93 



indicates that the level, in some places at least, has changed but 

 little since the preceding autumn. Elsewhere seedlings appear on 

 ground without dead stems, indicating that conditions were prob- 

 ably unfavorable for seed planting during the previous year. It 

 may be assumed also that certain tracts covered with plants during 

 one year may be bare the next, because of some slight change in 

 the velocity or direction of the sand movement. So the position 

 and extent of the blowsand association vary from year to year, 

 now extending lower and possibly surrounding some relics of the 

 basin association, now retreating tow'ard the summit of the slope, 

 but always appearing where the movement of the sand tends to bury 

 the seeds to a small, but sufficient depth. 



The species of the association vai-y in their ability to extend out 

 upon the sand. Aristida tuberculosa is ahvays the pioneer, and the 

 margin of the association frequently consists of that species alone. 

 This is probably due to the awned grains, which may be able to bury 

 themselves to some slight depth. The grains of Stipa spartea, with 

 much longer and stiffer bent awns, are known to bury themselves to 

 a depth of about two inches (5 cm.). Places most densely covered 

 with Aristida usually have several other species as well, and their 

 contour generally shows that upon them small deposits, generally 

 less than an inch (2.5 cm.) deep, have taken place. Other conditions 

 being- eliminated, small seeds are more apt to be buried than large 

 ones, and it is at once noticeable that the individuals of species with 

 small seeds are vastly more numerous than those with larger ones, as 

 Cassia ChauiaccJirista. 



The necessity of seed burial is strikingly illustrated by seedlings 

 coming up in rows over wagon tracks. This has already been men- 

 tioned for Cassia Chamaechrista (Hart and Gleason, igoy: 165) 

 and, in a short note (Amer. Botanist y: 91), for Diodia teres. A 

 blowout in the Hanover area illustrates the effect especially well (PI. 

 Vni, Fig. 2). At the very edg'e of the lee slope, where erosion has 

 probably exceeded deposition, there are several curving rows of 

 Diodia teres, marking the tracks of a wagon which had been driven 

 in a curve across the sand. To be effective this artificial planting must 

 be deep enough to prevent the seeds from being uncovered by any sub- 

 sequent erosion. 



The species comprising the association, arranged approximately 

 in the order of their abundance, are as follows: 



Aristida tuberculosa Comvielina virginica 



Paspalum setaceuui Cenchrns carolinianus 



Diodia teres Ambrosia psilostachya 



