Key points are not available for this paper at this time.
IF WE look away from certain rather new sorts of purely theoretical investigations which can be undertaken without very much of an observational base, geographical research almost by definition has to be founded on empirical data. However, we do not need just a few scattered observations but, more often, information in very large quantities. We further adhere to the habit of depicting on maps the information we have obtained, both for the sake of analysis and for the sake of communication. This habit forces us to assemble complete information concerning the area under observation to a degree which can often be avoided in other branches of research. Undoubtedly we have all at times met a situation where mapping or calculation was never accomplished for the simple reason that the amount of work proved to be insuperable.
Hägerstrand et al. (Fri,) studied this question.