diagram and joining them properly by very slightly curved lines shows that the span length of equal cost is reduced from 2,700 feet to 2,640 feet. This is no material amount, indicating, as it does, a variation of only 2.2 per cent.

ADDENDUM

The preceding was written in the summer of 1918. A year later the author was called in by some prominent citizens of Detroit to make a study of the governing conditions for a proposed highway-and-street-railway bridge over the Detroit River, practically on a line joining the business centers of the cities of Detroit and Windsor, and to determine upon the best type of structure to adopt. A few days of investigation led to the conclusion that a single span of 2,500 feet, crossing the entire river in the clear between harbor lines, would be obligatory; and, accordingly, the layout and the approximate cost-calculations were made for a suspension bridge. It became necessary to obtain pound prices for structural metal (both nickel steel and carbon steel) and wire cables in place; and the following values were found:

The last figure was simply staggering! Surely, such an enormous price can be only temporary, for the great difference between it and the other two figures is altogether illogical. Nevertheless, it shows the possibility of an abnormal price-condition existing long enough to affect temporarily the economics of cantilever and suspension bridges. It is not likely that there can ever be a worse condition than the one at present governing; consequently, the author has recast for existing unit prices the estimates of cost made for the preceding investigation, and has found the following results:

The span of equal cost for highway bridges has been advanced from 1,000 feet to exactly 1,200 feet; that for the particular combined bridges investigated has been increased by 170 feet; but that for the steam railway bridges has been augmented only 60 feet. The reason for the smaller increase in the last case is that, in cantilever structures the weight curves, and consequently the cost curves, rise very rapidly at a span of 2,700 feet, because such a length is really a little beyond the truly practicable limit for that style of bridge.

These variations are somewhat greater than the maximum which the author anticipated when writing his paper; but at that time he never would have deemed it possible that such a great variation in unit prices of structural steel and wire cables could hold as that which exists to-day; nor does he now consider it possible that it can be made to last for any great length of time.