80 BULLETIN 1301, U. S. DEPARTMENT OF AGRICULTURE 
called coulee orchards. The hillside orchard occupies a north slope 
of a morainal hill on one of the highest and most exposed sites at tee 
field station. Heavier soil is found here than in any other of the 
station orchards. The contour system of planting is used, the trees 
being placed in the center of plowed strips 20 feet wide which alter- 
nate with sod strips 30 feet wide. It was thought that excess rainfall 
would run off the sod strips and accumulate in the plowed land, thus 
increasing the water supplied to the trees. Seven rows of apples and 
crab apples were planted in 1915 and 1916. 
In general, treesin the hillside orchard have given poorer results than 
trees of the same variety planted in the orchards with more protection. 
Most of the trees are now dead, and nearly all the others lack thrift 
and show the effects of the severe exposure. Very little fruit has been 
produced, and the orchard is practically a faijure. 
The coulee orchards were both planted in the spring of 1920, the 
trees for the most part being common varieties of apples and plums, 
and the purpose being to note the growth and productivity of the 
trees in the more or less protected coulees compared to trees of the 
same variety growing on the more exposed bench land. Growth in 
the coulees has been quite vigorous, and some of the sand cherry— 
plum hybrids have already produced excellent fruit. However, it is 
too early to draw conclusions as to the value of coulee planting. At _ 
present it seems to be very promising. 
In this connection, a Wealthy seedling orchard planted in 1916 
may be mentioned. A small coulee passes guek the center of 
this orchard, and the stands in this coulee are good, whereas winter- 
killing has been very severe on the higher ground. ‘The trees in the 
coulee are also larger and growing more vigorously. 
FRUIT BREEDING 
Variety testing and cultura] experiments have their places in the 
development of fruit growing on the northern plains, but it is doubt- 
ful whether either line of work is of such fundamental importance as 
fruit breeding. Hardier varieties of all kinds of fruits are needed to 
build up a more successful northern Great Plains pomology. Much 
B to be looked for in the improvement of some of the hardy native 
ruits. 
The following three methods of fruit improvement usually have 
been attempted at the field station: (1) Selection from seedlings, where 
only one or neither of the parents is known; (2) crossing by hand or 
with bees in the field; and (3) crossing under glass by hand. 
The first method is the one most extensively used. A large 
number of seedlings of the different wild fruits are grown for selection 
purposes, and to study the tendency of the species to vary and break 
up into different types when grown under cultivation. This method 
has also been extensively used in growing seedlings of standard 
varieties of fruits, especially in cases where there seemed to be good 
opportunity for extensive natural crossing between desirable varieties. 
Some promising selections have been made from such seedlings. 
The second method, crossing by hand or with bees in the field, has 
been attempted in a limited way. It was planned to grow in groups 
two varieties which were to be crossed, two trees of each variety in a 
group. ‘The groups were to be covered with netting at blooming 
time and bees introduced to perform the crossing. Thus, all fruits 
