ECOLOGY OF NORTHERN MICHIGAN. 45 



a softer inter-bedded clastic. This contrast between the resistance offered 

 by the sandstone and trap sheets to erosion has given rise, in a small degree, 

 to what Marbut ('96, pp. 29-32) has called step and platform topography. 

 The cross section. Fig. 2, shows that the first ridge is formed by the pro- 

 jecting end of a stratum of trap; if this sheet had been vertical, erosion 

 would have been equal on both sides and the slope on either side would have 

 been the same. But, owing to the fact that the sheet is inclined, the out- 

 cropping edge protects the underlying sandstone and a precipitous escarp- 

 ment is formed. This section also shows that the valley of Carp river is 

 formed by the erosion of the softer stratum of sandstone lying between the 

 escarpment and the second ridge of trap, and that the stream runs along 

 the strike of the rocks, so that it is difficult to believe with Dr. Wright ('05, 

 p. 38) that the valley was formed by the submarine erosion of post-glacial 

 lakes. 



If the second sheet of trap was, like the first, underlaid by a bed of detrital 

 material, another escarpment would have been formed. But instead of 

 overlaying an interbedded clastic, it overlies the hard resistant quartzi- 

 ferous porphry that constitutes the central part of the mountainous district. 

 This results in the topography that would prevail if the resistant stratum 

 were vertical, i. e. equal slopes on either side. The topography, however 

 is not of recent origin, and although the physiographic history of this part 

 of Michigan has not been worked out by geologists, one is lead to believe 

 from the work of Wilson in Canada ('03), Weidman in Wisconsin ('03), 

 and Van Hise in northern Michigan ('94 and '96) that the baselevel that is 

 represented by the truncated strata, Fig. 2, probably represents a pre- 

 Cambrian peneplain. 



Prof. C. K. Leith has suggested to me in a letter that this structure possibly 

 represents a peneplain of later age than the Wisconsin pre-Cambrian, which 

 it probably intersects at a low angle in a manner similar to the intersection 

 of peneplains of different ages about the Laurentian of Canada as described 

 by Wilson ('03, p. 651). The Porcupine Mountains would thus represent 

 the remains of a peneplain, buried since early geological time under later 

 deposits. 



But whether the topography is of pre-Cambrian or later age, it is largely 

 determined by the geological structure and has been comparatively little 

 modified by the ice age. Thus Van Hise ('04, p. 35) has shown that the 

 rocks at present exposed in this part of the Lake Superior region were buried 

 under later deposits all through the Palaeozoic Era and had probably become 

 approximately adjusted to those conditions. Later when they were brought 

 to the surface by the extensive denudation of Cretaceous and Tertiary 

 Periods, and still later when the thin layer of weathered material had been 

 removed by the ice sheets of the Glacial Epoch, tliis equilibrium was disturbed, 

 and they at once began to adjust themselves to the new conditions — a 

 process that is not yet completed. 



At the close of the Tertiary Period, the Tertiary peneplain was destroyed 

 by an elevation of 3,000 to 5,000 feet (Upham, '04, p. 244) over northern 

 North America. Near the culmination of this uplift, three centers of ice 

 accumulation developed in this region and gave rise to the continental ice 

 sheets of the Glacial Epoch. 



The glaciers that were formed from these three centers covered Canada 

 and invaded the northern part of the United States at various times, but 

 the final invasion was the most important from the standpoint of the present 

 biota. At this time, ice sheets fed by the different centers united into a 



