Tree cover

A recent national tree survey established that most English trees are urban. Though there are patches of woodland in the countryside, and trees on some field boundaries, rural England is relatively treeless. Whereas town streets are often lined with trees, and they’re also well scattered across suburban gardens.

Maps based on the survey can be seen on Bluesky’s website. Write in a placename or postcode, and you can see an overview of yellow-highlighed trees in your area. It’s a landscape management tool. In the case of Oxford, patches of woodland stand out: Shotover, Open Brasenose, Bagley Wood, Wytham. Trees are scattered across colleges and suburban gardens. And they line many suburban streets, and parts of the bypass, and along the rivers. They’re especially thickly clustered on the east bank of the Thames from Christ Church meadows to Sandford, and on the Cherwell side of the University Parks. Also up Boundary Brook and the Lye Valley.

The Thames by Godstow, Iffley and Kennington Meadows

How does the modern pattern of tree cover compare with the past?

My guess is that in the eighteenth century, there would have been fewer trees overall than now. There existed the same woodlands; their peak lay rather further back: when Shotover was disafforested in the late seventeenth century it was already much diminished. There were trees in college grounds: in Merton’s parks; walks in Magdalen and Christ Church meadows.  But few other trees within the compact city centre. Perhaps, as now, some by city churches.

Probably there were trees along some, perhaps most of the rivers. Traces of poplars have been found along local rivers dating from Roman times. Religious houses, or grounds once occupied by religious houses along the rivers, may have maintained planted trees by rivers. Early maps are suggestive. A 1675 map of Oxford shows trees planted along most but not all of the rivers and streams, though not by the Cherwell at St Clements. In Taylor’s 1750 map of Oxford, there are trees regularly spaced along most river and stream banks, but again not all, suggesting in both cases that, though no doubt there’s a formulaic quality to what we see, it’s more than just a representational reflex – so in Taylor’s map not along the Thames in the St Ebbe’s stretch, or on the St Clement’s side of the Cherwell, or on the Cherwell as it comes down towards the Thames after Milham. Other maps based on Taylor’s show the same thing (no doubt they just copied Taylor’s trees).

As to fields around Oxford, early eighteenth-century prospects from the south show meadows or fields there parcelled out by hedges and scattered with hedgerow and other trees. But the town quickly gave way to open fields, probably relatively un-treed. The land was densely settled and cultivated, so when not woodland, or wood pasture, it was largely bare, and quite possibly barer than local farmland today. Once more extensive medieval tree cover may have been reduced also in non-forest places in the early modern period: thus the once large groves on the hill by Iffley.

A 1777 Christ Church estate map in the Bodleian map collection is suggestive here. The map shows the then parish of Cowley, extending from St Clements in the north-west to St James church, Cowley in the south and Horspath in the east. It marks both hedges and trees, though it’s hard to tell what’s formulaic. It shows Cowley Marsh with a fringe of hedges and a few, but not many hedges along other field boundaries. It shows ‘Lye Hill’ (between what’s now the Oxford Golf Course and Cowley Marsh) with a fringe of hedges dotted with trees. It shows strips of common woodland roughly corresponding to Open Magdalen and Open Brasenose, current Brasenose Wood. It shows some hedges along streets and the occasional tree, notably a tree in the street where Temple Road leaves the Oxford Road. A couple of plots around houses near St James are shown with trees, but others not. Overall not many trees!

A prospect towards Oxford from Ferry Hinksey by the local musician and painter John Malchair around 1800 shows meadows stretching away until trees hug around the city (perhaps along the Thames?). The second series OS map, a century later, suggests similar bareness. The map probably doesn’t show all the trees there were, let alone hedges, but it does show some patches of woodland (thus, on the left margin of the map below). Yet it shows little woodland in the hills above the Hinkseys: no equivalent of today’s Raleigh Park, where the growth of trees has almost obscured the view towards the city even from a height, nor does it show any clumps of trees on the meadows. (Tree cover varies across Malchair’s various renditions of this view, however – so it’s hard to be sure of the larger picture),

Hills west, meadows centre, town east, late nineteenth/early twentieth century (

Prospect from Raleigh Park, above North Hinksey

An 1892 photograph of beating bounds by the ferry shows trees dotted along the river bank  But now they’re denser than that around there (and the stream is scarcely navigable). The old causeway between Ferry Hinksey and Osney Mead runs through trees today.

For similar past bareness, see Malchair’s view from Shotover Hill, over what would be Headington and South Park today.

Or – better protected against charges of artistic licence — this 1891 photograph of a flooded Botley Road, at a point when building along it had just begun, and it lay largely open to meadows on either side, especially in this photo to the north. The road itself is shown as largely bare of trees.  Now much of the same, built-up road has tall suburban street trees dotted along it.

