By August, 1911, rows 1, 2, and 3 had lost 16 trees, while rows 4, 
5, and 6 had lost only one. This indicated that a more even stand 
might be obtained by leaving the trees in the nursery until the 
second summer after planting than by setting out small seedlings 
with five to six pairs of leaves. In September, 1912, row 4 was the 
only one which had lost no trees. 
In August, 1911, the average height per tree for row 4 was 7.3 
inches greater than that of the next tallest row and 10.9 inches 
greater than the average height of the trees of row 6, set at the 
same time but with the roots bare of earth. When considered at 
this time, in lines transverse to the rows and to the slope of the 
hill, 43 of the 56 living trees of rows 1, 3, and 6 were shorter than the 
correspondingly placed trees of row 4. Measurements made in Sep- 
tember, 1912, showed the average height per tree of row 4 to be 13.6 
inches greater than the average height per tree for row 6. 
In 1911 four to six trees in each of rows 1 to 4 produced some 
coffee, but no other trees in the planting gave any crop. This indi- 
cated that early production can be secured either by setting quite 
young plants or by transplanting older ones with a ball of earth 
incasing the roots. While the average yield per tree in 1912 for 
rows 1 and 3 exceeded that of row 4. the total yield of row 4 was the 
greatest of any row and doubled that of row 6. 
In transplanting trees about 20 months old, as in this instance, 
better development and earlier yield can be secured by moving the 
plants with the roots incased in a ball of earth. Transplanting seed- 
lings a year younger than these necessitates caring for the plantation 
for an additional year, a much more costly proceeding than keeping 
the nursery in condition. Also, the very small trees, less able to cope 
with adverse conditions, are more apt to be lost, leaving spaces to 
be refilled. 
The slight growth and meager crop of rows 7, 8, and 9 show these 
trees to have been retarded in their development a full year by the 
cutting back of the young trees to 4 to 6 inch stumps on trans- 
planting. 
PLANTING NO. 2. 
As a check on rows 4 and 6 of the preceding test, each of two 
small beds in a location shown by the growth of near-by coffee trees 
to be well adapted to this crop was set with 10 trees of Arabian coffee 
from seed planted a little more than 21 months before. In each bed 
five trees were set with roots in a clod of soil from the nursery and 
five with roots bare of earth. As the trees were to be left only a 
short time, a close planting with consequent uniform conditions was 
possible. 
The combined height at the time of setting of the 10 trees trans- 
planted with roots incased in soil was 306J inches, and that of the 
