THE PEATRIES 219 



the prairies would have prevented any such encroachment, but 

 in the hilly western part of the state, where the surface remained 

 practicallj- undisturbed, there has been only a slight extension 

 of the borders of the grooves. The oldest settlers in that part 

 of the state agree that since the cessation of prairie fires no new 

 groves have appeared except where set out by man. and that the 

 native groves have expanded very little, though they have become 

 denser and where not pastured have developed a dense under- 

 growth. This latter fact shows that the groves were also period- 

 ically swept by fires, and that the fires were not sufficient to 

 destroy the groves. 



That prairies appear in drier places has been shown in this 

 paper, but this very fact has often so depauperated the vege- 

 tation that it would oflier scant fuel for fires. The distribution 

 of native groves along streams is by no means consistent with the 

 view that the streams checked great conflagrations, for some- 

 times the groves are on one side of a stream, sometimes on the 

 other, and where the stream has not cut a deep channel neither 

 shore is wooded. The distribution of these groves is much more 

 consistent with the causes presented in this paper. 



The small bur-oaks have not increased in size in those places 

 where they have remained undisturbed since the cessation of fires. 

 in some cases more than thirty years, and it is evident that fires 

 were not responsible for their condition. These stunted oaks are 

 found invariably in dry. exposed places, and they produce large 

 roots and small tops now just as they did when the fires swept 

 the prairies. 



It is evident that fires constituted no more than a local second- 

 ary cause, but one of the most conclusive evidences that annual 

 fires were not the prime cause of the prairie is furnished by the 

 prairie openings which are still found in the forested sections 

 of this region, and which were formerly a very striking feature 

 of our upland forests. 



These openings differ from the forest not only in the absence 

 of trees, but also in the fact that the minor mesophytic forest 

 flora is also absent and the surface is covered with a typical 

 prairie vegetation. This is especially noteworthy in the smaller 

 areas which are often far remote from the broader prairies and 

 yet develop a typical prairie flora. 



