4 BULLETIN 1253, U. S. DEPARTMENT OF AGRICULTURE. 
It is felt, therefore, that the analysis here presented furnishes a 
trustworthy basis for comparison as to disease in the cars inspected 
between varieties, States, months, years, and inspection offices; also 
between the box and the barrel crops. Throughout the tables per- 
centages are given to the first decimal, though it is evident that any 
figure beyond the decimal point is of doubtful value except where 
large numbers of cars are involved. 
It should be added that the word '"percentage" as used in this bul- 
letin means the proportion of diseased apples to the total quantity 
in the carload or storage lot and has no reference to the amount of rot 
or other disease on individual fruits. Thus, 10 per cent of blue-mold 
rot means that 10 per cent of the apples showed the rot in some 
degree, but not that 10 per cent of them were totally rotted. For 
commercial purposes, the first condition is practically as serious as 
the second. On the market, an apple even slightly rotted is a cull. 
But with diseases other than rots the situation is different; 20 to 30 
per cent of slight scald or scab usually means a smaller reduction in 
the selling price than 10 per cent of bad scald or scab. The same is 
true for bitter-rot, Jonathan spot, water-core, and a number of others. 
It is apparent, therefore, that the present analysis is incomplete and 
that any subsequent report should take into consideration both the 
percentage of diseased apples and the extent to which those apples 
are diseased. 
In a number of the tables that follow there are, aside from the first 
or " Classification" column and the second column for total number 
of cars showing disease, four others for rots, scald, other diseases, 
and a disease index. The percentages shown in the third, fourth, 
and fifth columns were obtained by adding together, for each column, 
all percentages of rot, scald, or other diseases, respectively, that were 
recorded on the certificates and dividing by the total number of 
cars showing disease, for each month, variety, State, or office. Such 
a procedure may seem to involve the pyramiding of percentages and 
would actually do so were it not for the fact that when' as many as 
three diseases occur in a car they are usually found to be a rot, 
scald, and some other disease. It does not often happen that a car 
shows more than one rot or more than one ''other disease" aside 
from scald. And even when two rots or two "other diseases" are 
recorded, it has usually been found possible to segregate percentages 
and avoid pyramiding in the summary tables. 
As for the sixth or index column, pyramiding is involved in every 
figure which appears in it. That is to say, the very fact of obtaining 
such a figure implies that each disease appears on a different lot of 
apples in the car, which is not true. In a car which shows 5 per 
cent of rot, 20 per cent of scald, and 10 per cent of some other disease, 
there is not necessarily 35 per cent of the carload affected with 
disease. There may be only 20 or 25 per cent of it so affected. 
The disease index is, therefore, not a true percentage, but merely a 
sum of percentages, and has value, if at all, only when used as a 
basis of comparison between years, months, varieties, or any other 
coordinate elements in the tabulation. 
In Tables 3, 4, 5, 7, 8, 9, 11, 12, 13, 14, 15, 16, and 18 percentages 
are calculated on the basis of all cars showing disease ol any kind. 
In Tables 1, 2, 6, 17, 19, 20, 21, 22, and 23 the percentages are cal- 
culated on the basis of cars showing some particular disease or diseases. 
