3 °° 
Journal of Agricultural Research 
Vol. VIII, No. 8 
inches. Under the bandage the branch was girdled by at least one of 
the duplicate treatments. Zinc arsenite (California Spray Co.) caused no 
injury whatsoever. The amount of injury caused by the other chem¬ 
icals was intermediate, consisting of lenticel spots that were anastomosing 
more or less. 
Twelve other unwounded branches were treated in duplicate with the 
following chemicals, io gm. each: 
Calcium arsenite, Baker. 
Iron arsenate, Merck. 
Copper aceto arsenite, Baker, 
Zinc arsenite, ortho, Baker. 
Zinc arsenite, pyro, Baker. 
Zinc arsenite, meta, Baker. 
In this series calcium arsenite entirely killed the branch above the 
bandage. Streaks ran down from the bandages 16 and 20 inches, respec¬ 
tively. Copper aceto arsenite and the three zinc arsenites caused only 
lenticel spots, which in some cases were anastomosing. Iron arsenate 
caused no injury. 
It is to be noted that in this last series 10-gm. quantities of calcium 
arsenite and of copper aceto arsenite caused more injury than did 60 
gm. in the series just described. The 60-gm. series was begun on July 
30, 1911, and the 10-gm. series on June 17, 1912, both in midsummer. 
Had the injury in these two series been approximately equal, it would 
not have been especially remarkable, as we have found that, as a rule, 
the injury from these slightly soluble arsenicals is by no means propor¬ 
tional to the amounts used. Erratic results like these have made us 
very careful about drawing conclusions concerning the exact amounts 
required to produce injury, and have prompted us to make comparisons 
only between treatments given on the same date under conditions as 
nearly alike as possible. 
We may at least conclude that even unwounded limbs may be severely 
injured or ultimately killed as a result of girdling by some of the chemi¬ 
cals having only slight solubility. It appears, however, thaj tests on 
wounded limbs, where the exposure of living tissues to the chemicals is 
assured, are somewhat more suitable for comparing the toxicity of the 
chemicals than those on unwounded limbs where the corky bark limits 
their free access. 
Tables IV and V record the results of work specifically intended to de¬ 
termine the relative injuriousness of the more common insoluble arsenical 
compounds when applied to branches that are wounded. A close study 
of these tables shows some points of interest. One of the most striking 
facts is that similar tests, made under conditions as nearly alike as pos¬ 
sible, do not always yield the same result. Briefly stated, duplicate 
tests do not always check with each other, though in most cases they do 
so fairly well. In other words, there are factors that we can not control, 
and do not even know, that sometimes influence to a marked degree the 
