RELIEF MAPS. 
353 
look like inverted icicles as they have been represented in 
certain models designed for the schools. 
The exaggeration of the vertical over the horizontal scale, 
needful for yielding a map which shall best express the 
topography, appears to be a function of two variables. 
These variables are the horizontal scale and the character 
of the country to be depicted. When the horizontal scale 
is large the area represented by the model will be small, 
and small details can be shown. We are nearer to nature 
with the large scale. Our maps are less generalized. Our 
model should, therefore, be less generalized and nearer to 
nature, for the true proportions of things shown should be 
presented. If we should model the Capitol building all 
will agree that it would be objectionable to use different 
scales for its length and height. Proportion would be de¬ 
stroyed. Should we make a miniature, not only of the 
Capitol, but of the surrounding parks and buildings, true 
scale relations should still be preserved. And this remains 
true if we should even include the entire District of Colum¬ 
bia. But when we embrace a yet wider area, as, for exam¬ 
ple, the entire States of Virginia and Maryland or the whole 
United States, still keeping the size of our model within 
moderate limits, we can no longer show details. The entire 
city of Washington is reduced so nearly to a point as to be 
shown merely by a convention, and the surface inequalities 
of the city and whole District fade out relatively to the 
whole. Only the great inequalities of surface can be truly 
shown. But precisely this is the object of a relief map cov¬ 
ering a large area. It is to bring under a single view the 
great surface irregularities of an extensive area. This re¬ 
quires that we ignore small surface inequalities. 
If we model a large area, and of course upon a small 
scale, we sometimes have a mountainous or very rough 
part and a plains or prairie area which, relatively to the 
other, is flat, even though on closer view considerable un¬ 
evenness of surface appears. In such cases, shall we por¬ 
tray the mountainous parts in their true proportions and 
leave the small inequalities of the plains unrecognizable, or 
