local food they prepare for their families to eat. It may therefore be concluded that the local diet is basically quite uniform and that it changes primarily due to environmental conditions. The effect of imported food is not s0 much to change the elements of the local diet but simply to reduce them proportionately. The only exceptions to this tendency towards proportionate over-all reduction are Jekaru (coconut sap), Mokmok (arrowroot), and Jankwon (preserved mejwan breadfruit and preserved pandanus). This may be due to the intense labor invoived in the processing and preparation of these three foods. They appear to be the first traditional foods to be replaced from a total local food diet by imported sugar, rice and flour. However, further studies are needed to conclusively demonstrate this. With respect to community (A) where estimates showed the food prepared and eaten to be nearly 100% of the total diet, it is clear that these estimates exceed the actual amount that could conceivably be consumed, even by all the fam- ily members. This is especially so considering the fact that this group of fam- ily members includes women and children who could not possibly consume all that food on a daily basis when we know that they are eating significant quantities of imported foods as well. Table 4A and 4B represent a typical maximum diet. [It represents the most conservative estimate on the total gram weights of the various local foods which could conceivably be consumed under the assumption of a 100% local diet. , These estimates are based on the assumption that ali the Marshallese lLiving on outer islands regulate their dietary habits to a certain extent to a pat- tern parallel to environmental conditions and the natural food gathering cycles that are governed by these conditions. It is based on a general observation that most islanders do eat local foods. These estimates also indicate how much of a particular food is eaten (by a typical adult and child) during a given foods' peak season or seasons. They do not consider those periods when a partic- ular food is scarce or otherwise difficult to obtain. Since these estimates are based on a cycle of one year, it seems reasonable to assume that this method could provide an estimated maximum. It has also the advantage of being based on principles and assumptions which are scientifically verifiable. The various growing seasons are subject to yearly change. Also the length and production of each growing season varies somewhat from year to year. In calculating the maxice ne te ee mum diet the tabulations reflect a somewhat higher percentage of jekaro, coconut and pandanus than could reasonably be expected. It should be noted that an individual existing totally on such a diet would have to be carrying out a very active food gathering existence, and would therefore have very Little time for other endeavors. In short, he would have to return to the premodernized state his ancestors were living 200 years ago. It should also be noted that a higher maximum consumption of any one type of food is conceivable though it would be unlikely for two reasons. One, is the fact that the premodern Marshallese society as well as the contemporary society is very communal in its food consumption patterns. This means that food sharing is extremely important, and therefore if any ome person gathers a great deal of any one particular type of food, he is more likely to divide it up and give it away _al