This paper presents the results of the writer's experience with the geological use of aerial photographs. The writer is very optimistic about the future of aerial photography for geological discovery, not only in the wilderness but in certain regions with a long history of exploration by other methods. Yet in the field of topographic mapping it is apparent that too much has been expected of the stereoscopic study of photos--a mistake which is not confined to geologists. Stereoscopic mapping does not yield results of great accuracy unless pains are taken to establish a sufficient number of bench-marks by ground surveying.

Some of the common faults in stereoscopic vision are discussed, as are some of the causes of distortion in the stereoscopic model.

Photo-indexes in terrane of low relief, such as the low-strata-bench-lands and true plains, if properly prepared, are valuable "picture maps," instead of being, as they originally were intended to be, merely a guide to the location of the contact prints from the flight negatives.

In regions of low relief mosaics have a cost out of proportion to their geological value; a good photo-index sheet is to be preferred for geological study. In regions of high relief, however, the mosaic is thus far the only suitable device for securing a continuous picture map, and as such is very valuable. In studying terrane of either low or high relief, the geologist will always need stereoscopic contact prints for the most satisfactory work.

It is premature to make definite statements about the exact degree of usefulness of aerial photographs in geological exploration, since photos have seldom been made to the specifications of the geologists who study them. It happens all too frequently in photos of the low-strata-bench -lands that the bedding is not visible even when magnification is used.

Nevertheless, if aerial photographs are used chiefly for discovery the writer feels that the expense will prove to be a negligible quantity in comparison with the high returns.