ECOLOGICAL RELATIONS OF THE FLORA. 59 



However, more distinctly northern and southern species are found here 

 than in the interior parts of the state at the same latitude. Adams 

 (1902) has pointed out that the post-glacial migration of plants and 

 animals into Michigan was principally from southeastern United States. 

 That the southern forms tend to push farther northward on the western 

 side of the state is shown by the comparison of the species reported for 

 Roscommon County with those found in Kent County. (Livingston, 

 1903 and 1901.) Similarly the flora of Sand Point is more southern 

 than that of Roscommon County as also is the flora of Manistee and 

 Sturgis as reported by Daniels (1904) or of Manitou Island as reported 

 by Whitford (1901). This is probably due in part to the more equable 

 climate of the coasts. Whatever the explanation, however, the fact 

 stands that the vegetation has advanced in a V-shaped manner, the 

 central portion of the state lagging behind the border portion. 

 Typical southern species which show this advance are Benzoin aestivale. 

 Sassafras variifolimn, Asimina triloba, Liriodendron tulipifera, Celastrus 

 scandens, Gleditsia triacanthos, Cornus florida, Cercis canadensis, etc. 



Similar to this movement northward by southern forms is the move- 

 ment (or persistence) southward by northern forms for which no ex- 

 planation is offered. This is also V-shaped following the shores of 

 Lakes Michigan and Huron, such species as Pinus hanksiana, Pinus 

 strohus, Tsuga canadensis, Betula lutea, Taxus canadensis, Abies 

 halsamea, Equisetum hyemale var. robustum, Clintonia borealis, Salix 

 glaucophylla, being reported from Northern Indiana (Hill, 1896), 

 Chicago (Cowles, 1899), South Haven (Bailey, 1880) on the west and 

 from Port Huron on the east side (Dodge, 1897) but are rarely found 

 south of 43° in the interior of the state (Beal, 1904). 



Several classifications of the region have been made, all largel}" based 

 on the type of forest. Schimper (1903, p. 571) has followed Sargent 

 (1884) in including this in the Northern Coniferous Forest Area. 



A comparison of the floral lists given by Cowles for Chicago and 

 vicinity, Jennings (1909) for Lake Erie, Dodge for Port Huron, 

 Daniels for Manistee and Sturgis, Coulter (1904) and Whitford (1901) 

 for Manitou Island, Ruthven for Ontonagon County (see Adams, 1906) 

 and the compiled records from Beal and Gray's Manual (7 ed.) show 

 that, as far as Michigan is concerned, the flora of the Great Lakes region 

 is strikingly a unit, not alone as regards species but also in the com- 

 position of the plant societies. Furthermore, the genetics of the plant 

 formations are very similar and the formations are passing from a het- 

 erogeneous aggregation into a homogeneous climax if we follow Cowles 

 and Whitford. Now the basis of province delineation is similarity 

 of constituents of formation or climax formation, and the limits of the 

 province that includes the Great Lakes flora seem very well marked 



