Nov.--Dec To +6) Please circle the months that pandanus is in season. Jan.-7-Feb.--March-- April-May----~ June--- July--Aug. 777 Sept.-Oct .--Nov.--Dec.-7~- The feasibility of obtaining a total profile of a typical diet from an incerview stems from the prevailing environmental conditions in which the variety »£ available foods is quite restricted. There is also a very limited trading -conomy - both the variety and availability of imported foods being restricted ry the limited capital of those who import and retail such goods. Thus the lim.ted availability of cash affects both the variety of traditional foods and the amount of contemporary imports as well. Thus, the typical diet is very "day to day". This makes it possible to obtain relatively accurate estimates on a ques- clon and answer basis. Traditionally, one of the most respected talents is the ability to quickly divide large amounts of local food equitably among large numbers of families at island celebrations. The authors have observed the skill of both men and women at this task. Therefore, due to these environmental, economic and cultural fac- tors, it appears that the islanders themselves may eventually produce more accurate estimates of the foods they eat than those likely to be obtained by outside observations. A crucial problem for an outside observer is that of finding the “typical” family upon which to base his observations, since individual families consume variable amounts of local foods. Some appear to eat primarily a local diet, while that of others contain many imported foods. An analysis of the individual answers of the interviews shows the scope of this variability. However, observations indicate a large variance about the average which reflects wide variations in personal preferences for foods. This is not to suggest that direct observations, especially if made during a complete 365 day cycle, would not yield significant results - but only that such results could not be considered ‘average’ unless observations of a large number of individuals were made. Such a study would show a "typical maximum" or "typical minimum" diet of such families, due to the fact that they would represent such extremes from the norm that they would stand out to the observer whereas the "typical average" diet of the normal family does not. Therefore an outside observer would have no way of choosing which typical family to observe. i: a t- . bee