NS4274 : Linear mound near Milton Island
taken 16 years ago, near to Milton, West Dunbartonshire, Scotland

[By the end of 2010, the area shown in the photograph had been blocked off; it is as well that I took the picture when I did. As of 2011, there is no longer a railway bridge here; the tunnel below it has been filled in, so that the railway line is now simply at the top of an embankment: NS4274 : Cycle Route near Milton.]
This mound is, according to an archaeological report(*), aligned with the faint traces of a causeway (from the Roman period) across the River Clyde: see the latter part of Link
(*) For my own reservations about the causeway, see the link, given below, to my photograph of the supposed causeway on the southern shore of the Clyde.
[An aside: for such an indistinct mound, the report gives what seem to be very precise dimensions (10.97m and 10.36m, which come from the original DES journal articles). However, some quick figuring shows that the original measurements were, as I had suspected, simply given to the nearest foot (36 feet and 34 feet, respectively); when these were converted to metric units, too many figures were retained, giving a false impression of centimetre precision. Even the original measurements given to the nearest foot must only have been approximate, given how indistinct this feature is.]
On why the River Clyde, which is now navigable to Glasgow by shipping, could formerly be crossed on foot, see NS4273 : The Lang Dyke. For what has been described as the corresponding part of the causeway on the southern shore of the Clyde, see NS4273 : Supposed remains of a Roman causeway.
The course of the railway line can be seen on the left, and the background of the photo shows NS4374 : Sheep Hill, which is the site of a quarry (NS4374 : Sheephill Quarry) and of a vitrified Iron Age fort (NS4374 : The summit of Sheep Hill).
See Link
(at WoSAS) for the archaeological details; such archaeological reports described the remains as those of a causeway or, more specifically, of a "Roman causeway", and this interpretation found its way onto maps.
The author of this shared description disagrees, believing the visible remains to be merely those of a lateral jetty (indeed, the first-edition OS map of c.1858 clearly marks a jetty here), with the sheltering and current-slowing effects of the nearby Lang Dyke (a long training wall; see Link for details) making the remains better-preserved than they would otherwise have been. See NS4273 : Supposed remains of a Roman causeway for a longer account of the reasoning involved.
The fact that some of the land between the lateral jetties on the Clyde is known to have become silted up and grassed over (to the benefit of some land-owners) may account for the apparent continuation of this "causeway" for some distance inland from the shore; other nearby known jetties show similar continuations. It is worth adding that, in the time since I first questioned the nature of these remains, some maps that marked the site as an old causeway have stopped doing so.
The original reports of this site as a former causeway attempted to link it with remains of roads on both sides of the River Clyde. The archaeological reports (made in the 1970s) about the roads themselves are also problematic, for various reasons that are unrelated to the "causeway"; see Link (in a Geograph article), where the subject is discussed at some length.
