THE COUNTRY OF THE SHEPHERDS 51 



scattered by plundering bands and war parties. The remote and 

 isolated group may successfully oppose the smaller band and the 

 individuals that might reach the remoter regions. The fugitive 

 group would have nothing to fear from large bands, for the 

 limited food supply would inevitably cause these to disintegrate 

 upon leaving the main routes of travel. Probably the fullest ex- 

 ploration of the mountain pastures has resulted from the alterna- 

 tion of peace and war. The opposite conditions which these estab- 

 lish foster both kinds of distribution ; hence both the remote group 

 life encouraged by war and the individual's lack of restraint in 



Xote on regional diagrams. — For the sake of clearness I have classified the accom- 

 panying facts of human distribution in the country of the shepherds and represented 

 them graphically in "regional" diagrams, Figs. 17, 25, 26, 32, 34, 36, 42, 65. These 

 diagrams are constructed on the principle of dominant control. Each brings out the 

 factors of greatest importance in the distribution of the people in a given region. 

 Furthermore, the facts are compressed within the limits of a small rectangle. This com- 

 pression, though great, respects all essential relations. For example, every location on 

 these diagrams has a concrete illustration but the accidental relations of the field have 

 been omitted; the essential relations are preserved. Each diagram is, therefore, a 

 kind of generalized type map. It bears somewhat the same relation to the facts of 

 human geography that a block diagram does to physiography. The darkest shading 

 represents steep snow-covered country; the next lower grade represents rough but 

 snow-free country; the lightest shading represents moderate relief; unshaded parts 

 represent plain or plateau. Small circles represent forest or woodland; small open- 

 spaced dots, grassland. Fine alluvium is represented by small closely spaced dots; 

 coarse alluvium by large closely spaced dots. 



To take an illustration. In Figure 32 we have the Apurimac region near Pasaje 

 ( see location map, Fig. 20 ) . At the lower edge of the rectangle is a snow-capped 

 outlier of the Cordillera Vilcapampa. The belt of rugged country represents the 

 lofty, steep, exposed, and largely inaccessible ridges at the mid-elevations of the 

 mountains below the glaciated slopes at the heads of tributary valleys. The villages 

 in the belt of pasture might well be Incahuasi and Corralpata. The floors of the 

 large canyons on either hand are bordered by extensive alluvial fans. The river 

 courses are sketched in a diagrammatic way only, but a map would not be different 

 in its general disposition. Each location is justified by a real place with the same 

 essential features and relations. In making the change there has been no alteration 

 of the general relation of the alluvial lands to each other or to the highland. By 

 suppressing unnecessary details there is produced a diagram whose essentials have 

 simple and clear relations. When such a regional diagram is amplified by 

 photographs of real conditions it becomes a sort of generalized picture of a 

 large group of geographic facts. One could very well extend the method to the 

 whole of South America. It would be a real service to geography to draw up a set 

 of, say, twelve to fifteen regional diagrams, still further generalized, for the whole 

 of the continent. As a broad classification they would serve both the specialist and 

 the general student. As the basis for a regional map of South America they would 

 be invaluable if worked out in sufficient detail and constructed on the indispensable 

 basis of field studies. 



