12 The: Colorado Experiment Station. 
There is nothing in the results obtained by the ordinary agri¬ 
cultural analyses of this soil to indicate anything unusual unless 
it be the small amount of total nitrogen present. The method 
followed in these analyses was the conventional one, digestion with 
hydric chlorid sp. g. 1.115, etc., and the nitrogen determined by 
the Kjeldahl method modified for the determination of nitric acid. 
No. 785 is the first foot of soil from which the surface portion was 
scraped off in order to obtain the soil without incrustation of 
effloresced salts, and we have, as shown by the aqueous extract of 
this sample, a remarkably small amount of nitrates while the second 
and third foot show decidedly increased quantities. The third foot 
contains only a little less than the top two inches but is relatively 
only one-sixth as rich. We have for the top two inches 2,257 
pounds, for the next foot 528 pounds, for the second foot 1,332 
pounds and for the third foot 2,053 pounds per acre. I do not be¬ 
lieve that the nitre in the third foot of soil participates largely in 
doing injury to the trees unless it be brought nearer to the surface, 
say within two feet of the surface, by capillarity or some other 
agent, so that we cannot take this third foot into our reckoning 
except as a possible source of nitrates. The top twenty-six inches 
of this soil, however, contained at the time the samples were 
taken, in May, a little better than two tons of sodic nitrate, a 
comparatively moderate quantity, provided that these salts were 
not concentrated within the feeding zone of the roots. But, even 
as the case stands, there is, assuming 80 trees to the acre, fifty 
pounds of sodic nitrate to the tree, taking the soil to the depth of 
twenty-six inches and supposing that the succeeding foot with 
twenty-five pounds more for each tree does not in any way enter 
into the problem. I do not know how the amount of nitrates 
varied during the ensuing season, but, if they increased in this 
soil as they were observed to do in some others, the aggregate 
amount present during the season must have been at least twice as 
great as that given but, be this as it may, no trees occupying a 
continuous area about the point where these samples were taken 
died within the next four months in the same manner as that in 
which the trees we poisoned with sodic nitrate and as thousands 
of others, in lands richer in nitrates, died. The question of the 
amount of nitrates necessary to kill apple trees would be an in¬ 
teresting one to answer but it is not the object of this bulletin. 
There are also other questions involved, for instance, to determine 
the maximum quantity of nitrates available to trees in such lands 
at any one time, for, it is evident, that if a fatal quantity were 
available for only a few days, perhaps one day, it might suffice 
to kill them. We killed a tree in four days. Again it would be 
interesting to know whether a tree, say 15 years of age, would 
tolerate more saltpetre in September than in April, May or June, 
