ABSTRACT
In all the problems which have so far been considered in this book, the aim has been to reduce or eliminate the detailed differences between one particular value and another, so that the overall characteristics can more readily be appreciated. Even in the case of correlation, where individual values were more directly considered, the purpose was to obtain one index which would summarize the full set of individual relationships, whilst with regression lines one basic mathematical relationship between two variables is sought. With some problems, however, the geographer must necessarily concern himself with the details of the changes from one individual value to another. This is so when the data consist of values which change in relation to changes in the time-scale. Thus it is possible that changes in production, in climatic conditions or in population values bear some relationship to such time-scale changes.
