52 FUMIGATION FOR THE CITRUS WHITE FLY. 
injury to orange trees and apparently never to tangerine and grape- 
fruit trees. The liability of injuring trees through the emptying of the 
contents of the jars after fumigation close to or upon the base of the 
trees will be referred to under the subject of precautions. The injury 
to orange trees from the gas itself has never in the writer's experience 
been sufficient to offset the benefits of destroying the white fly and 
scale insect pests. Nevertheless the subject is one of considerable 
importance. The experiments conducted in January and February, 
1907, demonstrated the practicability of destroying the white fly 
with hydrocyanic-acid gas without injury to citrus trees. 
The fumigation of nearly 4,000 trees in the winter of 1907-8 has 
greatly extended our knowledge of the effect of fumigation upon the 
trees, but there remain several unsolved problems in this connection 
which it is hoped will be elucidated by future experience. The work 
of fumigating a grove should be completed if possible before the new 
growth appears in the spring. Under certain temperature conditions 
successful fumigations may occasion no injury to new growth, but 
there is danger of destroying the first spring shoots which normally 
produce the greater part of the blooms. When affected by the gas 
new shoots wilt and turn dark, appearing as though affected by frost. 
Under certain conditions there is more or less shedding of the old 
leaves following fumigation. The loss of 10 or 15 per cent of the old 
foliage can not be considered an injury, inasmuch as even more than 
this proportion is usually shed during the winter or in the spring. In 
fact, it has been demonstrated b} r experiments conducted by Mr. 
Yothers and the writer in February, 1908, that the leaves shed by 
fumigation when the percentage of the whole does not exceed 15 per 
cent are among the leaves which would normally drop in the course 
of a few w r eeks. 
In the experiments with the sheet tent of 8-ounce duck summarized 
in Table IV, the most extensive shedding occurred in experiments 
40.14. In this it was estimated that about 50 per cent of the leaves 
were shed. The tree was fumigated on January 29, beginning at 
4.07 p. m., about one-half hour before sunset. No shedding was 
observed until the morning of February 2, when it was estimated that 
from 15 to 20 per cent of the leaves dropped. On February 4 it was 
estimated that 50 per cent of the leaves had fallen, after which date 
the amount of the shedding was inappreciable. The winged petioles 
of the leaves remained attached to the tree in most cases and the 
fallen leaf blades showed distinct brownish areas due to burning by 
the gas. The tree consisted of five steins growing from the roots of a 
tree frozen to the ground in 1895. One of these stems was affected 
by foot rot or mal-di-gomma , and the defoliation of this was nearly 
complete, materially increasing the percentage of shedding from the 
