RELIEF MAPS. 
355 
Now it is the existence of rough country that awakens 
our interest in and shows the need of relief maps. Such 
maps are primarily devised to aid in gaining a true percep¬ 
tion of topographic relations in rough country. They are 
not needed for very flat areas unless for special purposes. 
The aim of the relief map is to give a comprehensive and 
true view of the large features over a wide area. We there¬ 
fore act unwisely if we distort the picture of these great 
masses to save lesser details of relatively minor importance. 
If we are to model Colorado, let the vertical exaggeration 
be chosen with a view to correctly portraying her majestic 
mountains even at a total loss of the character of the small 
but interesting topography of the outlying foothills and 
plains. If this plains topography is to be shown in relief, 
let it be done by itself. In the relief map as well as in its 
forerunner, the topographic map, the less must be subordi¬ 
nated to the greater. In choosing our scale of vertical ex¬ 
aggeration let us therefore have it (a) uniform throughout 
and (b) selected with reference to the best exhibit of the 
great relief features. 
If models of large areas and on corresponding^ small 
horizontal scales are carefully compared with those of 
smaller areas on larger scales, it is found that an equal 
amount of vertical exaggeration in each produces unequal 
results. If the fitting exaggeration has been chosen for the 
small scale, then the large-scale model appears over exag¬ 
gerated. For example, the model of the State of California, 
horizontal scale 4 miles to one inch, with a vertical exag¬ 
geration of 4, may be compared with the model of the Yo- 
semite on a scale of two miles to an inch, with no vertical 
exaggeration. Or again, compare Maryland, horizontal 
scale two-thirds of an inch to a mile, vertical exaggeration 
5, with the District of Columbia, horizontal scale 2 inches 
to 1 mile, vertical exaggeration 2. In each case the smaller 
area within the greater has been modeled on a larger hori¬ 
zontal scale and with diminished vertical exaggeration. An 
inspection of the models proves the wisdom of this choice 
and permits us to infer that, other things equal, large-scale 
