Design a site like this with
Get started

The Pillars of Larkhall, Bath

What is the origin and purpose of the elegant stone pillars around St Saviour’s Church in Larkhall, Bath? This article describes the pillars in detail and examines their connection to the development and improvement of roads in 19th century Larkhall.

Walking along St Saviour’s Terrace and Road today, one first notices the elegant stone pillars, walls and railings of St Saviour’s Church. Then, particularly on the northerly side, other stone pillars can also be seen bordering the pavements. The pillars are connected by stone wall, showing the stubs of railings that have now been removed.

While walls, railings and paired entrance pillars are common in Bath, ornamental pillars are much less so. This article attempts to answer questions on the pillars (and their associated walls) in Larkhall. Why, when and by whom were they commissioned and built?

The Pillars

The locations of the pillars are marked on the map (Fig. 1) below. The first pillar (abbreviated P1) is located beside the pavement just below Eastville. David Crellin and Alec Wood have told to me that this pillar was originally much taller (perhaps 8 ft or so), until some 20 years ago when the middle section was removed after repeated impacts from commercial vehicles.

The second pillar (P2) is at the corner of 1 Grosvenor Villas, Claremont Road with St Saviour’s Terrace, and is now barely visible under vegetation. This pillar and the next four pillars (P3 to P6) are also located on the north side of the present St Saviour’s Terrace and St Saviour’s Road. As shown in the early photograph (Fig. 2, below), pillars P2 to P5 are in line.

The third pillar (P3) is at the southern corner of Holland Road with St Saviour’s Road and opposite the church. The fourth pillar (P4) is at the northern corner of Wallace Road with St Saviour’s Road. The fifth pillar (P5) is at the northern corner of The Havory with St Saviour’s Road (and obviously predates the post-1950 development of these houses). Finally, the sixth pillar (P6) is just to the west of the newsagent McColl’s. This pillar, P6, was not noted by Melluish (1941).1

There are paired ornamental “entrance” pillars at the west and east corners of St Saviour’s Church (C1 and C2), and other pillars at the rear corners.

Finally, there is the “outlying” pillar (S1) at the eastern junction of Beaufort Mews with the south side of St Saviour’s Road, also not noted previously by Melluish. This pillar is embedded in the garden wall of 1 Victoria Place, somewhat obscured by ivy and recessed from the boundary line in St Saviour’s Road.

Fig. 1. Ordnance Survey Map 1951 – pillars are marked with red dots. (Note path to allotment gardens at P6.)
Fig. 2. St Saviour’s Church, Larkhall from the west, c. 1852, showing P3 and P4. (Photo attributed to Frederick C Bird.)

Photographs of the pillars are shown below. All the pillars are of Bath stone and clearly old, but none are marked on Ordnance Survey Maps.

Features of the pillars

Pillars P1, P2, P3, P4 and P5 are all octagonal, but not fluted, and have a single level of gabled buttresses and octagonal pad-tops. Thus, these five pillars appear to be part of a coherent scheme.

In contrast, P6 (though unfluted with a single level of buttresses and an octagonal pad-top) is four-sided, not octagonal. Possibly, P6 was erected at a different time and by other hands than the other pillars.

Pillar S1 is fluted and taller, and in this respect differs from the scheme for P1 to P5. Also, S1 has two levels of gabled buttresses with a pad-top, and so overall it appears identical with the church pillars. The overtly raised plinth makes the pad-top level (horizontal) with the church pillars on the opposite side of the road. The ugly rubble base (rather than solid stone) raises the possibility that this pillar may originally have been lower. Additionally, S1 is recessed back about 1 m. from the wall and railings in front of Victoria Place (and presumed boundary line on St Saviour’s Road).

The Church pillars are octagonal, with two levels of gabled buttresses with trefoil head panels to each facet and an octagonal slab on top. The listing text2 attributes these pillars to John Pinch, Jnr, who also designed St Saviour’s Church. The pillars of the church are listed, but none of the other pillars (P1 to P6 and S1) are.

The walls are horizontal between P2 and P5. The wall and some pillars (P3 and P4, and probably P5) are evident on the north side of St Saviour’s Road in the photograph of 1852 (Fig. 2, above) – P6 being further away and downhill is not visible. The top of the general wall is not at the same level as the front wall of the church. The wall continues today, but with sections removed later to allow entrances to houses. In many places cement mortar replaces the original (and minimal) lime mortar.

Between P1 and P2, the wall slopes markedly downhill in a southerly direction, and is not horizontal.

The walls comprise three ashlar blocks (each about 35 cm high) and have an average height of around 105 cm. The bottom block is inset in the ground or has a low plinth to offset changes in slope. Between P1 and P6 the wall is made of half thickness ashlar (c. 17.5 cm) with a rounded top and no separate coping stone. This contrasts with the church walls of full thickness ashlar blocks (35 cm) and an elegant coping stone.

