EXPERIMENTS WITH COCONUT PALMS 
least productive palm, No. 33. The removal of these two palms 
from their respective plats would have very materially altered the 
plat averages, and as a result, plats Xos. 1 and 2 would have inter- 
changed positions in rank, and plat Xo. 3 would have been moved 
from sixth to third place. These changes, however, would not have 
simplified greatly the explanation of production in relation to fer- 
tilization, with two incompletely fertilized plats outranking the two 
completely fertilized 
:s, and the 
outranking 
salted 
all 
^5 
plat 
plat 
others 
Figure 2 shows the 
curves of individual 
annual production for 
the period 1920-1927, 
and of the 10 most and 
the 10 least productive 
palms. At no time did 
the curves of produc- 
tion of the palms of the 
two great groups cross. 
Only 3 of 76 individual 
annual harvests from 
the high- production 
group fell below the 
average of the field 
for the respective year, 
and in no instance did 
a palm from the low- 
production group attain 
average production. 
Seasonal influence pro- 
duced variation in yield 
between one year and 
another. But in this 
variation the tendency 
toward high or toward 
low production re- 
mained clearly evident 
and consistent. High 
yielders remained 
high yielders and low 
yielders remained low 
yielders. The total pro- 
duction, including that of the half year 1928, of the one group 
was 7,242 nuts, and that of the other was 2.415 nuts, which was 
a ratio of 3 to 1. Still another difference between the groups was the 
earlier maturity of the more productive palms. Six of the ten high- 
producing palms yielded nuts in 1920, and the other 4 did so the 
following year, whereas none of the 10 low-producing palms fruited 
in 1920, and 3 failed to fruit prior to 1922. 
For further comparison of high and low producers, two larger 
groups were considered, consisting, respectively, of the 25 palms 
Figure 1. — Nut production by individual palms, San 
Jose plantation, 1920—1928. Relative position as in 
the field. To find the palm number, combine the ten's 
place and the unit's place numbers. Thus, the palm 
in the upper left corner is No. 10, that in the upper 
right corner No. 19, that in the lower left corner is 
No. 90, that in the lower right corner No. 99, while 
the missing palm indicated by a blank is No. 28 
