1 59 



hunch-formers. Three other grasses, Syntherisma filiformis, 

 Aristida tuberculosa, and Cenchrus tribuloides, and two sedges, 

 Cyperus Bushii and C. filiculmis, are abundant, but do not form 

 hunches. 



The distribution of the remaining members of the associa- 

 tion, about fifty in number, is very irregular, and a group of 

 representative species can not be chosen from them. Rosa hu- 

 miilix, Solidago missouriensis, Amorpha canescens, ()j>initi<< humi- 

 fusa, Callirrhoe triangulata, and Monarda punctata are the most 

 conspicuous, but as the quantitative study of the plants shows, 

 not so numerous as Teucriumcanadense, or Ambrosia psilostachy a. 

 With the exception of Monarda punctata and the last two species, 

 these grow in rather dense, rounded patches. The same is true 

 to a lesser extent of almost every other species in this class, 

 but since the individuals are smaller or less abundant, many 

 other species may be included, or two or more patches, each 

 with traceable outlines, may overlap. If the plants are at all 

 conspicuous the overlapping is plainly seen, and in any case it 

 is brought out by quantitative study of the area. This form of 

 distribution is characteristic of associations where there are no 

 progressive changes in any of the ecological factors, and conse- 

 quently no zonal arrangement of the plants. Annuals with 

 limited means of seed dispersal, and perennials spreading by 

 rootstocks or runners, naturally grow in rounded patches un- 

 der those conditions. It is evident, then, that while the bunch- 

 grasses are representative, the other species, of merely local 

 distribution, are to be considered as secondary members of the 

 association. 



Table I. shows the distribution of the species in a piece of 

 original bunch-grass prairie. The letters after each name here 

 and in subsequent tables, indicate the number of individuals, if 

 any, in one quadrat of one hundred square feet, a signifying 1 

 to 5; b, 5-10; c, 10-25; d, 25-50: e, 50-100; l\ 100-200; g, over 

 200; and o, none. These counts were estimated for the most 

 part, although care was taken to make actual counts at inter- 

 vals in order to avoid so far as possible any serious errors of ob- 

 servation. The quadrats in this table, as well as in all the ot.h- 



