ERADICATING THE COMMON BARBERRY 37 
jng barberries, did not make so good an average record. The reason 
for this is evident. The field treatments with kerosene were not 
begun until late in the fall of 1922. By that time the character ol ! 
the root system of the barberry and the most effective point of ap- 
plication of chemicals had been determined. A rather definite idea 
as to the quantity of chemical necessary to kill barberries also had 
'been acquired. 
One feature in connection with the use of kerosene for killing 
barberries which is not shown in Table 14 is the speed of the action 
• of the chemical. In the greenhouse the experiments ran for about 
• one month, at the end of which time there was no apparent injury to 
the plants, yet kerosene is one of the most effective chemicals known 
for killing barberries. In the field, kerosene is equally slow in its 
action. For instance, of the 20 bushes treated on May 5 only 1 
showed more than slight injury on July 3, but by October all of 
•these bushes were dead. 
The bushes treated July 3, 1923, were examined on August 10, at 
which time one of the five receiving 1 gallon was recorded as dead. 
The other four and the five receiving only one-half gallon showed 
from slight to no injury. On September 20, of the five receiving 
1 gallon each, three were recorded as dead and two had still a few 
green leaves. The notes on the bushes receiving one-half gallon are 
as follows : i; One small bush looks dead. The other four have 
dropped most of their leaves but retain a few green ones. One of 
them shows new growth from some buds." However, on July 14, 
1924, all 10 of these bushes were dead. 
Another interesting feature was observed in several instances. The 
data for the treatments made October 17, 1922, will illustrate it. 
On that date 10 standing bushes were treated with one-half gallon of 
kerosene each. On May 4, 1923, they were examined and the follow- 
ing notes were made : " Three plants without leaves. The others 
:i»ll have from a few to many leaves. They are not as far advanced as 
untreated bushes." By July 3 all this new growth had died, and 
later examination showed that all the bushes were dead. 
These instances show that kerosene is very slow in its action on 
barberries. In some ways this is a serious disadvantage and makes 
iJais chemical less desirable to use for eradication purposes than a 
rapidly acting chemical such as salt. It also is evident that brief 
greenhouse tests can not be taken as an absolute criterion of the action 
of any chemical in the field. Greenhouse tests undoubtedly are of 
value at times as indicators, but too much reliance should not be 
placed on them. 
The results with kerosene were so satisfactory and residts obtained 
by Baringer in Ohio corroborated them so well, that in the spring 
of 1924 the use of this chemical for eradicating barberries was recom- 
mended. A circular (16) published by the United States Depart- 
ment of Agriculture gives in detail the method of using this sub- 
stance. As a result, several thousand barberry bushes have been 
^treated with kerosene with uniformly satisfactory results. 
It is interesting to compare the four petroleum oils studied — gaso- 
line, gas oil, kerosene, and fuel oil. All were very slow in their 
:action. Gasoline was perhaps a little more rapid than the others, 
but it was slightly less effective than either kerosene or gas oil. The 
