86 BOTANICAL GAZETTE [AucusT 
taries, reached the prairie region they were unable to occupy the high 
lands, but occupied the flood plains and adjacent slopes. There is 
no evidence to indicate that the then existing topographical and cli- 
matic conditions differed radically from those of today. Again these 
elements have today their most widespread distribution, a condition 
at once intelligible when related to physiographic development in the 
working-back of streams and the increase of flood plain. ‘Two salient 
points may now be noted: the initial and the continuous subsequent 
preemption of the upland by the prairie formation, and secondly the 
continuation of initial climatic conditions. The initial causes which 
operated to restrict tree invasion and ecesis upon these prairie-covered 
uplands would thus seem to have their duplication in those natural 
factors which operate to that end at the present time. 
Under the discussion of ecological factors of the region it will be 
shown that there is a coincidence of factors operating most antago- 
nistically against tree growth upon the upland. The almost entire 
absence of fungi upon the prairie cannot but be significant. The 
roots of most of these deciduous trees are obligatively provided with 
symbiotic mycorhiza and the absence of their specific fungus would 
preclude advance. Hence fungal infection of the prairie soil must 
precede or at least accompany forest encroachment. Again, the 
difficulty of seed germination, almost impossible either because of 4 
dense sod or a lack of soil moisture, successfully checks invasion. 
In consideration of these precarious climatic factors, peculiar edaphic 
conditions, and the fact that if planted upon the prairie trees thrive, 
I am led to the view that the question of non-invasion upon the 
prairie proper is primarily and initially one of pre-occupation and the 
inability of seedlings successfully to withstand the extremely seve 
conditions of the first winter’s exposure. In the positive and coinc 
dent interaction of unfavorable biological and climatic condition 
may be found a cause sufficient to account for tree absence upo2 the 
prairie and the slow migration of tree species into this region. 
While these causes are all sufficient, yet we must not disregard # 
secondary and artificial though highly cooperative factor, which ™ 
recent times must have served in many places to prevent tree &'*" 
lishment. I refer to prairie fires; yet even in the absence of prairie 
fires for half a century the prairie stands uninvaded except in cas® : 
