GVII-58 
PHOTOGRAMMETRIC ENGINEERING 
one topic at a time. For example, all that is needed regarding transportation 
should be interpreted before proceeding to the topic of drainage. It is true that 
the interpretation of one topic aids another. Such aid is apparent as an inter 
preter proceeds from topic to topic, particularly if it is in the order listed and 
described below. 
Some research requires description or mapping of all of an area. However, 
unless full coverage is demonstrably essential, interpretation by the sampling 
procedure is highly recommended. With the use of other source material re 
searchers usually should be able to select areas, strips, lines, or spots and to 
place them in representative, mechanical, or random locations before interpreta 
tion is begun. Certainly research with air photos should not be assumed to re 
quire complete analysis of an area any more than if that area is analyzed by 
large-scale maps or detailed statistical data. 
2) Interpretation should he made from the general items to the specific. Or, 
we may say from the small-scale to the large-scale considerations, whether or 
not generalizations are to be the end product of the analysis. The first view of 
the photos should be with the naked eyes and of the smallest scale coverage 
available, such as photo indices, mosaics, and smaller scale verticals and 
obliques. By this step are discovered broad regional patterns which are too 
subtle to be seen on single photos as well as the discontinuous patterns which 
must be viewed over large areas to observe any continuity at all. After the naked 
eye study, examination of the smaller scale verticals with a mirror stereoscope, 
without binoculars, will be a natural step toward more detail. The study of 
large-scale vertical coverage is more effective when done last for here detail 
is the only thing apparent. In particular is this true when pocket stereoscopes 
are used because the diameter of the effective area seen directly beneath the 
two-power lenses is less than three inches, and that beneath the four-power 
lenses is less than one and one-half inches. 
In general, there is too great a tendency for photo interpreters to become 
lost in detail, and too little effort to conquer the tendency; the prevention, and 
more efficient interpretation, is by working from the general characteristics 
toward the specific ones of each topic interpreted. 
3) Interpretation should be done from the known to the unknown features. 
Just as we “screen” the photos a topic at a time, so should we establish early 
what is clearly recognizable from other source materials and previous experience. 
Generally there is little need to methodically identify a feature on a photo when 
an available map or report already identifies it. For encouragement and efficient 
interpretation it is valuable to establish rather soon in the research what parts of 
a topic are recognizable on a photo. These items must then be verified by com 
parison with source materials, by examination of the photo pattern, and by 
measurement of the object to determine its size (particularly necessary with 
respect to cultural features). What is absolutely identifiable on the coverage 
is then available for comparison with the unknown features; this comparison 
of known with unknown is very basic procedure. 
Then two thought-provoking questions must be applied rigorously to the 
unknown. Is it a part of or directly related to the topic being analyzed? If the 
answer is in the negative the item should be skipped immediately. If there is 
doubt or if the answer is in the affirmative, identification should be sought. 
Any possible identification must then be recognized as only a possibility by 
repetition of the second question: what else could it be? This question must be 
posed many times until identification is certain or until the number of possibili 
ties is as limited as it can be. It is this last question which is one of the more