That’s the Victoria Bridge over the Saint Lawrence River at Montreal, as seen in 1901. The river is quite a bit wider than the main channel, leading to the bridge configuration seen here: a long (close to two miles) low series of relatively short truss spans with a single higher and longer span over the main channel. It’s not going to win any beauty contests, but it’s a practical bridge still serving its original purpose of carrying both rail and road, and there is a certain machinelike charm in the repetition of the Pratt trusses.
The superstructure is steel and completed in 1897, which was Queen Victoria’s Diamond Jubilee year, hence the name. The piers were constructed in the 1850s for the first bridge on this site, which had a similar layout and was built for railroad use. That bridge had wrought-iron boxes rather than trusses and was designed by the Robert Stephenson, one of the most famous engineers in the world in the middle of the nineteenth century. Stephenson’s presence and the use of tubes are really two aspects of the same fact, because he had pioneered the use of tubes for railroad bridges. He designed the Britannia Bridge and the Conwy Bridge in the 1840s using wrought-iron boxes as the main structure. Those designs were based in part on materials research by William Fairbairn and Eaton Hodgkinson, whose worked contributed to the use of cast iron, wrought iron, and steel in the US as well as the UK. Here’s Stephenson’s bridge at Montreal:
The trains ran inside the boxes, making the bridge effectively an elevated tunnel. Since the trains were all pulled by steam engines, the small amount of ventilation provided by the windows visible in the box sides was probably not enough to make the trip pleasant. There’s not a lot of information readily available about why the tubes were replaced by trusses after less than 40 years in service – which is a polite way of saying that the old bridge did not fail catastrophically – but I’d guess that the internal clearance was a bottleneck, because trains had increased in size so much in that period. It’s possible that it was not capable of carrying the weight of modern trains, but I doubt that was true.
The tube design had its drawbacks, with expense, corrosion from the pollutants in coal smoke, and general unpleasantness during the crossing at the top of the list, but it was very strong. All truss and girder bridges can be modeled as beams, and the box layout was very strong, particularly with the cellular top and bottom sides (the flanges of the beam) that Stephenson used. In beam design, you want to push the mass of material away from the neutral axis (the midline of bending) and the box did that quite efficiently. Trusses do the same thing, with top and bottom chords that are significantly heavier than the web members. It might seem like the big “holes” in a truss reduce its strength and they do, but not very much. It’s the material at the top and bottom that does most of the work in bending. So switching from a box to a truss meant saving a lot of metal material, and saving on the total number of rivets required, but increasing the complexity of design, fabrication, and connections. Given that the vast majority of nineteenth-century iron and steel bridges constructed after Stephenson built Britannia Bridge were trusses, the economics seem to have spoken.