The: Fixation oe Nitrogen. 43 
trees at this time, May 1, 1911, were burning quite badly though 
the leaves were not yet fully developed. It is unusual to observe 
this burning of the leaves in any degree so early in the season and 
very much more so to meet with such marked cases of it as these 
trees presented. Unfortunately I do not know how high the water 
plane rose during the season of 1910, but information elicited by 
specific inquiry in regard to this point did not suggest an undue 
amount of water as the cause of the trouble or as the probable source 
of the nitrates. I visited this orchard again on May 30, 1911, at 
which time the destruction of the orchard not only seemed a cer¬ 
tainty but was already far advanced. The following analyses rep¬ 
resent the surface soil and the water from a well located near by 
a burning tree. This sample of water was sent by the owner. 
ANALYSES 
LIV 
LV 
• 
Water-Soluble 
Water Residue 
Laboratory 
Laboratory 
No. 1073 
No. 1200 
March 28, 1911 
May 12, 1911 
Percent 
Percent 
Cialcic sulfate . 
. 24.275 
36.596 
Calcic chloric! . 
. 8.658 
Magnesic sulfate . 
2.194 
Magnetic chlorid . 
. 8.577 
11.654 
Potassic chlorid. 
. 1.951 
2.107 
Sodic carbonate . 
5.040 
Sodic chlorid . 
. 39.187 
41.992 
Sodic nitrate . 
. 17.075 
Minute trace 
Iron and Aluminic oxid . 
. 0.088 
0.078 
Silicic acid 
0.189 
100.000 
100.000 
Sample No. 1073 represents the surface soil, an inch and a 
half or at most two inches deep, taken beneath a tree which was 
already burning. This sample yielded 5.768 percent of salts to 
water, of which a little over 17.00 percent, 17.075, was nitrates, 
giving us 6,500 pounds, three and a quarter tons, of nitrates in 
the top two inches of this soil per acre. Sample No. 1200 is the 
ground water taken from a well near to another burning tree which 
has since died. This water contained 10,094 p. p. m. of total 
solids, but contained so little nitric acid that five grams of the 
residue gave with hydric and ferrous chlorids only a minute trace 
of nitric oxid. This test, like all of the nitric acid determinations, 
was done in duplicate. It is a striking fact that this ground water 
within four and a half feet of the surface on which an abundance of 
nitrates occur should be so nearly destitute of even a trace of nitric 
acid, but this is the second case; we have already given one case 
in which this is so and we have still another which will indicate 
the same thing. In the case of sample 1201, a drain water, we 
have given even fuller data, showing the presence of 8,489 parts 
of salts per million with only 0.018 p. p. m. of ammonia, 0.00003 
