58 PSYCHE [April 



Central America and Cuba, the latter being a branch of trend C, the possible 

 course of D being indicated on the basis of the trend of diffusion of the intro- 

 duced Aphodius lividiis. Cy /as formica ri us, the pest of the sweet potato, may have 

 followed this course, but of this the proof is lacking. The Antillean trend may, 

 and again may not, extend northward, as in C. What I have termed the Trans- 

 appalachian trend, E, is intended to indicate the trend of diffusion especially, 

 though not exclusively, of species introduced through the agencies of man. Most 

 of these, whether they become established along the coast from Virginia north- 

 ward, or in Canada as far up the St. Lawrence river as Quebec, make their way 

 into New York and west, passing between the low mountains and the south shore 

 of Lake Erie into the comparatively level country beyond. The Boreal trend, F, 

 from the northwest, may perhaps follow more closely to the coast, as indicated by 

 the dotted line, but sooner or later it sweeps broadly to the east to the Atlantic 

 coast and New England, not infrequently sending subtrends into California, Col- 

 orado, and even Mexico. I have not indicated southward trends so fully, because 

 there are less of them ; they are less important to this discussion, and too many 

 lines, especially in the far south, would complicate and obscure the map, the land 

 as shown being too narrow to display them clearly separated. 



Illinois State Laboratory of Natural History, 

 Urbana, 111., Mar. 9, 1903. 



