11 



creased, and the temperature probably became somewhat cooler. It 

 is an unpublished opinion of Professor Gleason that the eastern, or 

 prairie-grass formation, had not developed before the change in cli- 

 mate, and that this gradual change favored the evolution of a meso- 

 phytic type of prairie over the whole eastern extent of the province. 

 The uplands on both sides of the river, surrounding the sand-plain, 

 were soon covered with this black-soil prairie, to the exclusion of the 

 older xerophytic type, which was, however, preserved with little or no 

 change on the sand ridges. At the same time, or a little later, hydro- 

 phytic members of the present swamp-prairie association appeared on 

 the sandy loam flats and began to compete with the northern bog 

 plants, which gradually gave way to their western rivals, except in a 

 few localities unfavorable to the invaders. 



Another far reaching effect of the increase of rainfall and humid- 

 ity w^as the spread westw'ard and northward of the deciduous forest, 

 which during the arid period was developed only a short distance 

 north of the Ohio River. The northern migration in the eastern states 

 (Adams, '05) was aided materially by a northward moving wave of 

 temperature increase, which followed the retreat of the ice-sheet. 

 This extension of the forest to the north and west is evidently still 

 in progress, though it is obscured by the clearing of land for agricul- 

 ture. In the Havana region the forest probably first came in from 

 the southwest by means of the Mississippi and Illinois rivers. The 

 bottomland association was the first forest type in the up-stream mi- 

 gration, willow and poplar being the pioneer species ; and this forest 

 type came to occupy a narrow strip along the margin of the sand- 

 plain, and probably the whole new flood-plain of the western edge of 

 the valley. The semi-xerophytic and mesophytic oak forests followed 

 the bottomland type up the Illinois, and spread laterally from the 

 river, replacing to a large extent the black-soil prairie of the uplands. 

 The forest could not at first invade the sand region : the level areas 

 were covered wath xerophytic prairie and swamp prairie ; the sand 

 ridges in their open condition were distinctly unfavorable to forest 

 growth. In time, however, the xerophytic black-jack and black oak 

 were able to establish themselves on the sand ridges. The succession 

 from bunch-grass to forest is probably still going on very rapidly in 

 the Illinois River valley region. This fact and the records of early 

 explorers lead us to think that the forest invasion is very recent (see 

 Gleason, '10: 120). The manner in wdiich the succession from bunch- 

 grass to forest takes place is not known. Once the forest is well es- 

 tablished, the ^■egetation assumes complete control of the environment, 

 and save for the factor of leaching, the peculiar unfavorable qualities 



