Ten pounds of eolian sand were dyed fluorescent red and released on the crest of a large shadow dune at Windy Point, San Gorgonio Pass, California. The dune surface was sampled by pressing 3×3-in. Vaseline-coated cards onto the sands at predetermined stations downwind of the tracer point source. Westerly winds blew at 15-25 m.p.h. during the test.

Sample cards were examined under ultraviolet light and the number of fluorescent grains per square inch determined for elapsed times of 3, 20 and 60 min. after tracer release. Isopleths of equal tracer concentration at the three elapsed times all delineated lobate patterns presumably in response to variable wind "streamlines" over the dune.

At 2.5 hrs. after tracer release a lag deposit of very coarse fluorescent grains remained at the point source. The lag grains formed ripples with lengths three times those of the natural ripples. This suggests that ripple length is governed primarily by grain diameter and wind velocity.

Analysis of tracer distribution revealed that fluorescent sand entered and left the sample grid at a constant rate. A decay curve of tracer loss from the point source indicated extinction of the point source occurred 157 min. after release. Difference between decay value and per cent of the total tracer on the dune at any moment was equivalent to per cent of tracer loss from the sample grid at any moment. Knowing the distance of grain movement, this relationship yielded an average tracer grain velocity of 30.36 in./min. Time-lapse motion pictures established that creep velocity, assumed to be equivalent to ripple velocity, was 0.153 in./min., indicating that grains in saltation were traveling roughly 198 times faster than grains in creep. This magnitude of difference between creep and saltation velocity is physically inconsistent with Bagnold's classic division of eolian sand load.

End_of_Article - Last_Page 1762------------