Nineteenth-century urban expansion may not originally have been good for trees: one imagines relatively bare new streets. But nineteenth-century suburbs favoured trees: the wealthier the suburban neighbourhood, the more trees. Trees appear in late nineteenth and early twentieth-century photographs as features of streetscapes: in 1922 there were trees by Catte street that are not there now   A 1914 photo shows large trees lining St Giles. The Plain 1922 is well planted with trees. Heading further out (and back in time), an 1890 view from Magdalen Tower over St Clements shows large trees dotted around near East Oxford. Tanis Hinchcliffe’s history of North Oxford regrettably has no entry for ‘trees’ but late nineteenth and early twentieth century photos in her book show them, eg. scattered around front gardens in Chalfont Rd.

In some cases we know when trees spread. In my post on the Jubilee Walk, I noted the tree-ing of Boars Hill as it developed as a district of grand houses. Now the trees have grown to such a point as to obstruct views from Arthur Evans’s Jarn Mound, which was intended to provide a prospect over the city.  The similarly grand-house district of Harberton Mead and Pullen’s Lane, on the slopes up to Headington, is equally heavily planted, with many majestic trees. In this case, the second series OS map shows some woodland in the neighbourhood, but not covering as extensive an area as now.

First series OS maps appear to show trees with obsessive concern for detail. How reliable they are I can’t say, but they would bear further study from this point of view.

First series (1876-81) OS map (from Percy Manning collection)

Evidence for what was the case around the rivers increases in the era of photography – at this point we start to be beholden less to cartographers’ choices and more to photographers’s notions of what made a good picture. Insofar as we can trust them, early photos suggests that trees were a common feature of river banks, adjacent to and further beyond the city, if in varying density. On the Thames, they were to be found not only by Christ Church meadows (a location often to the fore in photos of boats on the river), but also on the opposite bank and stretching down river. Thus 1860, 1922.

Heading up the river, this photo c1880 probably shows the stretch of river between the Botley Road and the southern end of Port Meadow. Here the trees on the bank look newly planted,  but there will no doubt have been felling and planting cycles  Trees are abundant in an 1895 photo on the west bank of the Thames at Medley, though, as now, Port Meadow looks treeless along its bank.

An 1880 photo labelled as showing the Thames at Binsey surely doesn’t show that, but a lesser stream/ditch, but that has pollarded willows along it, perhaps to help give structure to the banks.

Heading south along the Thames, trees are much in evidence in old photos, not always in exactly the same patterns as now. An undated image shows some huge poplars by the weir at Iffley  

The Cherwell appears yet more heavily treed, an arboreal idyll: 1906, 1912, 1922.

The Cherwell 2020

Aerial photos even from the post war period — accessible eg through Google Earth’s timeslider — suggest that some now treed areas adjoining rivers, the kinds of areas that showed up as densely treed on the national tree map, weren’t so in the past. Aston’s Eyot was bare (but, having been a city dump, had during the war been used for some military purpose); Rivermead had tree fringes but was less treed than today – a case where trees seem subsequently to have self-sown; tree cover has been thinned in recent conservation efforts. A belt of trees now masks Osney Mead industrial estate from the towpath walker; post-war photos suggest they didn’t exist when that was open land.

Focussing on rivers alone, before the age of photography, maps tell us more about the immediate neighbourhood of the city than further out. But they already suggest a management regime, or aesthetic, in which it was thought good to line the banks and immediate surroundings of rivers with trees. Hard to say how far around the city this pattern extended before the mid nineteenth century – or who would have done the planting further out.

I’ve been thinking about the tree cover issue for a while, but it’s engaging me especially now, as, in default of flowers, I’m turning my attention to trees.

Thanks to Stephanie Jenkins for allowing me to use the Botley Road flood postcard from her collection. In relation to early maps, I have linked to an online source, but have myself relied in the first instance on Daniel MacCannell, Oxford: Mapping the City.

Thanks to Nick Daisley for drawing my attention to the Christ Church estate map, and for noticing the tree in the street at the end of Temple Road.

St Clement’s

St Clement’s is by, but not whole-heartedly of the river. Ironic, then, that St Clement is the patron saint of mariners. (In the usual perverse way of such associations, this is because he was supposedly martyred by being tied to an anchor – so is traditionally portrayed with one).

Historically, the main focus of settlement in this district was close to the Cherwell, by the crossing, later a bridge: known as East or later Magdalen bridge. It’s an ancient settlement, certainly pre-Conquest. It’s hypothesised that it may have been home to a Danish garrison at the end of the tenth century: apparently they liked to site their garrisons with river protection on one side. And St Clement appealed to these Vikings (as also in St Clement Dane’s). The modest church that bore this dedication stood for many centuries in the area now occupied by the Plain, where houses around the church formed the settlement’s core, with further ribbon development spreading along ‘St Clement’s High St’, the route to London. Or at least, the route to London via Wycombe. The Cowley Road was once contrastingly known as the road to London via Henley. And one 1789 map, following the improvement of the Iffley Road, labelled the three main roads radiating out from St Clement’s the road to London, the old London Road and the new London road respectively. The settlement derived its function from its ambiguous relationship with the town of Oxford: it provided an entry point to and from the town, but its inhabitants were not bound by its restrictive practices.

