14 BULLETIN 1018, U. S. DEPARTMENT OF AGRICULTURE. 
it is remembered that Lloyd's observations were made under condi- 
tions where the soil-moisture content was increased only by rainfall, 
and this usually in such small quantities as to affect only the super- 
ficial layers of soil. 
It will be observed from the curves of shedding (displaced) and of 
moisture content of the deeper layers of the soil as shown in figure 
2 that there are indications of a correlation. Shedding began to in- 
crease coincident with the date (July 2) when the first important- 
drop took place in the curve representing the moisture content of 
the fourth foot. Several years of experience in studying the sub- 
terranean conditions influencing the growth of the cotton plant have 
convinced the writer that the depth of the taproots and their rate 
of progress downward can be traced fairly accurately by means of 
frequent determinations of soil moisture. It seems evident from the 
curves that the roots had not reached the fourth-foot layer of soil 
in sufficient number to reduce the moisture content to any extent prior 
to July 2. Subsequent to this date, and especially following the re- 
duction in moisture content of the fifth and sixth foot layers during 
this week, July 8 to 15, there was a rapid increase in the shedding 
rate. The increase was temporarily checked by the irrigation of 
July 28, which increased the moisture content of the fourth foot. 
Another node of high shedding rate followed soon after and continued 
until checked, evidently by the effect of the irrigation on August 21. 
The general trend of increasing shedding as the season advanced 
was finally checked in early September by the gradual cessation of 
plant growth. 
That the depletion of moisture in the deeper soil is responsible for 
the increased shedding rate may be supported by other evidence noted 
in a different part of the Salt River Valley, where the water table has 
risen to within 5 or 6 feet of the surface, because of leaky 7 canals, etc., 
and the soil is of a gravelly loam nature permitting the free move- 
ment of water. In this locality it was observed that cotton plants, 
although of large size, widely spaced, and having considerable 
transpiring surface, shed but few of their small bolls during the 
period when the shedding rate was highest in localities where the 
water table was not within reach of the cotton roots. In 1919 yields 
of* a bale to the acre were not uncommon in this district, and one 
7-acre field yielded at the rate of 1.7 bales of lint per acre. In other 
localities where the water table is still nearer the surface and where 
the soil has a greater capacity to retain moisture, it often happens 
that the plants grow to a height of 7 or 8 feet, and here the shedding 
is excessive as the season advances. This condition undoubtedly 
limits the roots which are functioning properly to a zone near the 
surface, and the strain imposed upon these roots by the great rate of 
transpiration apparently is a factor in the magnitude of the shedding 
rate. 
