region, and insist in the first place upon the same itinerant field
methods of notebook and camera, even for museum collections and the
rest, as those of the natural sciences. The dreary manuals which have
too long discredited those sciences in our schools, are now giving place
to a new and fascinating literature of first-hand nature study.
Similarly, those too abstract manuals of civics which are at present
employed in schools[5] must be replaced by concrete and regional ones,
their abstract counsels of political or personal perfection thus also
giving place to a corresponding regional idealism which may then be
supplemented from other regions as far as needs demand and circumstances
allow.

[5] For a fuller review of these, compare the writer's "City
Development," in _Contemporary Review_, October, 1904.

C--GEOGRAPHICAL DETERMINISM AND ITS DIFFICULTIES

To interpret then our tangle of ideas, both of the city and its
citizens, let us now bring more fully to our transverse valley sections,
and to each occupation separately, the geographical view-point which we
have found of service to elucidate the development of towns and cities
upon its longitudinal [Page: 64] slope. But this is neither more nor
less than the method of Montesquieu, whose classic "Esprit des Lois"
anticipates and initiates so much of that of later writers--Ritter,
Buckle, Taine, or Le Play. Once more then let their common, or rather
their resultant, doctrine be stated in terms expressing the latest of
these more fully than the first. Given the region, its character
determines the nature of the fundamental occupation, and this in turn
essentially determines the type of family. The nature and method of the
occupation must normally determine the mode of its organisation, e.g.,