142 The Philippine Journal of Science 1922 
compare for themselves the costs of spraying in Japan with 
those in America; the disadvantage to American conditions, if 
it exists at all, is very slight. 
To summarize, then, a reduction from 80 to 96 per cent of 
cankered fruits on untreated plats, to 6.5 per cent on plat 3, 
cost 92.1 sen or 46 American cents per tree. This included very 
careful pruning of the trees. A reduction from 80 to 96 per 
cent to 34.5 per cent was obtained without pruning for 33.7 
sen or, roughly, 17 American cents per tree. Formalin 1 to 100 
solution, the least-effective spray, was incidentally by far the 
most expensive. The figures are presented in detail here and 
probably can be best interpreted for each country by those men in 
the industry closely acquainted with the markets. In America, 
however, with the highly developed competition for markets, 
this reduction of the blemish caused by canker would apparently 
be profitable at these costs. In oriental countries, where a 
blemish upon the fruit is not so important, possibly the expen- 
diture for control would not be advantageous. These statements 
apply only to that class of citrus fruits having the suscepti- 
bility of the Washington navel. 
It should be considered that there are factors which made these 
results less profitable than would be the case in ordinary years. 
First, the typhoons which occurred in early August are unusual 
in normal years. These typhoons were entirely unexpected, and 
the trees were entirely unprotected by preventive-spray coat- 
ings. One typhoon at this time would have been a considerable 
setback, but a second one is far from a normal seasonal occur- 
rence. Another factor has been pointed out previously ; namely, 
that continued spraying, year after year, would materially re- 
duce the sources of infection, so that canker control should mere 
nearly approach completeness with successive years. The results 
presented here, being those of the first year of control attempts, 
are not so favorable as they would be in a succeeding year. 
RELATION OF WIND PREVENTION TO CANKER CONTROL 
Another factor, that of protection against winds, could be 
improved upon in commercial work over that in the experiments. 
Although a row of a coniferous tree surrounded the orchard, 
such a windbreak was small and not sufficient to protect the 
whole orchard; a much more desirable condition could be pro- 
vided. The data available to show the definite value of wind- 
breaks follow: 
A row of loquat trees ran across the experimental orchard 
from east to west. The situation of these loquat trees is shown 
