Nitrogen Fixation 25 
The last sample naturally suggested the examination of other 
waters issuing from the shales underlying irrigated mesas. The 
same point had been reached, starting from an entirely different 
point of departure, i. e., the search for an answer to the question, 
how much alkali may an irrigation water contain without detriment 
to the land to which it is applied? No attempt is made to answer 
this question at this time; it suffices that no one has yet complained 
with good cause that he has reached the injurious limit. An in¬ 
spection of the next two analyses will indicate clearly why'they are 
given in this place. The land was planted to corn and rye. It had 
passed into the hands of a man who had no experience with irriga¬ 
tion; the water was handy and plentiful, the people on the mesa 
above had only a moderate supply of water or there would have 
been more. The corn had been rather deeply furrowed for the 
purposes of irrigation. There was a band of efflorescent alkalis at 
least two inches wide on the side of each furrow. My curiosity 
was excited to find some reason why the corn should grow at all 
and not die. The rye was dying, but the corn was in pretty fair 
condition, which was a matter of surprise. Inquiry elicited the fact 
that small stuff had been a failure on this ground. I took a sample 
of the worst of this surface soil where even the corn, which seemed 
to have endured better than anything else, had almost failed. The 
irrigation was very excessive. The sample of soil yielded 4.67 per 
cent of water soluble, which gave me the following analysis: The 
determination of NO was made four times, because at the end of 
Hie reaction there was a rather copious evolution of C 0 2 , more than 
the sodic hydrate in the measuring tube could readily absorb. 
ANALYSIS XXII. 
WATER SOLUBLE PORTION. Laboratory No. 589. 
Per Cent. 
Calcic Sulfate . 16.645 
Magnesic Sulfate . 14.518 
Potassic Sulfate . 2.3 60 
Sodic Sulfate . 35.731 
Sodic Chlorid . 19.914 
Sodic Nitrate . 7.352 
Sodic Silicate . 0.328 
Loss (water, organic matter, etc.) . 3.152 
100.000 
The nitrate according to this analysis is 0.342 per cent in this 
portion of the soil. This can only be considered as an approximate 
result. The big fact is that nitrates were present in notable quan¬ 
tities. This land was irrigated with the seepage water which issued 
from or above the shale underlying the mesa. This water con¬ 
tained 418.2 grains per imperial gallon or 5,975.7 parts per million. 
No nitrites were present. 
