supposed to be 

 extinct on the coast of New England and Scandinavia, such as Yoldia 

 arctica Gray (Nucula Portlandica Hitchcock), are still living in the 

 seas of Greenland and Spitzbergen. The quaternary fauna of Maine 

 indicates a much more purely arctic assemblage than is at present to 

 be found. This is also the case with the Scandinavian quaternary 

 fauna, according to the researches of Prof. M. Sars. As we have 

 before shown, the circumpolar marine fauna runs down along the 

 coast of northeastern America and of Europe, and the forms common 

 to the two shores have not come one from the other. Europe has not 

 perhaps borrowed in quaternary times from America, but both have 

 been peopled from a purely circumpolar fauna. If there has been any 

 borrowing it has been on the part of Europe, since the fossil musk 

 ox of France and Central Europe is said to be identical with the musk 

 ox of Arctic America. So also on the coast of northeastern Asia and 

 Alaska are circumpolar forms, which have evidently followed the 

 flow of the arctic currents down each coast. The forms which are 

 identical or representative on these two coasts are species derived from 

 the circumpolar fauna; so the forms which are so strikingly similar in 

 northern Japan to those on the coast of New England are, if we 

 mistake not, also derived from the northward. I believe it to be a 

 matter of fact that the Atlantic States species of insects which are 

 common to the two countries, are, if not of circumpolar, at least of 

 subarctical or boreal origin. From these facts we are led to accept the 

 conclusions of Gray and Lesquereux, that co-specific or congeneric 

 forms occurring in California and Europe and Asia, are the remnants 

 of a southward migration from polar tertiary lands during tertiary, 

 and even perhaps cretaceous times; and in proportion to the high 

 antiquity of the migrations there have been changes and extinctions 

 causing the present anomalies in the distribution of organized beings 

 which are now so difficult to account for on any other hypothesis. 



For this reason it is not improbable that those species of insects 

 which are more or less cosmopolite (and independently so of human 

 agency) are the most ancient, just as some forms taxonomically the 

 most remote are remnants of earlier geological periods. For exam- 

 ple, the curious anomalies in the geographical distribution of Lirnulus, 

 the genus only occurring on the eastern coasts of Asia and North 

 America, accord with its isolation from other Crustacea. Geological 



