applied to obtain final demand total. No previous studies of this kind had been done in the
West Indies and all the data presented in part three was gathered particularly for this
study.

 On the supply side, studies of supply conditions for various commodities provided most
of the raw material for part four. In general the method was to take into account all fac-
tors likely to influence acreages and yields, and to project total supplies on the basis of
these findings. Naturally a great number of factors of an economic, technical, and physi-
cal nature were taken into consideration and our methods of treating these is explained in
part four.

 In reconciling demand and supply and arriving at import demand we have assumed a
constant price relationship between products as a first assumption. Where a product is
mainly imported or is produced for export as with sugar and citrus we have generally
assumed supply to consumers will be elastic and no considerable changes in price rela-
tionships are likely. We have had to make an exception in the case of bananas, where
domestic supplies tend to be a residual from exports rather than exports a residual after
domestic needs have been met, and have predicted that price adjustments will follow from
discrepancies between demand and supply projections. As regards commodities produced
mainly for local consumption, we have also predicted price adjustments rather than im-
portations to meet a deficiency between demand and supply, particularly for certain roots
and vegetables and fresh milk.

 Only for population have we presented two possible projections, but we have only used
 one throughout the study. It would be possible for us to present any number of projections
 of the important indicators based on different assumptions regarding income growth,
 population growth, demand changes and supply changes. We do not favour a method of
 presentation which proliferates the number of alternative projections and we feel that to
 present two projections would in most cases emphasise maxima and minima rather than
 the "average" or most probable course of events. Our technique has been to present one
 set of what we feel to be the most probable projections, at the same time discussing the
 modifying influences which certain different assumptions are likely to have on them and
 in particular we have endeavoured to present our information in such a way that, if proper
 mechanical equipment becomes available, alternative major assumptions can easily be
 fed into the base model.

 One point which must be mentioned here since it had some influence on our method is
 that the West Indies is at the moment very much at cross roads as regards its economic
 future. For most territories the fast tempo of economic advance experienced in the
 1950's had already slowed down by 1958 and many of the major trends of the fifties have
 not been continued into the sixties. This has made it almost impossible to project straight
 line or curvilinear trends solely from past trends. In fact the economic situation has
 been such as to discourage the use of purely mathematical relationships and to encourage
 the assimilation of empirical observations into our general conclusions, and this we can-
 not help but feel has been of great benefit to the study.