plates of similar cells in fig. 4: ; or forming transverse or vertical 

 plates, as in fig. 3. 



The prairie plants also exhibit the ordinary narrow, elongated 

 palisade cells so common in xerophytes which are exposed to 

 strong light on the treeless tracts which form their habitat. 

 Where the leaf is covered with strong cutin, or with abundant 

 trichomes, either of which reduces the intensity of the entering 

 light, the palisade cells are more rounded, as illustrated in Plate 

 IV, fig. 2. 



It is because of these xerophytic adaptations that the prairie 

 flora is able to take possession of the exposed tracts on which it is 

 ordinarily found. The niesophytes of the forest usually perish 

 in a few houi^s if transferred to these exposed tracts, as the writer 

 has determined by repeated experiments. 



The flowering plants form the dominant vegetation of the 

 prairies. The few Thallophyta are for the most part parasitic 

 on flowering plants, the prairie Bryophyta are almost a negli- 

 gible quantity in this region, and the Pteridophyta are repre- 

 sented only by the genus Equisetum. 



An inspection of the list of xerophytic plants of this region 

 (compare especially columns 1 to 4 and 5 to 6 in the vascular 

 list) again shows that the flora of the sandy areas is essentially 

 the same as that of the prairie, a fact which emphasizes the con- 

 clusion that fineness of the soil does not determine the distribu- 

 tion of the prairie plants. 



Very much more could, and probably should be written con- 

 cerning the peculiarities of plant associations on the prairies. 

 Most of the species have their seeds distributed by wind, and the 

 variability of this factor produces marked changes from season 

 to season in the relative abundance and distribution of annual 

 plants, and in the end strongly influences and finally deter- 

 mines the distribution of the perennial plants. The accident of 

 distribution because of this fluctuating factor enters largely into 

 the determination of particular groupings, and suggests the in- 

 advisability of basing ecological classifications or determination 

 of prairie plant formations on the particular grouping which is 

 presented in a comparatively restricted area. It is for this rea- 

 son that the quadrat method of study of relative distribution of 

 prairie plants has proven unsatisfactory and misleading. 



A comparison of the floras of two very similar localities will 



25 



