32 BOTANICAL GAZETTE [JANUAR 
The Norway pine type occupies gravelly ridges and loamy saudi 
plains. The soil here is somewhat lighter than in the locations hei! , 
by the last-named type, but it is generally too poor for profita’! 
agriculture. As will be seen by a glance at the map, most of + 1° 
uplands here studied were originally covered by this type 
The jack pine type occupies only the most thoroughly washed 
the sand plains. The localities held by this open formation lie » 
the valley of the Au Sable. The parts lying about the headwaters 
of the Muskegon have abundant plains of loamy sand, but these 
support the type of Norway pine. This fact has no connection w''” 
the rivers themselves, however, for farther down the Muskegon «'- 
to be found typical jack pine barrens. The soil of the jack pine ty) 
is almost worthless for agriculture; it is light and dry, and where t»« 
surface is broken it is apt to be wind-blown, and often forms sm 
traveling dunes. 
Lowlands. 
The distribution of the lowland type was not worked out wi) 
accuracy. Great difficulty was experienced in studying such areas, 7: 
the swamps are often almost utterly impassable. The greatest are“: 
of open marsh encountered are marked on the map, as alreac) 
described. There are doubtless many areas of like nature whic’ — 
were not seen at all, but these cannot be of very great extent. in 
all the swamps the ground is covered with a layer of humus, i: : 
of the nature of peat, and there seems to be no difference in ti is” 
substance between the sand and the clay areas. Neither is there any | 
apparent difference in the swamp enaiee whether it is upon sa 
or clay. a 
RELATION BETWEEN DISTRIBUTION OF FOREST TYPES AND THAT 75 | 
| SOIL TYPES. ee 
The upland types seem to be very closely dependent upon "| 
nearness of the underground water level to the surface and upon "8° > 
nature of the soil. The former factor determines at once whet)" 
the vegetation shall be classified as upland or lowland. The © 
- tinction between these two classes is more evident on the surfs 
than it is after closer study; it is difficult to state just how far © 
water level may recede from the surface and still support a lowl 
