﻿4^ BOTANICAL GAZETTE [january 



western part of the county is found the only marked instance of 

 the presence of hemlock. This tree belongs typically with the 

 hard wood group in northern Michigan. 



In the northern part of the county white pine was almost 

 universally present in the uplands at the time of settlement. This 

 can be proven by stumps which are still in place or have been 

 used in the construction of stump fences. There are pine stumps 

 and a few trees still standing even in the beech-maple group upon 

 areas north of an east and west line drawn through a point about 

 midway between Cedar Springs and Rockford. South of this 

 line the pine disappears in society I and becomes very rare in 

 all but IV and V. 



3. Generalizations on the upland flora. — Any sort of general- 



r 



ization upon the study of such a limited area as the present must 

 necessarily be a hazardous undertaking. I venture to call atten- 

 tion to the following points, however, all of which must be 

 looked upon as merely tentative suggestions: 



a. The soil factor, — It appears that the general distribution of 

 the upland societies is based primarily upon the nature of the 

 superficial soil. This must be so, since the roots of the smaller 

 plants never penetrate very far into the soil, and since, in the 

 case of trees and shrubs seed germination and the growth of 

 seedlings is conditioned by the surface layers. If seedlings 

 cannot develop it is clear that there can be no mature plants. 



It is very probable that the original till material covering 

 Kent county was reasonably uniform in chemical constituents. 

 That it was thoroughly mixed by the movement of the ice sheet 

 is shown by the wealth of different minerals to be found in any 

 small region. In spite of the marked washing, the most sandy 

 soils contain a considerable amount of minerals other than 

 silica.^ The plant can make use of the soil constituents only 

 after they are in aqueous solution. The great bulk of the soil is 

 practically insoluble in water, and it makes no difference to the 

 plant what may be the chemical nature of these undissolved sub- 

 stances. Thus it would seem much more to the point to make 



1, 



8 



Kedzie, R. C, Analysis of soil of jack-pine plains near Grayling, Michigan. 



Annual Report Mich. Board of Agriculture 27 : 211. 1888 ; also Bull. 99, i893 