On the non-church walls, no original railings remain intact. Overgrown hedges often make observation of the top surface of the wall difficult, and sometimes impossible. Nevertheless, a distinct pattern of railing stubs can still be seen: a) between P1 and P2 only round stubs occur; between P5 and P6, also only round railing stubs occur; but b) between P2 and P5, round stubs alternate with square profile stubs in the walls. It is not known whether the square pillars preceded the round, or vice versa. A possible explanation could be that the original railings of P2 to P5 were square, but were replaced with round profile railings some 10 years later, when the Claremont Road railings were first erected. Whatever, the final railings were removed in 1943 during World War Two.

Fig. 12. Round railing stubs on wall in Claremont Road, Larkhall.
Fig. 13. Alternate round and square stubs on wall in St Saviour’s Terrace, Larkhall.

When were the Larkhall pillars built?

Crucially, there is at present no hard evidence to determine the dates when any of the pillars and their associated walls were built. However, some reasonable speculation is still possible if it is assumed that the building of pillars and boundary walls requires the position, width and level of the associated pavement and road to be firmly established. On this basis, the pillars and walls on St Saviour’s Road and on Claremont Road were completed in two separate stages and some 10 years apart (c. 1848, and September 1858).

St Saviour’s Road

From the photograph of St Saviour’s Road, the pillars P3, P4 (and probably P5) with associated walls were already in position by 1852. Indeed, the location of P3 and P4 at the corners of the roads beside the church, suggests they were built with firm knowledge of the church plans, or after 1832 when the church was finally erected. In June 1848,3 the parishioners of St Saviour’s Church considered a proposal to improve the roadway leading from St Saviour’s Church to the road near Beaufort House (for which purpose ground had been generously offered by Mr. F. Dowding). This need for land (probably opposite Victoria Place) suggests the proposed improvement probably involved widening (and straightening) the existing carriage road. In turn, this would have allowed pillars and walls to then be erected with confidence.

Claremont Road

A newspaper account in 18594 mentions vandalism to the top of a stone pillar in the new road leading from St Saviour’s Church to Beacon Hill, belonging to Messrs Dowding (i.e. the lower part of Claremont Road). If this was pillar P1, rather than P2, then this suggests a build date before 1859.

A plan to widen and level the road from Camden Place to St Saviour’s Church had been mooted in 1836, but little or nothing happened. When the proposed new road was discussed at a parish meeting in April 1845, the existing road was described as “a little dirty crooked lane about 6 feet wide (on and over which Sir Henry Rivers would construct a road 30 ft in width). In June 18485 it was stated at a parish meeting of St Saviour’s that F. Dowding, Esq., had generously offered ground for the purpose (presumably on the east side of Claremont Road), but another 10 years passed before success was achieved. The necessity for constructing a road from St Saviour’s towards Lansdown and Camden Place was raised at a meeting of the Bath City Act Committee in 1858.6. A week later, the Surveyor said the City Act Committee would take on the work, which he thought could be done for £129, but the committee would still have to decide whether the improved road would have a kerbed footpath and pitched gutter on each side. Finally, in September 1858, the Bath Chronicle recorded7 that the road from near St Saviour’s Church to Bay Farm (on the road to Lansdown) was at last completed and was a public road.

The mention of Mr. F. Dowding in this latter newspaper account makes it likely that lower Claremont Road (bordered on the east by P1 and P2) was also widened at that time. Consequently, it would seem to have only been prudent and practical to erect the pillar P1 and the wall down to P2 in 1858.

Attention has been previously drawn to pillars P6 and S1 – on account of their atypical features (and position in the case of S1). Unfortunately, there is no evidence available to determine their actual build-dates. P6 is out of line from P2 to P5, is not on a corner, and is unrelated to the site of the church. Also, P6 does not mark the extreme easterly boundary of the land of Upper Furlong.

S1 is the only pillar on the south side of St Saviour’s Road, and was not identified by Melluish. As noted earlier, its fluted design very closely resembles the fluted (but much less weathered) pillars of St. Saviour’s Church. The plinth is raised on poorly maintained rubble blocks. The original height of this pillar might well have been lower, and it might also be older than the pillars of the northerly side of St Saviour’s Road.

Functions of the pillars and walls

There are several possibilities:

Firstly, the position of many, but not all, of the pillars appears predictable. Thus, P1 was the north-western boundary of Dowding’s substantial holding of land (some 12 acres) in the Upper Furlong, while P2 to P4 marked corners of roads (or embryo roads) on the southerly boundary of much but not all of the Upper Furlong (see Fig. 14, below).

