maintain this flexible attitude. 
Contrary to the opinions advanced by some manufacturers of orthophoto 
equipment, use of orthophoto maps for scales smaller than 1:25,000 entails 
certain difficulties which definitely make it not as advantageous and straight- 
forward as in the large-scale range. This is simply because many details cease 
to be recognizable at smaller photographic scales and thus require symbolization. 
Therefore, the question arises: What is the advantage of using the orthophoto 
approach for small-scale mapping (1:25,000 and smaller) if extensive symboliza- 
tion is necessary? From personal experience, I would like to make the following 
observations. 
Unless the photographic background is undesirable (which in some 
instances is the case) there is no convincing argument why symbolization should 
not be added to the orthophoto manuscript. Most often the characteristic 
features of the terrain can better be presented by an adequately processed ortho- 
photo map than by using the conventional technique. This applies particularly 
to less developed areas, where the need for mapping is urgent. In such areas, 
there are relatively few artificial details requiring symbolization; the greater 
part of a map consists of natural features such as topography (contour lines), 
bodies of water and vegetation. In this respect, there is no doubt that the 
orthophoto technique offers a superior form of cartographic presentation. In 
addition, the basic sheets are obtained simultaneously and at great speed. 
If, however, a conventional line-drawn map is required, the orthophoto 
technique [6) may present an attractive and efficient solution. The usual 
orthophotos, the contour line sheets, and the shaded relief sheets of the land 
to be mapped are produced in one scanning operation. Then the orthophoto 
manuscript is observed stereoscopically together with the enlargement of the 
Second photograph of the stereopair (or its contact copy, if a special observa- 
tion device is available). By this mode of observation, superior interpretability 
of the viewed "stereomodel" is achieved in a manner very similar to that on 
Stereoplotters and the details that must be symbolized are drawn or scribed. 
In this regard, a very special note must be made of the Stereo-ortho- 
photo developed at the National Research Council of Canada [7]. In this novel 
technique, a "stereomate" of the orthophoto is produced, preferably in the same 
Scanning process as the original orthophoto. The stereomate is a rectified 
photograph, similar to the orthophoto, with the difference that the horizontal 
parallaxes of terrain points are proportional to the elevation differences. 
The important feature of this approach is the fact that the vertical parallaxes 
between the orthophoto and the orthophoto-mate have been eliminated in the 
scanning process and the horizontal parallaxes follow a simple, linear function 
of the terrain elevation differences. As a result, stereoscopic viewing is 
particularly simple and convenient and the vertical differences in the "stereo- 
model" can be determined at any point with great ease. It is obvious that 
stereo-orthophotos, complemented of course by contours, offer an advanced concept 
in the presentation of terrain [6] and at the same time, are very suitable for 
converting orthophoto maps into line-drawn maps, if this should be required. 
There is a great advantage in this approach as far as conventional map 
production is concerned, which seems to be particularly valid for less developed 
areas: the basic map manuscripts are produced with great rapidity in a simple 
semi-automatic scanning process and the actual line drawing and symbolization 
are removed to the stage of final drafting where they more appropriately belong. 
The principles sketched above have been experimentally tested in our 
laboratories. The purpose of discussing them here is simply to indicate that 
the orthophoto technique is only in its infancy and that its various possibilities 
greatly exceed the initial "map substitute" concept. Although at the moment 
small-scale mapping may not be the field in which this technique will have the