12 BULLETIN 3 4, PORTO EICO EXPERIMENT STATION 
palms in the one group was not due to high production for a brief 
time. In each of the three periods they gave a high yield. As a 
whole, their production varied little from one period to another. 
Of the 25 palms in the high-yielding group, 8 received no ferti- 
lizer and 17 received fertilizer. Of the 8 unfertilized palms, 5 gave 
a higher production per picking in the first than in either the second 
or the third period, and 3 gave a higher production per picking 
in both the second and the third periods than in the first period. Of 
the 17 fertilized palms of this group, 5 produced more per picking 
in the first period than in either the second or the third period, and 
8 produced more in both the second and third periods than in the 
first period. The remaining 4 palms showed in the second period 
an opposite trend from that shown in the third period when com- 
pared with the first period. If the second and third periods are 
considered as a unit, the 17 high-yielding fertilized palms will be 
found to have averaged in this period 11.7 nuts per palm per picking, 
whereas in the first period the yield was 11 nuts per palm per 
picking, the average interval between pickings being the same, 2.7 
months. These high-yielding palms thus failed to show any pro- 
nounced response to fertilization. 
Twenty-one of the twenty-five lowest yielders were fertilized. 
Nineteen of the fertilized palms yielded more in the third than in the 
first period. The increase in yield under fertilization was pro- 
nounced. Notwithstanding the increased production, the low-yield- 
ing palms maintained their relative positions as low yielders in 
comparison with the other palms. When ranked in order of in- 
creasing yield, only 1 of the 25 palms got beyond thirty-ninth place 
in either the first or the last period, a fact which shows that these 
palms were consistently low yielders. The highest rank held in the 
first period by any palm of the 13 lowest yielders was twenty-first 
place, and in the last period only 2 of the 13 ranked higher than 
twentieth place. 
BOQUILLAS PLANTATION 
In 1912 a fertilizer experiment was begun in the Boquillas coco- 
.nut plantation near the coast southwest of Ahasco. (PL 1.) Both 
as to palms and soil the area chosen was typical of the average 
plantation of old palms. The palms were in fairly good alinement 
and spaced about 27% feet apart. They were average to tall in height, 
and evidently had been in production for many years. The soil was 
slightly rolling sandy loam. The water in a small spring about 6 
feet below the surface of the ground between the experimental plats 
and the ocean was said to be salty in flavor. Analysis of the water 
by the station chemist showed that it contained practically no salt 
(sodium chloride), but was heavily charged with calcium bicarbon- 
ate. Its chlorine content was less than 0.05 part per million. 
TREATMENT 
For the work with fertilizers seven parallel plats were set aside , 
each containing two rows of 24 palms, excepting palms missing or 
excluded, with guard rows between plats. During the course of the 
experiment several trees were eliminated on account of bud rot or 
other cause. Hence, of the 303 palms included in the first period, 
