PAPERS ON BOTANY 245 



apparent permanency of a primitive type of forest within 

 a region of a mesophytic climax seems established be- 

 yond qnestion, and also the further fact that this scrub 

 pine forest is in both Michigan and New Jersey asso- 

 ciated with a soil of almost pure sand and a topography 

 of low relief. 



In Michigan there exist other areas of sandy soil ex- 

 hibiting quite different conditions of forest development. 

 These are to be found more particularly along the west- 

 ern shore of the lower peninsula in the form of sand- 

 dune areas of varied extent. The soil does not seem to 

 be essentially ditferent than that of the sand plains some 

 miles farther inland, iii fact, if there is a difference it 

 seems to lie in the direction of the dunes having more 

 nearly absolutely pure sand, free from any other soil 

 whatever, than the sandy plains. The latter have fre- 

 quently some small admixture of loam or similar fine 

 material. But while the soil is similar the same cannot 

 be said of the vegetation for while primitive stages 

 closely resembling those of the sand plains are evident 

 in portions of the forest cover of the dunes more par- 

 ticularly in newly formed areas, much of the older por- 

 tions are covered by the climax forest of the region. 

 Such a forest has been described by Waterman (8) in the 

 Frankfort region, and it has also been seen by the writer 

 at many different points, extending from Sawyer, at the 

 south end of Lake Michigan to Bay View, Fox and 

 Beaver Islands well to the north. This climax forest, 

 although developed upon pure sand, differs very slightly 

 from that upon other soils in the region the dominating 

 trees being beech, maple and hemlock. 



It is true that some dune areas such as those covering 

 a portion of the Big Sable Point area, near Ludington, 

 seem to have remained for a very long time with a very 

 primitive vegetation cover, but adjacent portions of the 

 same area are in climax forest. 



What has caused this great difference in the rate of 

 succession in these sand areas? In the opinion of the 

 writer the differences in the topography relief have 

 played an important part. It is a fact established by 



