EXPERIMENTS WITH COCONUT PALMS 
17 
»• 
taken into consideration, a comparison of production in the fifth 
period with that of the fourth period would appear to be the best 
criterion for judging the effect of salt on production. Plats Nos. 1 
and 7, which received salt at the rate of 2y 2 pounds per palm per 
application, yielded in the fifth period 
at the average rate of 42.2 nuts per palm, 
whereas in the fourth period the rate 
was 46 nuts per palm, the yield in the 
fifth period being only 92 per cent as 
great as that in the fourth period. Simi- 
larly plats Nos. 3 and 6, receiving salt at 
the rate of 5 pounds per palm per applica- 
tion, produced in the fifth period 33.2 nuts 
per palm in contrast with 37.9 nuts in the 
fourth period, the fifth-period yield being 
88 per cent as great as that of the fourth 
period. The average yield of plats Nos. 2, 
4, and 5, which received no salt, in the fifth 
period was 36.5 nuts per palm, which was 
91 per cent as great as that of the fourth- 
period yield of 40.3 nuts per palm. Under 
the three treatments the production fell in 
the fifth period to 92, 88, and 91 per cent, 
respectively, of what it had been in the 
period preceding, thus in the uniform de- 
cline failing to indicate any effect of salt 
on yield. If comparisons are made of ac- 
tual numbers of nuts produced, it is seen 
that the check stood between the treated 
plats in the fifth period and that the groups remained in the same 
sequence as before receiving salt. 
The production of 287 palms recorded throughout the five periods 
is shown graphically in Figure 10. The variation in individual 
productivity was very great. 
Section 
C 
Figure 9. — Average number of 
nuts produced per palm per 
subdivided plat in 16 har- 
vests, July, 1913, to June, 
1917, at Boquillas plantation 
Figure 10. — Variation in total nut production per palm in 40 harvests at Boquillas 
plantation. Each line represents the nuts produced by one palm 
Figure 11 contrasts the yield by periods of the 50 most productive 
palms with that of the 50 least productive palms for the entire time 
covered by the record and shows that a group of palms which were 
low or high yielders in one period were respectively low or high 
yielders in each of the other periods. Whereas the average produc- 
