That’s the 1916 Murtaugh Bridge over the Snake River, near Murtaugh, Idaho. It was recored by HAER in 1980 and demolished three years later. The two main spans were 102 feet long and 16 feet wide; if it were two lanes wide instead of one, it might have survived longer. Or maybe not.
The HAER description calls it a Pratt truss, but I have some questions about that. The two main spans are each six panels long and the center two panels have diagonals both ways, as is common in Pratt trusses to deal with the minor moment reversals that pattern loading can cause. That leaves two panels on each side of the center, between the center and the supports, and in a proper Pratt truss the diagonals should be oriented in the tension direction: from the top chord closer to the support, to bottom chord closer to the center. But the heavy diagonals at the end panels run the wrong way. At the mid-river supports, that might be to better brace the pier, but why at the end piers? In other words, this sure looks more like a Warren truss to me. Most of the details are fairly standard, although the deck-truss configuration was less common than through-trusses.
The next picture was labelled “General view showing bridge in its setting during high river flows” so I’m gong to assume this was during a high, but not necessarily flood, stage. if you look at the rocks on the river banks, you can see higher water marks.
If this part of the Snake River floods badly, like a lot of shallow rivers in the semi-arid parts of the west, that might explain the single pier in the middle. Putting two piers on the banks far enough apart to completely avoid flood stage would have significantly lengthened the span, and putting two piers low on the banks would double the chances of scour or other water damage.
The most interesting detail to me, and the thing that truly marks the age of this bridge, is that the two main spans are independent at the center pier. You can see it in the first and fourth photos above: there are two steel frames resting on the concrete pier in the river (which rests on rock.) The most obvious reason for doing that is that the analysis of a simple span is significantly easier if the work is being performed by hand. There’s a secondary reason that failure of one span doesn’t cause failure of the other, but it’s hard to see the scenario where that would matter here, since the most likely cause of failure is high water and that would still affect both spans. A more modern design would be to have a single frame that supports a continuous truss; a still more modern design than that would be to have a continuous truss whose depth reflected the bending moment; and then for the really modern look we get to plate girders or post-tensioned concrete. The analysis and construction concepts for all of these options would have been readily understood at any time in the last century, but the available analysis tools and construction technology have changed a great deal.