tives, as compared with that assumed for the loaded cars, diminishes as the span-length increases; but in highway bridges the lessening is due to an application of the theory of probabilities. For the electric-railway loading the diminution in the equivalent live load per lineal foot exists only when the length of span is greater than the assumed total length of train, and the greater the difference in length the greater the said diminution of loading per lineal foot of span. On pages 110 to 116, inclusive, of "Bridge Engineering" are given curves of equivalent uniform loads for trains of two, three, four, five, and six electric-railway cars and for all the classes of loading from 15 to 40. These are carried out to a length of only 600 feet; and in case of longer spans it would become necessary to lengthen them either accurately by computation or approximately by visual extension.

For ordinary railroad bridges the determination of the proper live load to adopt is generally a simple affair, because the company is likely to have a standard of its own; but sometimes it may be advisable to depart from this, in order to provide for some unusual condition or complication. For highway bridges, though, it is a different matter, because the amount and character of the traffic will depend greatly on local conditions; and, consequently, the size of the loading is a question of judgment. In exercising it, one should look to the possibility of a material increase in the weights to be transported; and should be governed accordingly.

The standard steam-railway loadings given in Chapter VI of "Bridge Engineering" are still sufficient for maximum requirements, and it is not likely that the greatest (Class 70) will ever be materially exceeded, unless some fundamental improvement be made in the character of railway roadbeds. The distribution of weight on the axles of the locomotives is gradually being changed, but the weight thereof per lineal foot of track does not appear to be augmenting to any great extent.

The heaviest live loads for electric railways given in the aforesaid chapter have not yet been reached, except in the case of electrified steam-railroads; and these are not treated as real electric railways, as far as bridge loadings are concerned. Ordinarily, Class 25 is heavy enough for the street-railway live-loads on highway bridges; but sometimes it is advisable to proportion for Class 30, in order to provide for future heavy suburban cars.

Classes A, B, and C for highway loadings are sufficiently heavy, even if it be possible to crowd vehicles and pedestrians so close together as to cause the actual loadings to exceed them; because, when much crowding occurs, the speed of travel is materially reduced, and then the effect of impact may practically be ignored. Class A loading is so great that, for comparatively long spans and wide decks, it may be used to cover the average loading over the full width of deck from electric-cars, vehicles, and pedestrians-and this applies in nearly all cases to suspension bridges. But in that type of bridge it must be remembered that for the stiffening trusses the equivalent uniform live load per lineal foot of structure must