108 
Research Bulletin No. 3 
localities on the loess, no difference was observed in the readiness 
with which roots penetrated the soils or in the extent to which 
the free water could be exhausted, the latter probably being de- 
pendent upon the former. The eastern soils, which gave up their 
free water least completely, were much less calcareous than the 
others. 
In all the cylinders a hard crust developed below the mulch. 
EXPERIMENTS WITH PERENNIAL DESERT LEGUMES. 
Three species were employed more or less, viz, Prosopis velu- 
tina (Mesquite), Acacia Greggi (Catsclaw), and Acacia con- 
stricta. The seeds of the first two were gathered near Benson, 
Arizona, and those of the last near Columbus. New Mexico. Both 
places are in desert regions, the normal annual precipitation be- 
ing a little under 10 inches. 
The seeds were planted on December 31, 1910, in small flower- 
pots filled with surface soil .from the Experiment Station farm. 
Later, part were transferred to larger flowerpots filled with the 
same soil. Until they were transplanted to the metal cylinder and 
left to die all were watered freely. 
Experiment 1. — In this there were used two cylinders of gal- 
vanized iron 6 inches in diameter and 1 foot high, in the bottom 
of each of which ten small holes had been punched. In the bottom 
of each a layer of coarse gravel f of an inch in thickness was 
placed and then H O subsoil tamped in to within 3 inches of 
the top. On February 4, 1911, one small mesquite plant (Pro- 
sopis velutina) was transplanted, soil and all, to each cylinder. 
The space between the black soil and the cylinder was filled 
with H O subsoil. The surface was covered with a one-half 
inch mulch of coarse sand, the top of this being one-half inch 
below the top of the cylinder. Then the soil was thoroly satu- 
rated and the cylinders left in the greenhouse without further 
addition of water and without any protection from the sun's 
rays. 
The mesquites when transplanted were 4 inches high and 
bore 7 and 9 compound leaves, respectively. On March 18 the 
plants were both 10.5 inches high and bore 15 and 14 compound 
leaves, respectively; even on this date some of the lowest leaves 
had fallen from each. On April 11 the tip of the plant in cylin- 
der No. 3 was dead and only 6 compound leaves remained, while 
the tip of the plant in cylinder No. 4 was still alive altho the 
number of compound leaves on it also was only 6. Nine days 
later the numbers of leaves were 4 and 6 respectively. On May 7, 
just before No. 3 was opened, when the plants were photo- 
