reference to the yellow and red pigmented plant foods. Hence, many 

 of the values in these tables should be considered tentative. 



Most of the ascorbic acid present in fresh foods occurs in the reduced 

 form, and it is this form that is represented by most of the values in 

 the literature. To the extent that some foods, particularly after 

 storage or processing, may contain also small amounts of ascorbic acid 

 in the oxidized form (dehydroascorbic acid), values for the reduced 

 form may underestimate the total vitamin C content. Where authors 

 reported both the reduced and the total ascorbic acid, values for the 

 total were used. On the whole, however, averages in these tables are 

 based on values for the reduced form. 



Problems in Derivation of Averages 



The objective in developing these tables was to derive values repre- 

 sentative of foods as they are used throughout the year, in the country 

 as a whole. However, it is not possible with present knowledge to give 

 averages that are equally satisfactory for all foods or for all nutrients 

 in a given food. For one reason, it is a rare case when the same sample 

 of food is analyzed for all nutrients listed in these tables. One inves- 

 tigation may be concerned with protein, fat, and carbohydrate, another 

 with minerals, and still another with one or more vitamins. There 

 seems to be much more interest in studying the content of a food with 

 respect to one vitamin than another, hence the great inequality in the 

 volume of data available for various nutrients. In the case of tomatoes, 

 for example, averages in these tables were based on 956 analyses for 

 ascorbic acid, 73 for thiamine, and 46 for vitamin A. 



Problems involved in the compilation and summarization of food 

 composition data are numerous and difficult. One must reckon not 

 only with natural variability of foods as affected by variety, cultural 

 factors, and maturity, but in the case of vitamins, with the effects of 

 handling foods after harvesting. As a result, considerable judgment 

 must be exercised in the selection of data to be averaged and in de- 

 ciding whether simple averages or weighted averages are most suitable 

 for obtaining representative values. 



Sometimes it may seem best to exclude experimental strains of fruits 

 or vegetables not in common production, especially if their inclusion 

 might greatly alter the average. Again it might be argued that ascor- 

 bic acid determinations made within 2 or 3 hours after harvesting a 

 fruit or vegetable should be excluded on the grounds that store-bought 

 products are never as fresh as this. However, in these days of victory 

 gardens it seemed logical to include such data provided they did not 

 unduly weight the averages. Many times these problems are compli- 

 cated by insufficient information about the foods which were analyzed 

 so that there is little basis for making a decision. 



In some cases production figures were studied as a basis for weighting 

 values for different varieties or types of foods. For instance, ascorbic 

 acid values for early market varieties of cabbage differ from those for 

 the late "storage type," therefore an attempt was made to learn the 

 relative quantities of each that are ordinarily grown in the United 

 States, in order that a representative average might be obtained. 



