Photographs of channels like those of the Potomac River and its tributaries will be commercially as well as scientifically valuable. The deep-water channel of the Potomac is well known and has been charted; but very little is known of many of the small tributary channels, such as that of Powells Creek (Fig. 75). Where the channels are not well known, such a photograph could be used to advantage in avoiding the shoals, and, by surveyors, first in exploratory work and later as a general guide in charting. Small boats entering this channel could use the photographs either for the original location of the deep channel in case no chart were available or for detecting changes in its course after the chart was made. For uncharted channels, like those of many of the tributaries of the Potomac River, air photography furnishes a quick and accurate means of location. No amount of sounding, charting, or description could produce so accurate a mental picture of a drowned valley as that produced by Figure 76. In Figures 78 and 79, both of which were taken near Miami, Florida, is illustrated the difference in appearance between natural and artificial channels. The straightaway course and regular outlines of the dredged channel contrast sharply with the winding course and merging outlines of the natural channel. To the student of physiography and earth history the photographs furnish a means of observation of a definiteness heretofore quite unthought-of. On them the actual shape of the channels, submerged terraces, and drowned land forms are shown in detail. IMPROVEMENTS UNDER WAY POINT TO PROMISING OUTLOOK FOR AIRPLANE PHOTOGRAPHY There is, however, need of careful research to determine the conditions under which the best results can be obtained. The height and time of day for exposures with a certain lens, the emulsion and kind of ray filter best suited under certain conditions, the effect of light as it enters and emerges from the water, and the effect of polarization are subjects demanding consideration. Chief among the experiments now under way is the determination of the kind of emulsion and ray filter or color screen that will give the best results. It is a demonstrated fact that, with an emulsion sensitive to red light, objects in the air invisible to the eye because of intervening haze can be photographed through a red filter. It is possible that water can be penetrated in the same way and that filters of other colors will prove advantageous.