Don't know how I missed the entrance to the ?souterrain before, just inside the fence on the S side of the road at HY2706715113. Guess I looked for a channel further down whereas it starts as a hole like a badgers set (no sign of the masonry walls in the dark, should have used the camcorder's nightlight) literally a couple of feet from the actual fence behind three flat stones. I assume these are a few of the lintels from the intermittent roof of the built passage. This is described as 800cm square to begin with, though perhaps narrowing further on - it goes south-easterly to a ditch but after some 2.5m turns quickly eastwards and then the hillslope and presence of mud stopped investigations progressing to what they describe as the inner end (or else it goes ~100m down to the loch). By the western stone I could see a masonry wall of thin stones behind a grass fringe. Not wishing to be disturbed by the ploughman in the next field I did not enter for a more thorough going over. This side of the road the 'broken' mound is ascribed to outbuildings on account of projecting stones but it is obvious that the passage is within the tower's circuit - at The Howe we found a souterrain which had been struck through the tomb at the centre of the broch so the road might have demolished a northerly continuance (an odd coincidence otherwise).
Carry on up the Brodgar road past Buckan farm and you could easily miss the small sign on the right pointing out the Lyking road . No great distance from Lyking ( why does the road take such a strange turn there when it could easily have kept straight on it seems ) is the Stackrue site on the right . Oh I thought , another scrappy broch remnant . And so it appeared as I approached , only when I came up to it was the ruined splendour revealed as the other end a broch corner shone with large intact blocks still standing several courses high . Around it runs a large deep ditch possibly similar to that at the Broch of Gurness . At the resplendent end are a few standing stones and at one point against the bank several largish slabs . The rest looks like your average (?post- ) broch remains .
That a round steatite plate with a partial runic inscription was discovered here shows that like the Broch of Gurness and The Howe occupation continued till Viking times , explaining why the broch wall remains stand out from the rest .
In a massive thesis on early ecclesiastical sites there is a survey of this area that shows the passage 9 fenceposts from the stream and just inside the downhill field. There it is referenced as a souterrain, which is at least literally true. On March 9th finally found the entrance at HY2706715113.
John Fraser in the Proceeding of the Society of Orkney Antiquaries mentions a well near the broch's presumed centre. On the other side of the road the passage mentioned goes in an easterly direction, and he believed that it was heading for the burn like a similar passage at the broch of Nettletar.
RCAHMS NMRS record no. HY21NE 9 bears out my comparisons as they have a two-foot gap at the top leading down to a 1.5m high beehive cell that is equated with the well at the Broch of Gurness. And over the fence t'other side o' road is described a curving subterranean passage 800x800mm and at least 2.5m long (souterrain but not a souterrain?).
Coming towards the broch from Lyking, where the Burn of Lyking passes under the field wall (nice double bridge in drystane), I saw a short line of good stones along the bank running to the fieldwall at an acute angle. There appear to be remains of another bank wall over in the broch field section of the stream too. Going in for a closer look at these I saw that there is an earthwork with stones, taking the shape of a circular quadrant (HY27121514) rising from the burn, occupying the field corner this side of Lyking against a loop of the stream [it may be relevant that CANMAP shows mound of the Viking burial as having been in the next loop, north of this]. There appears to be an inner ditch below the top of the rise and the mound's arc appears to continue on the other side of the road, though no earthwork is shown there on the first O.S. map. In 1882 a ford is shown at this area of the burn and the subterranean passage is indicated downhill from this area.
Though this would appear to be a continuation of the broch settlement it is not presently recorded and may instead relate to a vanished mound from which a Viking burial came (RCAHMS NMRS record no. HY21NE 4). This is currently indicated on CANMAP at HY271152 but does not appear on the 1882 map like the Viking cists do (HY21NE 8 at HY27051532, but on the 1182 O.S. at HY26961532), unlike this earthwork.