2 CIRCULAR 50, U. S. DEPARTMENT OF AGRICULTURE 
cultural Chemists.2, In many cases the records, however, were not 
definite about the analytical methods, and in others variations from 
the official procedure have been described. Such data have been 
included if the indications were that the results were comparable 
and of sufficient accuracy for the present purpose. 
Fruits are grouped in various ways, for the most part according 
to species. The Bureau of Plant Industry has given much helpful 
advice in this classification. The data on some fruits were not suf- 
ficient to permit of distinction within the genus. In the case of others, 
the data lent themselves to varietal, type, or geographic classification. 
Such groupings were used if it appeared that the differences between 
groups were significant. In still other cases such classes as red and 
black raspberries were reported separately, although it was doubtful 
whether any significance should be attached to the differences. 
DEFINITION OF TERMS 
REFUSE AND EDIBLE PORTION 
For each fruit the ‘‘refuse’”’ is defined according to the part rep- 
resented, so far as this could be determined from the records. The 
basis of selection of these figures was similar to that used by Atwater 
and Bryant.’ An effort was made to exclude data that represented 
any considerable loss from spoilage in addition to the normal refuse, 
but many of the records were not definite on this point. Percentages 
of refuse are expressed in terms of the fruit “‘as purchased.”’ 
All of the data on chemical constituents refer to the “edible 
portion.’”’ The values in the line designated as A. P. refer only to 
the constituents of the edible part, but are calculated as percentages 
of the ‘“‘as purchased”’ weight. 
The part of the fruit that was analyzed and here reported as 
edible can be inferred usually by the definition of refuse, but there 
are certain discrepancies on this point. In apples, for example, 
the chemical sample of the edible portion was not peeled in all cases. 
It included skins in some studies; from other reports it is not clear 
whether the skins were included or excluded; and in three cases the 
entire apple was analyzed. Enough data were available, however, 
to indicate that in these cases the variations were of minor importance. 
For some fruits such as crab apples and currants no data on refuse 
percentage were found, but this does not imply that the whole fruit - 
was analyzed ‘‘as purchased.’’ With a few unimportant exceptions 
no data have been included from samples in which parts that are 
usually discarded were analyzed. 
In fruits with skins that are sometimes considered edible and 
sometimes discarded as refuse, removal of the skins from the chemical 
sample would be expected to have a noticeable effect on the fiber 
content and possibly on the content of ether extract and ash. But 
the variations in these constituents in individuals are so wide that 
the difference due to presence or absence of skin is relatively small. 
The seeds of grapes were counted as refuse in some studies and as 
edible in others. Such a difference in basis would, presumably, have 
a marked effect particularly on the fiber content, but owing to a lack 
of sufficient original data, it was impossible to make a sharp dis- 
tinction on this point in selecting analyses for the present summary. 
2 ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS. OFFICIAL AND TENTATIVE METHODS OF ANALYSIS. 
COMPILED BY THE COMMITTEE ON EDITING METHODS OF ANALYSIS. REVISED TO JULY 1, 1924. Ed. 2, 535 
p., illus. Woastinetons D.C. 1925. 
3 ATWATER, W.O., and BRYANT, A. P. Op. cit. 
