26 
Colorado Experiment Station 
per million, and to cause its destruction. This sample of disintegrated 
bark may have contained some arsenic in the form of the spray original¬ 
ly used. We tried to remove all the soil and adhering spray by sift¬ 
ing, this being the only feasible means of removing them. 
The Source of the Arsenic. 
We have found arsenic present in every sprayed tree which we have 
examined. I am fully convinced that every sprayed tree contains 
arsenic enough to permit of its detection in the woody tissue without 
the use of any very large amount of the wood. I believe that from two 
to 4 or at most 6 ounces of wood will suffice for the detection of 
arsenic. The perplexing feature in this work has not been to detect 
the presence of arsenic but to arrive at a definite conclusion in regard 
to its source and action. I have already answered both of these ques¬ 
tions. 
The presence of arsenic alone is not taken as sufficient proof that 
the arsenic in the trees came from spray material, though such might be 
the case, but when arsenic and copper or arsenic and lead or when all 
three are present at the same time I believe it to be sufficient evidence 
that the tree has taken it up mostly, if not wholly, from the sprays 
which have been used. If there is arsenic already in the soil the tree 
might without doubt obtain a trace from this source. This question, 
however, will be more fully considered under the subject of systemic 
poisoning, where it properly belongs. 
The Cause of Death. 
The crowns of trees growing in our neglected orchards are not 
corroded although the trees mav be in bad condition. The trees that 
have been killed by nitre have perfect crowns. We have described 
the condition of the bark on trees showing cases of corrosion. We 
have shown that arsenic destroys the bark in this manner and have 
cited at least eight instances in which I have seen the trees. We have 
shown that this bark contains in the case of the pear trees arsenic 
equivalent to 24.71 part - ’ of arsenic acid per million of the bark. This 
observation is supported by the findings of Mr. Watson, who obtained 
15.8 parts of arsenic per million, which, assuming that Mr. Watson’s 
figures mean metallic arsenic as I suppose they do, would correspond to 
24.22 parts arsenic acid per million. These figures are so high that 
any other source than the spray material is out of the question; besides 
we know that the spray material doe^ run down the trunk and into the 
ground at the crown of the tree. That this is a sufficient amount of 
arsenic to destroy both bark and woody ti sue is plainly shown by the 
amount of arsenic found in the disintegrated tissues of roots which 
we know were killed by arsenic, in which was found arsenic correspond¬ 
ing to 24.02 parts arsenic acid per million. The maximum amount of 
arsenic found in any part of a tree, even when the arsenic was known 
to have been applied in such quantities as to have killed the tree quick- 
