Nitrogen Fixation 43 
were attacked. A number of cottonwoods were affected in this 
way. As elsewhere stated the condition of the soil in regard to 
excessive water was immediately investigated but we could not 
find any reason for attributing the trouble to this cause. One case 
in particular was instructive, it was in the remnant of the orchard 
designated Orchard No. 2. See Plates II, VI and VII. Plate VII 
gives a good idea of the size of this barren area as well as the killing 
of the remaining trees. The trees in the background are of the 
same age as those in the foreground. The dead tree in the middle 
of the background is the same tree that appears in the foreground 
of Plate II. 
There are a few trees on the east side of the orchard which 
had up to that time remained in very good condition, in fact, were 
quite thrifty trees. These trees are shown in the background of 
Plate VI. These trees died suddenly. Plate VIII shows a tree killed 
late in the summer. The crowns and roots of these trees were ap¬ 
parently healthy. The roots were followed for six or eight feet 
from the trunk of the tree, nothing could be found to indicate any 
cause for the deportment of the tree. I had known for three years 
that this ground was very rich in nitrates and locally in chlorids. 
This determined me to make some experiments with nitre and salt. 
The nitre in this soil amounts to many tons per acre foot, 56.7 tons 
in the surface foot. It was evident that I had to deal with large 
quantities so I applied nitre to some young trees in an experi¬ 
mental orchard and irrigated them to bring the nitre solution in 
contact with the feeding roots. The amounts applied were from 5 
to 20 pounds per tree. The results were that in those trees, about 
which I buried five pounds of nitre each, some of the limbs were 
reached and others were not, in other words in applying the nitre, 
(Chile saltpetre was used in this experiment), I had reached some 
of the feeding roots but the others had escaped. With the applica¬ 
tion of 20 pounds I succeeded in reaching all of the roots, at least 
I killed the tree in about four days. Judging from the deportment 
of these trees, I can conceive of ones failing in an experiment of 
this sort, even though he applied large quantities of nitre or other 
poison provided the poison was not put within the feeding area of 
the roots for the effects produced on different trees was by no 
means proportional to the amounts applied to the soil, the applica¬ 
tion of five pounds produced just as pronounced results as the appli¬ 
cation of 10 pounds. This of course depended upon the root sys¬ 
tem of the tree and the location of the nitre. The effects were in 
all respects similar to those produced in the other orchards, the be¬ 
ginning and progress of the effects, the killing of the leaves, the de¬ 
portment of the tree in throwing out a few whitish-yellow leaves and 
the appearance of the bark and wood after death, were identical. An 
