[Illustration: FIG. 18--Map of the same area shown in Fig. 17 enlarged from the corresponding sections of the 1:62,500 Columbus and Seale, Ga.-Ala., and the 1:125,000 Talbotton and Opelika, Ga.-Ala., topographic sheets surveyed mainly in 1906 and 1907 and published by the U.S. Geological Survey. The cross section at the bottom lies along the line indicated on the map and extends somewhat beyond the right border of the map. The section shows the broad shelf upon which the city rests and its relation to the river and to the terraced hillsides east of it. Scale, 1:38,000.] CHAPTER VI MARSHES AND MARSH DRAINAGE (FIGS. 19 TO 27) Mention has been made of the objects seen better from the air than from any viewpoint on the ground; but there are some objects which as a whole can be seen only from above. Swamps, parts of everglades, peaks in the midst of difficult country, precipitous canyon walls, and many volcanic craters cannot be seen from the ground without undue effort. Photographs of bluffs, terraces, and other slopes facing bodies of water have hitherto been adequately obtainable only from the water. All of these can be readily viewed and photographed from the airplane. Pictorial representations of drainage systems were rare until photographs such as Figure 19 were taken from airplanes. The intricate drainage of marshes like those along the Pamunkey River in Virginia pictured in Figure 20 was never accurately shown until photographed from the air. Of frequent occurrence on the Atlantic Coastal Plain of the United States are swamps and marshes inaccessible from the ground. Much of the surface material is so soft that they cannot be easily traversed; and, even where firm enough to support a man’s weight, few of the details are deemed of sufficient importance to warrant the trouble and expense of mapping by ordinary methods. Yet the trapper would scarcely admit that these details are unimportant, and, to the student, they are an interesting feature of marsh topography that has thus far received little attention. Figure 25 is part of the excellent New Kent, Va., sheet of the topographic map and is probably as detailed as a map of this character should be when made from ground surveys only. However, on comparison of the map with a photograph of the same area (Fig. 24), there is no difficulty in detecting errors; and it is probable that, had the photograph been available when the map was made, the marshes would have been represented differently.