﻿^903] UPLAND PLAN! SOCIETIES 53 



margins of the numerous ponds and lakes where societies IV and 

 V hold the ground. In slight depressions along the margin of 

 the plain the humus is deep and society I is common. There 

 are also many rather large areas of societies I and III well out 

 in. the plain. Usually these are in slight depressions, not low 

 enough to be swampy, but well covered with humus. We may 

 say, then, that in the most southern of the three sand plains 

 which have been studied, and therefore the one which has been 

 out of water and fit for vegetation the longest, the predominant 

 society is III ; but I and 11 are not uncommon ; while IV and V 

 occupy a relatively small portion of the area. 



In the Grand Rapids sand plain, society IV is predominant, 

 with a good part occupied by V and comparativ^ely very little 

 by I, II, and III. And in the plain which extends from Rock- 

 ford northward, the only upland society found is V. Of course 

 the last plain has been out of the water a much shorter time 

 than the other two. In fact, a great part of it is at present 

 swamp and is occupied by lowland societies. 



In these three plains, perhaps , are seen successive stages 

 of vegetation occupying successive stages in the formation and 

 accumulation of humus. Of course the extensive destruction of 

 the natural vegetation which has taken place since settlement of 

 this region began will make it impossible for the natural course 

 of events to continue here, even if the above hypothesis be the 

 correct one. Clearing and burning have often reduced the soil 

 from a condition suited to society II or III to one only fit for 

 society V. This is probably also true on the moraines in the 

 northern part of the county. It will probably be impossible 

 ever to trace the history much farther than it had gone at the 

 time of settlement. 



CONCLUSIONS. 



The present observations appear to justify the retention, as a 

 broad general hypothesis, of the physiographic idea advanced 

 Cowles, namely, that physiography determines vegetation. 

 But this hypothesis does not come close enough to* the ulti- 

 oiate factor upon which depends the distribution of the plant 



* 

 h 



