218 MINNESOTA BOTANICAL STUDIES 



Clements (56, 57). The former region is characterized especially 

 by the presence of turf-forming grasses, the latter by species that 

 have acquired the bunch-grass habit. 



The chief association of the prairie-grass formation in the 

 sandy portions of the formation, especially where definite sandhills 

 occur, is the bunch-grass association which is to be considered as a 

 temporary climax association of the formation in such regions. In 

 those portions of the formation with a more stable substratum (of 

 an argillaceous, loessal, drift, or loamy nature) various associations 

 of long-stemmed grasses possessed of the sod-forming habit pre- 

 dominate. With these we are not especially concerned in this paper 

 although of necessity they must be noted in connection with certain 

 possible successional phenomena. 



I wish to insist furthermore that this region is dominated by 

 grass species of a similar physiognomy but at the same time quite 

 different from those grasses that control wide stretches of upland 

 between this formation and the foothills of the Rocky Mountains. 

 As we go westward we find that certain other types of grassland 

 become conspicuous under certain soil conditions and an arid climate. 

 Certain prairie-grasses and short-grasses become associated as alter- 

 nating associations. This common condition is revealed in various 

 portions of the Great Plains. I wish then to emphasize the desir- 

 ability of maintaining in our ecological classification a prairie-grass 

 formation toward the eastern portion of the Trans-Mississippi region^ 

 as ecologically distinct from a short-grass formation lying between 

 the former formation and the base of the Rocky Mountains. After 

 a rather extended investigation including both areas, the facts of 

 distribution seem to warrant the above conception which appears to 

 coincide with the ideas of Shantz (60) and others who have had an 

 even wider experience in the great belt of grassland stretching from 

 the broadleaf and conifer forests of the east to the coniferous forests 

 of the Rockies. 



There can be little doubt, especially after the most admirable 

 work done by Shantz (59, 60), that the two complexes of limiting 

 factors in the differentiation of these two great plant formations are 

 climate and competition. Precipitation and competition are cer- 

 tainly the crucial phases of the problem instead of temperature as 

 Merriam (47) and his followers conclude. 



The prairie-grass formation is limited on the east by that com- 

 bination of climatic and edaphic factors that makes tree growth 



