388 BULLETIN 339, U. S. DEPARTMENT OF AGRICULTURE. 
show that ight but rather frequent applications would give good 
results with these soils and that the application of water in this 
manner would result in a great saving of water, an experiment was 
carried on that it is believed has determined these factors in a satis- 
factory Manner. — 
A water-tight galvanized-iron tank, 2 feet in diameter and 6 feet 
deep, with an outlet at the bottom terminating in a three-fourths inch 
galvanized-iron pipe 10 feet long, was installed in the soil flush with 
the surface of the ground. This tank was filled with 6 feet of soil in 
as near its original position as possible, three bands of hot asphaltum 
1 foot wide were painted around the inside of the tank at 1-foot inter- 
vals to prevent percolation between the soil and the inside of the tank, 
after which the tank was irrigated each time an adjoiming experi- 
mental plot was irrigated by applying the same amounts to the tank 
that were applied to the plot. Whatever water percolated through 
the 6 feet of soil m the tank drained out of the outlet pipe and was 
caught in a tub in a curbed pit 10 feet away. A total of 6.6 feet in 
depth in seven irrigations was applied to the experimental tract 
handled by the owner of the land and to the tank, and a quantity 
representing an equivalent of 5.5 feet per acre, or 83.5 per cent of the 
amount applied, percolated from the tank and was caught in the tub 
in the curbed pit, yet alfalfa which was planted on the tank when 
the experiment was started grew luxuriantly, demonstrating that the 
supply retained in the tank was adequate for the proper growth of the 
alfalfa. It was found that the tank during the season retained an 
average in its 6 feet of soil of 0.15 foot of water per acre for each 
application, and the experiment was continued during the following 
year, that of 1912, by irrigating the alfalfa in the tank whenever it 
appeared to require irrigation with 0.15 foot in depth at each apph- 
cation, no limit having been placed on the number of irrigations that 
were to be applied to the tank. It was found that the alfalfa in the 
tank required 10 irrigations during the season, and that there was 
but a trace of percolation from the tank. The alfalfa grown on the 
tank was cut, cured, and carefully weighed, and produced at the rate 
of 7.147 tons per acre, with an equivalent of 1.5 feet of water per acre 
plus a precipitation during the growing season of 8.51 inches. While 
this experiment was not conducted upon a scale broad enough to 
demonstrate the feasibility of this small amount of irrigation on large 
tracts or that this volume would produce the same comparative yield 
on them, it is believed, after taking into consideration the other 
observations, that it has been demonstrated that porous soils actually 
require but a very small supply at each application. 
The securing of similar results from a practical application of the 
experiment just described requires a careful adaptation of methods 
of applying water to the types of soil to be irrigated. If an irrigation 
