Reading RCTS more carefully and looking at the Vulcan drawing the original 34 and 35 had the exceedingly unusual feature of an oval section boiler. This boiler had the long dimension vertically, which one supposes explains why there was room for railings round the footplate, otherwise only seen on the broad gauge. I'm tempted to speculate that perhaps these oval boilers didn't last very long, and that some time before 1866 these locomotives had been rebuilt with more conventional boilers, and when Armstrong came to rebuild the chassis he retained the replacement boilers, which would have been of more recent and more conventional construction.  If boiler and at least some motion components could be retained then the reconstruction looks rather more sensible. Just about everything in the chassis must have been discarded with even the wheels of a different size. 

I'm still struggling to fully understand the Vulcan drawing, but I'm also coming to think that at least the valve rods ran under the front axle rather than above.  I'm really struggling with the drawing as regards the piston rods. As far as I can see at the moment the pistons are directly in line with the leading wheel axle, which is surely impossible. I'm increasingly tempted to believe that the locomotives had a major reconstruction that hasn't been recorded, because I find it hard to believe that something so very unconventional ran for 12 years. 



 

I posted a link to the sketch on the GWR Elist and a fellow member suggests that the drawings show two piston rods per cylinder, one above and one below the axle, and (bearing in mind the possibility of pareidolia) I now reckon I see that too. Besides how else can it possibly have worked? But it doesn't seem to me that this setup would have worked, presumably adequately, for twelve years under Joseph Armstrong. I think it reinforces my (utterly without evidence) conjecture that they must have seen a significant rebuild fairly early in life, maybe even in S&C days, into something more conventional and more in the style of their renewals. Maybe even a smaller wheeled long boiler 0-6-0? But I don't know what sources there are for this era, what RCTS drew on for their research. And even if anything does exist in some register in one of the archives, can I face working through pages of impenetrable victorian copperplate  handwriting to find it?

The valve chest was under the cylinders, hence the odd, sloping bottom to the cylinder block. 

There was only one piston rod in each cylinder, but because the front axle was in line with the centre of the cylinder there was a yoke to take the drive past the axle. This yoke's top and bottom elements were round bars which must have worked in some sort of bearing, though this part is not clear on the drawing. This whole arrangement was a substitute for conventional slide bars. 

1 hour ago, billbedford said:

The valve chest was under the cylinders, hence the odd, sloping bottom to the cylinder block. 

 

There was only one piston rod in each cylinder, but because the front axle was in line with the centre of the cylinder there was a yoke to take the drive past the axle. This yoke's top and bottom elements were round bars which must have worked in some sort of bearing, though this part is not clear on the drawing. This whole arrangement was a substitute for conventional slide bars. 

Could you enlarge on that please? I'm having trouble working out what you mean. As far as I can see there is maybe a few inches, no more between the cylinder and the axle, so I don't see how there could be any kind of central piston rod or yoke external to the cylinder - I don't see where it can move.
  I've sketched my interpretation below, please could you clarify what I haven't  understood. 
It seems to me as if the valve (red in front elevation) is about 30 degrees off horizontal, as well as sloping downwards on the plane of the valve rod (very faint pink), so pretty much below.
The cylinder is in green, and I don't see there's room for any kind of yoke. We've certainly got the two round bars top and bottom, but my interpretation had them as piston rods. I think I see bearing surfaces for conventional slide bars on the crosshead (red)

I've produced a sketch of the as built condition. It took an awful lot longer than I had hoped. A lot of it was not finding it very congenial to work up, and a lot more was having a very  detailed but very indistinct works drawing to use as a source.
Livery is largely guesswork. I have a black and white shaded drawing which gives some clues, but little more. I'm choosing to believe that it was painted up something in the style of GWR Wolverhampton, the S&C having been the foundation of that line, but that's completely and utterly a guess. 

I'm still unsure of some constructional details, and cannot make sense of @billbedford's  comment re valve gear above. I don't think it really matters for the sketch though. I do remain unconvinced that they would have stayed in this configuration for all their twelve years of service, but there's no evidence otherwise in any source available to me, it's only surmise.  Its so easy to construct a grand theory, but I'm all too aware its easy to construct grand theories that then get utterly demolished when subsequent evidence comes to light. 

Update: 5/11/2023. I've found a more detailed description of these locomotives in E.L. Ahrons. The early Great Western standard gauge engines, The locomotive magazine No 260 1914, which amongst other things states the dome was painted, so I amended the sketch. Ahrons is quite positive about one of the oval boilers having been reconstructed with a circular profile in 1866, which would seem to torpedo my theory about an earlier reconstruction, although these were events from before Ahrons was born.

As a light hearted aside, whilst talking about another ex S&C locomotive in the same piece, Ahrons' dry sense of humour is evident. He records that the locomotive in question "had but a short career, the chief incident in which appears to have been a collision with an itinerant horse, as a result of which, No. 32 sustained damage to the extent of one broken split pin on the outside valve gear. Structural alterations subsequently required by the horse were not stated.". Thanks to Steamindex for the extract.