Fig. 14. Thomas Thorpe Map (1740). Upper and Lower Furlongs coloured green. Both parcels were acquired by Charles Gunning and Daniel Tanner in 1792, later by John Tanner (1767-1823), penultimately by his sister Ann Tanner (1775-1837), who finally bequeathed them to Frederick Dowding (1796-1861). Roads in yellow.

Overall, the ornamental nature of the pillars may reflect the important civic status of the landowner, Alderman (later Mayor) Frederick Dowding. The likely date of the erection of many of the pillars along St Saviour’s Road just preceded Dowding’s election as Mayor of Bath.

Walls are used traditionally to mark land boundaries. They also allow soil retention, which has been particularly important in controlling the sloping ground down to St Saviour’s Road, so preventing erosion and land slippage onto the improved new pavement and road. A retaining wall also allows better and safer use of the retained land.

However, this suggestion also requires P5, P6 and S1 to be accounted for:

  • Perhaps P5 (at the later-built Havory) was only ever intended to enhance the view, particularly from Victoria Place opposite. Unusually, the rear elevations of nos. 8 to 17 Beaufort West are all or mostly ashlar, suggesting the aesthetic appearance of St Saviour’s Road was important.
  • P6 might originally have defined the boundary for commercial and non-domestic developments in St Saviour’s Road near the Larkhall crossroads. Indeed, this area for future commercial development is delineated in the Harcourt Masters Map of 1795,8 so P6 could have marked a boundary from an early date. These non-domestic developments were a linear building (perhaps a workshop and the initial parish school), and the shops and pub of Lambridge Buildings. Indeed, No. 1 Lambridge Buildings (aka the White Lion Inn) was built and licensed before 1832.
  • Pillar S1 is the only one on the south side of St Saviour’s Road. As noted earlier, its fluted design very closely resembles the fluted (but less weathered) pillars of St. Saviour’s Church. a) If, as I favour, S1 was built earlier than the church, it would presumably have been by one of the Tanner family. Such a pillar might have had symbolic and/or practical significance c.f. Beaufort House (built c. 1798), where a pair of ceremonial pillars at the rear was erected by the then Commissioners. Also, S1 might have marked a significant stage in the building of Beaufort East and provided a template for the subsequent church pillars. However, while the weathering of S1 is compatible with an earlier date, it could also just reflect poorer quality stone. b) If built after 1832, the fact that S1 is directly opposite the church and matches one of its entrance pillars may have been of aesthetic significance in terms of giving more visual prominence to the church.

Secondly, Crellin has suggested (personal communication) that the non-church pillars and their associated walls in Larkhall relate essentially to St Saviour’s Gardens (the undeveloped land left when the Worcester Square project was abandoned). Thus, St Saviour’s Gardens began after the church was built in 1832, and continued as an ever decreasing area of land until at least 1946. Indeed, the OS map of 1886 suggests a pillar at the position of P6 would have marked a major entrance to the Gardens from St Saviour’s Road. However, this suggestion does not specifically account for the position of P5, or the position and function of S1.

Thirdly, Melluish originally suggested that the pillars and walls of Larkhall marked the boundaries of the planned but never built Worcester Square scheme. P1 and P2 and their connecting walls are on the line of Claremont Road, and mark the western boundary of both the original Upper Furlong fields and of the intended Worcester Square. However, as a specific purpose, the association with Worcester Square may be a coincidence in view of a) the late build-dates (discussed above); and b) the general impracticality of building pillars and walls before roads and boundaries have been finalised. In particular, erecting P1 and walls in Claremont Road for a building project that never started seems unlikely, impractical and imprudent, as there was initially no road, and later a narrow and dirty lane until 1858. However, P1 is still a general boundary marker both for the extreme north-west corner of the original land of Upper Furlong as shown in Fig. 14, above, and P2 is a natural corner. Also, P3 and P4 which mark the corners of the culs de sac beside St Saviour’s (embryos for the later Wallace and Holland Roads) seem unlikely to have been erected at these sites, decades before St Saviour’s Church was conceived. Melluish had suggested that P3 and P4 marked the proposed entrance to Worcester Square (estimated to be some 28 yards from Harcourt Masters’ map of 1795), but the actual measured distance between P3 and P4 is some 78 yards. This seems too large to have marked an intended but never built grand entrance from St Saviour’s Road. Any connection may be a coincidence. It seems unlikely that any of the pillars and walls of Larkhall were specifically intended to mark the intended, but unbuilt, Worcester Square.

