phere in the tube perfectly safe for breathing, especially by people of feeble constitution.

There is another factor that is likely to have considerable influence in deciding between bridge and tunnel, viz., that driving over the former is much more agreeable than driving through the latter. Again, walking through a long tunnel is so unpleasant that it would seldom be worth while to make any provision whatsoever for pedestrian travel therein.

Then, too, the difference between the height of climb over a bridge and the depth of descent into a tunnel will have some effect upon the choice of structure. In general, it may be stated that the dip of a tunnel is greater than the rise of a low-level bridge, about equal to that of an ordinary high-level bridge, and less than that of a structure below which pass ocean-going vessels.

Supposing, however, that all the pros and cons of the two types, barring cost only, about balance each other, the question for settlement would be one of economics; and then the decision would favor the bridge, because for equal facilities the cost of the tunnel almost always greatly exceeds that of the bridge. Of course, the conditions affecting the latter might be so onerous as to increase the cost of the structure beyond any reasonable limit—for instance a span of unprecedented length; but, in general, the bridge costs less than the tunnel.

In the author's practice he has twice had occasion to pit bridge against tunnel. In the first case a highway suspension bridge with a span of 1,750 feet and a total clear roadway, including sidewalks, of 96 feet was contrasted with two tubes which together gave a clear roadway of 56 feet. The result was that the tubes, in spite of their smaller carrying capacity, cost about fifty per cent more than the bridge.

In the other case, a single-track-railway tube cost considerably more than a double-track, low-level bridge-and even a little more than a low-level, combined-railway-and-highway, double-deck structure. Comparing the tunnel and a high-level bridge with a clearance above high water of 150 feet, the ratio of costs of a two-tube, railway tunnel and a double-track, railway bridge was 1.1; and comparing a three-tube tunnel for both railway and highway traffic with a combined-railway-and-highway, high-level bridge, in which comparison the facilities afforded were nearly equal, the ratio of costs was 1.12—both of these results being in favor of the bridge.

There is one advantage which, theoretically, the tunnel possesses over the bridge; and under certain conditions it might become practically operative, viz., that, while a wide-decked bridge has to be built all at once, several tunnels of the same aggregate width can be constructed from time to time as the traffic necessitates, thus saving for some years the interest on the difference between first costs.

A claim has been made that any bridge having more than about forty feet of deck-width will congest the traffic to such an extent that there is no advantage to be obtained from the extra width above the said forty feet;