Hmmm - I'm having trouble figuring how the focal length helps, as I'd think you need more independent variables. The focal length would need to be related to the negative size/film format to get an angle of view -- is anything known about that? There's also the possibility the image is cropped, which would argue for using proportions and ignoring the focal length.
Basically everything is tied together by some sort of proportion. You've got the width of the tops (presumably the same) of the two stone piers for one ratio. But is there any information (or at least educated guess) where the camera was positioned? Are there any old-time engineering drawings giving some dimensions of the pathway width or rail spacing or even the dimensions of the stone piers? Height and width are self-evident, but the depth needs some way to establish a reference. If it could be assumed the camera was approximately at the river's edge, I suppose that and the far edge that's visible could be used to guesstimate the away-from-the viewer distances.
Odds are the intermediate pier is halfway between the two sides of the river, although some peculiarities of the riverbed could no doubt shift that.
Anyway, this caught my eye because it's not too far from home (and I've always been intrigued by things mechanical and structural!)