[Illustration:  Bridge over the Blaauw Krantz ravine, Cape Colony.]

[Illustration:  Bridge over the Blaauw Krantz ravine, Cape Colony.]

In designing the structure the following points had to be considered:  (1) That, on account of the great height above the ground, and on account of the high price of timber at the site, the structure could be easily erected without the use of scaffolding supporting it as a whole. (2) That, on account of the high freights to Port Alfred, the quantity of iron in the structure should be as small as possible. (3) That the single parts of the principal span should be easy to lift, and that there should be as few of them as possible.  For this latter reason most of them were made in lengths of 20 ft. and more.  The question of economy of material presented itself as a comparison between a few standard types, viz., the girder bridge of small independent spans; the cantilever bridge, or the continuous girder bridge in three large spans; the single girder bridge with one large span and several small spans; and the arch with small girder spans on each side.  The suspension bridge was left out of question as inadmissible.  A girder bridge with small independent spans on rocker piers would probably have been the most economical, even taking into account the great height of the piers near the middle of the ravine, but there would have been some difficulty in holding those piers in position until they could be secured to the girders at the top; and, moreover, such a structure would have been strikingly out of harmony with the character of the site.  On the other hand, a cantilever or continuous girder bridge in three spans—­although such structures have been erected in similar localities—­could not enter into comparison of simple economy of material, because such a design would entirely disregard the anomaly that the greater part of the structure, viz., the side spans, being necessarily constructed to carry across a large space, would be too near the ground to justify the omission of further supports.  The question was, therefore, narrowed to a comparison between the present arch and a central independent girder of the same span, including the piers on which it rests.  The small side spans could obviously be left out in each case.  The comparison was made with a view not only to arrive at a decision in this particular case, but also of answering the question of the economy of the arch more generally.  The following table contains the weights of geometrically similar structures of three different spans, of which the second is the one here described.  The so-called theoretical weight is that which the structure would have if no part required stiffening, leaving out also all connections and all wind bracing.  The moving load is taken at one ton per foot lineal, and the strain on the iron at an average of four tons per square inch.  The proportion of the girder is taken at 1 in 8.