greatest good for the greatest number." The limitation of efficiency will generally be found to be the capacity for transferring automobiles; and this can be greatly augmented by making the cages three-decked, taking care of the street-cars and the auto-trucks on the lower one, the automobiles on the middle one, and the pedestrians above.

From the preceding reasoning it may be concluded that in the comparison between a double-track transbordeur with short-cages and a standard, 60 ft.-wide, highway bridge, the former is two-thirds as effective for both pedestrians and street-cars, and only one-tenth as effective for automobiles. If the 100 ft.-long cages were used and only one street-car were carried per cage, the rest of the space being occupied by automobiles, the preceding ratios would be about one and a third, two-thirds, and one-third; or, if the long cages carried two street-cars each, these figures would be one and a third, one and a third, and one-fifth. The comparative efficiencies of the two structures would, consequently, be dependent upon how the traffic is divided between pedestrians, street cars, and automobiles. For the first two the transbordeur can readily be made but little lower in efficiency than the low-level bridge, but for auto-traffic it will always be found decidedly inferior; and, as that traffic is on the increase in both amount and importance, it must inevitably be concluded that, in respect to general carrying capacity, the transbordeur is never as satisfactory as the said low-level bridge.

The making of this comparison was advisable, although by no means essential, because there should never be any choice between a bridge and a transbordeur for any proposed crossing. As stated at the beginning of this chapter, if the low-level bridge is permissible, it should be adopted; but, if not, the transbordeur should be used as a pis aller. The author's object in making the attempted comparison was to confirm in a general way his a priori conclusion.

As previously indicated, there will not be many occasions for the building of the transbordeur; and the conditions of traffic, navigation, river-width, and property for approaches are so variable that each case will require a thorough and systematic compilation of them all, for both the immediate and the distant future, and an exhaustive study of the question how best to compromise between conflicting interests and to develop in general the greatest possible efficiency.

The question might sometime arise as to whether a transbordeur or a high-level bridge would be preferable for the crossing of a waterway navigated only by river steamers, for which the Government's clearance-requirement generally varies from 50 to 60 feet above high-water elevation; consequently, it would be well to know in advance the approximate ratios of cost of a transbordeur to the various costs of the corresponding high-level bridges having differing vertical clearances. For this purpose Fig. 32e was prepared by using as a basis a slight modification of some of the results of the computations for the New Orleans Bridge study. That figure gives