Feb. i 9( 1917 Arsenical Injury through the Bark of Fruit Trees . 315 
These comparisons are worthy of further analysis. There is, on the 
whole, a rather striking similarity if we restrict our comparisons to in¬ 
juries on the thick, rough-barked crowns, trunks, and limbs, where the 
treatment was not too severe and the in3 ury was correspondingly slow. The 
streaking effects were produced as a result of very rapid absorption, either 
through wounds or thin places in the bark, and a very close similarity 
need not be expected. It would appear that up to a certain limit the 
slower the injury and the longer the killed bark remains in contact with 
the moist earth the darker and more friable it becomes. It is possible, 
therefore, that if the injury in the Colorado orchards under consideration 
has been slowly developing on old trees for a considerable number of 
years this would account for the difference. As for the small and highly 
colored fruit, the yellowish foliage, and the restricted terminal growth, 
it is well known to horticulturists that these conditions may follow a 
nearly complete girdling of the crowns from any cause whatsoever. The 
purplish color of the foliage noted in some cases (Table X) late in the 
season both on our experimental trees and in some of the Colorado 
orchards is not restricted to arsenical injury. At the time it appeared 
in our treated trees it could be found in untreated ones where small 
branches had been partly broken down or the bark scraped off by farm 
machinery. 
Realizing the difficulty of making comparisons of symptoms from 
written descriptions and the importance of such studies where the etiology 
of a disease is unknown, we made, in 1910, a rather extended trip into 
the affected districts of Colorado and Utah. On this trip we received 
most courteous consideration and many helpful suggestions from members 
of the Experiment Station staffs in these two States, especially from 
Headden and Ball. We found that the extent of injury suffered by the 
orchards was not overstated in Headden’s bulletins and that the symp¬ 
toms were most accurately described. 
One additional feature impressed us very much—namely, the amount 
of injury done the aerial portions of the trees relative to the size of the 
corroded area on the crown. To be sure, many trees were entirely 
girdled and others nearly so, but some trees 8 to 12 inches in diameter 
had a dead patch of bark on the crown no larger than a man's hand and 
yet showed in fruit, foliage, and terminal growth a degree of injury 
wholly unexpected if one judges by the trees, frequently encountered, 
that had the crowns more or less injured by mice, blight, or other agencies. 
We dug up a few of the sick trees completely and uncovered the crowns 
and upper roots of a great many others. We did not dig up all, as this 
would have been a herculean task which time did not permit. The only 
possibility that some of the trees recorded as having only a small dead 
patch on the crown could have had other injuries is that such injuries 
were on the deeper roots below the attachment to the crown. However, 
