PAPERS ON BIOLOGY AND AGRICULTURE 159 



Thus the wooded area on either side of the Rock, in its 

 original extent, was about equaL See the sketch map on 

 page 160. In the typical prairie river wooded region the 

 gi-eater extent of the woods is on the west of the river. 

 The potential mesophytism, if not the actual mesophytism, 

 is greater in degree and in extent in the Rock River wood- 

 lands region than in that of any true prairie river. An 

 examination of the Rock in Ogle County shows that the 

 intrenched valley must expose here on either sloping side 

 a set of soils that is different from the prairie soil. This 

 is typical of the middle and lower courses of the prairie 

 rivers. In general, too, the east side of the valley is wider 

 than the west side, and the east is also generally of some- 

 what lower slope. This, again, is typical of the prairie river. 

 On the Rock River this is locally greatly broken into an ac- 

 count of both the preglacial character of much of its course 

 in Ogle and the physiographic diversity of the postglacial 

 portion, so that the statements hold true only for the general 

 aspect of the entire course in the county. 



In part this transition belt is a savanna. This term is of- 

 ten used in order to compare it with other similar areas of 

 the world that go by this general name. Many parts of 

 this North American transition unquestionably furnish ex- 

 amples of true savanna, chiefly of the patchy type, where, 

 in a grassland, patches of tree growth frequently occur. 

 This is almost the same thing as the "oak openings" often 

 mentioned in literature descriptive of the early days of 

 settlement, with the possibility that the oak openings com- 

 prised rather more frequent wooded portions than character- 

 istic patchy savanna. In some places the park-like savanna 

 occurred. Here the grassland was set not with patches of 

 tree growth but with isolated trees far enough apart to 

 make it possible to drive about through the area. The 

 early records of Ogle County make reference to such a 

 form in Mount Morris township and the reference makes 

 it highly probable that this example arose as the result of 

 prairie fires. Other examples must have existed and there 

 is likelihood that prairie fires were an important if not de- 

 cisive element in their creation. 



