January 29, 1904.] 



SCIENCE. 



181 



of the greatest values of systematic geog- 

 raphy; it enforces attention upon the 

 idealized type ; by means of this increased 

 attention the type is more fully conceived, 

 and both observation and description of 

 actual things are greatly aided. Let me 

 illustrate. 



The breezes that descend from mountain 

 valleys at night are well known and well 

 understood phenomena. As a result, one 

 may form a well-defined conception of such 

 a breeze— a type mountain breeze— im- 

 agining its gradual beginning, its increase 

 in strength with its extension in area, and 

 its gradual extinction; all its phases of 

 waxing and waning being duly related to 

 the passing hours of the night and to the 

 associated changes of temperature. It is 

 safe to say that no actual mountain breeze 

 is as well known by direct observation of 

 all its parts and stages as is the type 

 breeze, in which all pertinent observations 

 are properly generalized, and in which the 

 deficiencies of observation are supple- 

 mented as far as possible by inferences 

 deduced from well-established physical 

 laws. It is entirely possible that there may 

 be some errors in the deduced elements of 

 the ideal type-breeze, but it may be confi- 

 dently asserted that the errors will be re- 

 placed by the truth through the methods 

 involved in observing, imagining and 

 checking, guided by the conception of the 

 type, sooner than the truth will be discov- 

 ered by blind observation unguided by the 

 aid that a well-defined type affords. 



It is the same with an alluvial fan; an 

 element of land form that has, by the way, 

 more similarity to a mountain breeze than 

 appears on first thought. Observation 

 shows only the existing stage of the surface 

 of a fan; the fully developed type-fan in- 

 . eludes the structure as well as the surface, 

 the process and the progress of formation, 

 extended into the future as well as brought 

 forward from the past. There can be no 



question that the explorer who is equipped 

 with a clear conception of a type-fan can 

 do much better work in observing and de- 

 scribing the fans that he may find than 

 will be done by an explorer who thinks 

 he can dispense with all idealized types, 

 and who proposes simply to describe what 

 he sees. The shortcomings of the simple 

 observational method would be less if it 

 were not so difficult to see what one looks 

 at and to record what one sees ; but any 

 one who has had experience in field studies 

 knows how far short seeing may be of look- 

 ing, and how far short recording may be of 

 seeing. The best results in geographical 

 investigation can only be obtained when 

 every legitimate aid to observation and 

 description is summoned; and of all aids, 

 that furnished by carefully considered 

 types, reasonably classified, is the greatest. 

 When large and complicated features, such 

 as valley systems or cuestas, are to be de- 

 scribed, the need of types is vastly in- 

 creased. Hence one of the most impor- 

 tant and practical suggestions that can be 

 made toward the maturing of geographical 

 science is to cultivate the geographical 

 imagination in the direction of acquiring 

 familiarity with a large, systematic series 

 of well-defined ideal types. As progress is 

 made in this direction there will be profita- 

 ble advance from that narrow conception 

 of geography which is based on the school- 

 day study of names, locations and bounda- 

 ries — the only conception of geography 

 that many mature persons in this country 

 possess— to a wider conception in which 

 everything studied is considered as an ex- 

 ample of a kind of things, so that it shall 

 appeal to the reasonable understanding 

 rather than to the empirical memory. 

 Progress of this sort is already apparent 

 in the schools, but it has not yet reached a 

 desirable measure of advance. 



One of the best results that follow from 

 the systematic recognition of a large num- 



