forming plants, remains more or less in motion and affords an en- 

 vironment snitable for the development of the blowsand associa- 

 tion, which then becomes dominant. This condition is much like 

 that on the broad flat deposits without dune- formers, where the rate 

 of deposition is slow, because of the large surface to be covered, and 

 the vegetation accordingly consists of the blowsand association. On 

 the deposits this association persists longer than in the basin, be- 

 cause the greater exposure of the sand to the wind keeps it longer 

 in motion. Finally the motion stops, and the deposits are also com- 

 pletely stabilized by various outside species, chiefly bunch-grasses. 

 On the steeper slopes, held by Rhus canadensis, var. illinoensis, Pan- 

 iciini I'irgatuiu, or other species constituting the real deposit asso- 

 ciation, there is less opportunity for the development of the blow- 

 sand vegetation, because the perennials persist and retain their dom- 

 inancy. They are finally joined by additional species, until eventu- 

 ally the siu'face is covered and the sand completely stabilized. 



Summing up the successions within a. single blowout (Fig'ure 5), 

 it is seen that there exists a perfect correlation between the vegetation 

 and the physical conditions of the environment. The original Panicinn 

 pseudopubescens association is succeeded by each of the four blowout 

 associations. Between these four, the successions depend partly upon 

 the direction of movement of the blowout as a whole. The wind- 

 ward slope association may be succeeded by the basin association, but 

 the reverse does not take place. Between the other three, the suc- 

 cession may be in either direction. The blowsand association shows 

 a general tendency eventually to succeed l^oth of the others, and may 

 be regarded as the climax association of the blowout formation. This 

 is directly correlated with the general dynamic effect of the w-ind, 

 which leads, on the average, neither to erosion nor deposition, but 

 merely to the movement of the sand. This condition is most favor- 

 able to the blowsand association, and is the cause of its dominancy 

 It is probable that the blowsand vegetation would also appear on the 

 windward slope after it has become static, but the relic bunch-grasses 

 become at once the controlling feature, among which the blowsand 

 species play a secondary part as interstitials. 



When one blowout is filled l)y the deposits of another, or when 

 secondary blowouts appear on the lee slope or deposits of an older 

 one, there may be deviations from this normal series of successions. 

 In the latter case the young blowout may be wholly or partially sur- 

 rounded by a blowsand association, and the windward slope associ- 

 ation is never developed because of the absence of any relic plants. 

 When two or more blowouts unite to form a complex w^aste of sand, 



