Mr. Cudworth. near Whitby, where the cross bracing of the girders was very defective (Figs. 18). It was a shallow viaduct of no great height, and it had been possible in that case to take out the old girders from

STRENGTHENING OF THE TEES VIADUCT, BARNARD CASTLE.

under the floor, leaving the old flooring, and to put new girders under the latter, so that the upper part of the bridge remained as it was, but with new girders. The work had been executed

A Mr. Cudworth.

successfully at a cost of £2,600 for five spans of 60 feet. similar viaduct over a very deep ravine at Upgang, near Whitby, had required strengthening where it was not possible to get even temporary support from below. In that case, while the girder was of the same type as the girders of the West Lynn Bridge, it had been the bracing which was at fault, and not the flanges, the former having been badly designed and much crippled. Fresh bracing was put in without removing the old (Figs. 19). Verticals were put in at the centre of each panel; end bracing was then put in connecting the panels, and that bracing had to be carried round the old bracing. It had rather a singular appearance, but it had been Fig. 18.

DEFORMED STRUTS REMOVED IN COURSE OF STRENGTHENING.

very successful, and not at all costly. About £800 was spent on the viaduct, which consisted of five spans of 60 feet, and no trouble whatever had been experienced since. The Author had referred to the great difficulty of adding to the flanges of the West Lynn Bridge. Mr. Cudworth remembered one case on the North Eastern Railway where the flanges of a bridge had to be added to, although not to so large an extent as on the West Lynn Bridge. The bridge, which crossed the River Eden at Musgrave, near Kirkby Stephen, consisted of three spans of 66 feet, the width of the flange being just the same as that of the West Lynn Bridge, namely 2 feet. The method adopted (Fig. 20) was much the same as in the work carried out at the West Lynn Bridge, rivets being cut out and

Mr. Cudworth. bolts substituted, but the work was done half at a time. The old cover-plates were cut down the middle of the girder and one-half was removed first; then new plates were added, which were secured by bolts, and riveted up during the week, the plates being

STRENGTHENING OF THE UPGANG VIADUCT, NEAR WHITBY.

put in on the Sunday when there was no traffic. Such treatment had been much more feasible in the Musgrave Bridge than in the West Lynn Bridge, because the dead load did not bear nearly such a high ratio to the total load; so that while the rivets were all out, and the bottom plate was being added, what was left of the bridge

was quite sufficient to carry it during the operation. The effect Mr. Cudworth. of the work had been to reduce the deflection of the bridge from a little more than inch to about inch. The bridge had stood very satisfactorily ever since, and the cost of the work had been only £350. The Paper did not refer to the renewal of wrought-iron bridges, but he would like to say a word in favour of what might be called the flat-floor type of bridge. The Author, in referring Fig. 20.

STRENGTHENING OF THE MUSGRAVE BRIDGE, NEAR KIRKBY STEPHEN.

to Figs. 14, Plate 5, spoke of the floor as being "tied," no doubt meaning thereby a road not resting on sleepers, and which could not readily be adjusted to suit the superelevation or curvature of the rails. There could be no doubt that a good stiff floor covered with flat plates and a layer of good asphalt, on which ballast and ordinary sleepers and permanent way were laid, was the ideal floor for railway-bridges. It could not always be got, but, wherever possible, he strongly advocated the use of a floor of that type.

Mr. Sadler.

Mr. H. W. SADLER thought there were three principal reasons why old bridges required strengthening; first, on account of improper or bad design, which was a cause of the weakness of many bridges; secondly, the increase in the weight of locomotives; and, thirdly, corrosion. Some old bridges were very badly designed; the girders were too shallow, some being only one-thirtieth of the span in depth. A very common defect was that the webs were too thin, a usual practice having been to make webs of cross girders only inch thick throughout. There was a want of sufficient rivet- and bearing-area, and the connections were weak. The girders were seldom deficient in flange-area. The increase in the weight of locomotives was becoming a serious matter, and had necessitated the strengthening of many bridges. In 1870 engines weighed 40 tons; now they weighed over 78 tons. In those days engineers took 1 ton per lineal foot as the usual weight for an engine measured over buffers, and 2 tons per lineal foot measured on the wheel-base; at the present time the corresponding loads were 1.82 ton and nearly 4 tons. Of course many of the old girders were incapable of standing such an increase. The Author's statement that no iron bridge, however old, rusted so quickly as new steel bridges did, was a very serious one, in view of the fact that so many steel bridges had been built; and it was certainly well worthy of discussion. Text-books gave the coefficient of corrosion for wrought iron in pure air as 0.0123, and that of steel as 0.0125; while in towns and manufacturing districts it was 0.1254 for wrought iron and 0.1252 for steel. If those figures were correct, steel should not corrode more than wrought iron. He did not wish to assert that steel did not corrode, but he thought it was questionable whether it corroded more than wrought iron, and his experience did not agree with the Author's. Certain steel bridges which he had examined had been in work more than 18 years and were in good condition, whereas some wrought-iron bridges had corroded all to pieces in 23 years. Corrosion was largely a question of maintenance and situation. He would like to ask whether the mill-scale was taken off the plates to which the Author referred; because if it were not removed, rust would result. Mill-scale could not at first be brushed off with wire brushes; it was impossible even to hammer it off; and until it was removed rust was sure to occur. The Author stated that no plate should ever be allowed to rust; whereas, many leading engineers at the present time were purposely leaving their girders to rust so that the mill-scale should be taken off. He would be glad to hear the Author's opinion on that point, because both contentions could