THE FORTH BRIDGE. 429 



bolts provided, however, about double the resistance to overturning, so 

 that 4 cwt. of wind-pressure per square foot upon the single surface would 

 be required to upset the bridge ; and under this ideal pressure, though 

 the wind-bracing would, it is true, be on the point of failing, none of the 

 great tubes or tension members of the main girders would even be per- 

 manently deformed. 



Chancres of temperature as well as gales of wind will affect the stresses 

 upon the steel but to a comparatively insignificant extent. In designing 

 a certain class of work, nothing embari-asses an engineer more than ques- 

 tions arising out of the expansion and contraction of materials. An expe- 

 rienced engineer will know that in certain instances he maj' entirely 

 io-nore the influence of temperature, whilst in others it is all-important. 

 For example, if a provision were not made in the case of point rods, the 

 points would fail to close, and great loss of life might follow the conse- 

 quent derailment of a train, whilst, on the other hand, infinite complica- 

 tion would result if a similar provision had to be made in a structure like 

 the Crystal Palace, or in long station roofs, or in girders carrying build- 

 ino-s. Only experience could decide whether it would be safe to carry 

 an ornamental ashlar-faced building on a girder 150 feet in length with- 

 out running the risk of inducing cracks in the masoni'y ; but the author, 

 havino- tried it, can say that the effects of temperature in such a case 

 may be safely ignored. Similarly he had used, without inconvenience, 

 girders G30 feet in length in the floor of a building, and purlines of equal 

 lencrth in the roof of a station. Each case must, however, be dealt with 

 on its own merits ; for in another instance, where a gasworks roof had 

 fallen down like a pack of cards, he had traced the cause of the accident 

 unmistakably to the expansion of the iron slate-battens, which were ex- 

 posed to exceptionally high temperatures at times. An engineer is apt 

 to consider he has taken sufficient account of expansion if he adopts the 

 conventional course of placing one end of his girders on rollers or sliding 

 plates, so that the bridge may be free to expand in the direction of its 

 lencrth. The superstructure, it is true, may be twice as wide as it is 

 lontr, but it must take its chance as regards expansion in the direction of 

 thewidth. Arched station-roofs, such as the St. Pancras roof of 240-feet 

 span and 700 feet length, seem to require no provision for expansion in 

 either direction. Again, the Victoria Bridge at Pimlico has longitu- 

 dinal girders 913 feet in length, and cross girders 100 feet in length, 

 with no provision for expansion in either direction. On the other hand, 

 at the Southwark Bridge, grooves had to be cut in the masonry to 

 admit of the free expansion of the cast-iron spandrils, which had 

 been fractured by expansion in several instances, though the arch was 

 only 240 feet in span, and the spandrils were not continuous across 

 the piers. However much the engineer may desire to eliminate stresses 

 due to expansion, he will find it impossible to do so, for the very ob- 

 vious reason that the temperature is not uniform at any instant, and 

 that even in this climate there may be a difference of 50 degrees between 

 the tempei'ature in the sun and in the shade. Take the case of an ordi- 

 nary lattice girder with trough-shaped top and bottom members and 

 double lattices. One side plate of the trough may be highly heated by 

 the direct rays of the sun, and the other be completely in the cool shade, 

 and one tie bar may similarly differ from its companion bar in tempera- 

 ture. If the girder were free to warp in any direction like a thin plank, 

 little stress would result ; but as it is not free to adapt itself to the un- 