Sources and Notes

  1. J. G. Melluish (1941) Grosvenor might have been different – Worcester Square project. Bath Chronicle, 27 Sept, page 8. The relevant text reads: As originally planned, this block of buildings, known as Worcester Square, was to have had a grand entrance leading up to it from the centre of Grosvenor Place. In that event, the erection of Beaufort Buildings, both east and west, would not have taken place. The boundaries of the proposed ‘Worcester Square’ were outlined, as may be seen to-day. The land its roadway edges was enclosed by walls and railings, and at its corners pillars were erected. These stand at the top corner at Deatsville [P1], the lower corner of Grosvenor Villas [P2] and opposite Beaufort Place [P5] towards the Larkhall end of St Saviour’s Road. The entrance pillars [to Worcester Square] were erected at the end of St. Saviour’s Terrace [P3] and continued on the farther side of the church at the corner of Ellenborough Villa [P4].
    St. Saviour’s Church – – Its situation appears to have been the entrance to the suggested Square, which was to have led up to it from the centre of Grosvenor Place.
  2. Listing details for St Saviour’s Church wall, piers and railings (15 October 2010, Source ID: 1394887, English Heritage Legacy ID: 5102960).
  3. Meeting of St Saviour’s Parish to consider a plan to improve the roadway from St Saviour’s Church to the road near Beaufort House. Bath Chronicle, 29 June 1848. St. Saviour’s Parish.—The text reads: Yesterday evening, a meeting of the parishioners was held at the National School-room, Larkhall, for the purpose of considering a proposal to carry into effect a plan for improving the roadway leading from St. Saviour’s Church to the road near Beaufort House. The Rev. Mr. Campbell, curate of the parish, was in the chair.
    Mr. Sims, – – – stated that F. Dowding, Esq., had generously offered the ground for the purpose. Mr. Sims also said that the Commissioners would contribute the sum of £20 towards the expenses; and that about £40 more, raised by subscription, would probably be sufficient to meet the whole cost of the improvement, which would be a great public accommodation. A committee was accordingly appointed to collect subscriptions, and to see the proposal carried out. On the motion of G. Adams, Esq., a vote of thanks was unanimously passed to Mr. Dowding for the gift of the ground, and other acts of liberality to the parish.
  4. Vandalism of stone pillar in new road (Claremont Road) belonging to Mr. Dowding. Bath Chronicle 28 July 1959.
  5. Improving the road from St Saviour’s Church to the road near Beaufort House, London Road – St Saviour’s Church Parish Meeting, Bath Chronicle. 29 June 1848.
  6. Improving the road from St Saviour’s Church to Camden Place and Lansdown. Bath Chronicle, 22 April 1858. The text reads in part: –BATH CITY ACT COMMITTEE – Mr. Dallaway stated that, at present there was only a dirty lane leading from St. Saviour’s to Lansdown; – – – In the discussion which ensued, it was stated that the Committee had surveyed the road sometime since, and had come to the conclusion that they had no objection to take it as soon as it was put in proper order. — – another amendment was carried on the motion of Mr.Cox to the effect that the services of the Surveyor be allowed to the gentlemen interested in the road, with a view to its being properly constructed.
  7. New Road from near St. Saviour’s Church to Camden Place and Lansdown. Bath Chronicle, 23 September 1858. The text reads – The perseverance and liberality of a few gentlemen at Grosvenor, aided by Sir James Rivers and Mr. F. Dowding, have last completed and thrown open to the public one of the most interesting roads that have been formed in Bath for many years, leading from near St. Saviour’s Church to Bay Farm, on the road to Lansdown, and the Town Council having accepted it on the part of the public, it will be free for ever. We now hope Sir James Rivers will extend his liberality, by continuing what has been done to the extent of his property at Tything Lane, which will open up some of the most charming scenery in the neighbourhood of Bath.
  8. “Plan of the City of Bath” by C. Harcourt Masters (1795).
  9. The land of Upper and Lower Furlong was for many generations owned by the Hayne family, and then inherited by John Bragge, Esq., of Ladborow, Devon (nephew of James Hayne). In February 1791, John Bragge’s son and heir (another John Bragge) sold the land to Charles Gunning, a gentleman of Bath and Daniel Tanner, a master builder of Bath. At that point, development became possible. After Charles Gunning’s death, Daniel Tanner acquired sole title, and after Daniel Tanner’s death in 1802, his son John (1760-1823) acquired Upper and Lower Furlongs. In turn, after John’s death, John’s youngest sister Ann Tanner (1774-1837) inherited the land, and her Will left her land interests to Frederick Dowding (1795-1861). He was a Bath solicitor, J.P., Alderman and Mayor of Bath in 1849 and 1850.After Ann Tanner’s death, Dowding and his Trustees developed the domestic housing on Upper Furlong. However, the commercial development at the easterly tip of Upper Furlong (Lambridge Buildings) started in Ann Tanner’s time and after her death in 1837, was continued by her heir, Frederick Dowding.


The patience and expert help of the staff in the Bath Record Office in providing maps is gratefully acknowledged. My thanks also to Stella for helpful discussion and patient acceptance of my preoccupation with local history; to Susannah Straughan for commenting on and editing the manuscript; and to Chris Straughan for expert help with image editing and illustrations.