The Plain, from the Iffley Road

Round the settled core, the parish’s fields fanned out: up to hillsides now occupied by Headington Hill Park and South Park to the north, then down the line of Divinity Road (Divinity walk as it was initially known)  – so encompassing much of what’s now the most developed area along the Cowley Road. From the point southwards it became messier. It’s hard to find good maps of pre-enclosure St Clement’s parish, and once you look at the 1853 tithe map, you can see why: because to the east and south the parish was something of a jigsaw, with a detached part to the east (in Cowley Marsh), while around the modern Iffley Road, chunks of St Clements were interspersed with chunks of Cowley. There may be chance reasons for that – to do with manorial or other jurisdictional or property boundaries. Alternatively, if the parish provided the unit within which common cultivation was organised (and the fact that enclosure took place with reference to parishes suggests that it did) then this may reflect efforts to ensure that inhabitants of neighbouring parishes each had access to certain kinds of useful land: marshland for peat-digging (something St Clement’s people may have wanted), lush pasture or meadow land (which Cowley residents may have wanted). From the account in the VCH, it seems that parishioners, following a similar logic, had some forest rights, presumably in Shotover, since they were ultimately completely bought out, in the late nineteenth century, by the owner of Shotover Lodge.  The fact that people had this range of rights doesn’t necessarily imply that they exercised all or any of them directly: they could choose to lease them or, if this could be organised, sell them, and over time the general trend was towards commodification. But some common fields remained, however cultivated, down to enclosure, in the mid nineteenth century.

St Clement’s didn’t occupy much of the river bank going south – at some point it gave way to the embrace of the large rural parish of Cowley: the tithe map suggests very quickly, though Mary Lobel in the VCH thinks that two medieval mills along the stretch south of the bridge fell within, or at least had some relation to this community (the parish was probably not the operative unit in this regard). There were no mills adjoining parish land north of the bridge – and no mills surviving into the modern period — until you got to King’s Mill, at the parish’s northern boundary.

During the Civil War, with royalist forces concentrated along this front, St Clement’s was the site of a major defensive bastion. Fighting caused much destruction here. But after the war, the community was rebuilt and expanded its footprint.

Later eighteenth and early nineteenth-century ‘improvements’ were more enduringly though selectively devastating.  When, under the Mileways Act of 1771, the space on the St Clements side of the bridge was opened up, part of the churchyard and some houses were removed to give better access to the road running between the city and Iffley. In effect, the core of the settlement was assailed. In 1829, the church itself was removed to allow further widening. A new church was built at the NW corner of the parish, at the corner with the Marston road, where it still stands, in some weathers (not all) looking gloomy and secluded.

The new St Clement’s on a fine November day: not so gloomy

On the site of the former churchyard (which was definitively reconfigured as a roundabout only in the mid twentieth century), there stands, obscured (though listed) a stone marking the supposed end of the Napoleonic wars in 1814 (supposed, because Napoleon staged a come-back and had to be defeated again at Waterloo). It now holds up the lamp post on the side of the roundabout opposite Sainsbury’s. Stephanie Jenkins reports a very similar stone in Carfax Tower (which at the time was still part of St Martin’s church), and suggests that this St Clement’s stone was probably also originally attached to the church, remaining when the church was relocated. Jackson’s Oxford Journal reports that after peace was proclaimed ‘a number of barrels of strong beer were stationed in different quarters of the town to regale all those who choose to partake of them’. Let’s hope that the existence of the stone implies that, though not yet part of the town as such, St Clement’s residents nonetheless got their share of beer.

The reduction of housing in the area around the Plain through to the early nineteenth century had the effect of making St Clement’s High Street – modern St Clement’s St – all the more the heart of the settlement.

In the 1820s, largely though not entirely low-grade housing started to be built to fill the space between the high street and the river. Grander houses making up ‘London Place’ at the northern end of the High St represent higher end development in the same period. John Henry Newman, who was briefly curate of St Clements and helped to raise funds for the construction of the new church), attributed low-end expansion to the tightening grip of colleges within the city, which had the effect of raising rents and forcing poorer residents out. (Plus ça change…) The extra-mural suburbs of St Ebbe’s and Jericho developed in parallel and perhaps in part for the same reasons. So gaps between the old city and the rivers were filled in – in ways that in the medium term proved bad both for the rivers and for those living on their banks.

London Place

The Mileways Act, which applied to the city of Oxford and the parish of St Clements, was a portent of things to come. In 1832, St Clement’s was included in the Oxford city parliamentary constituency; in 1835, within the new municipal incorporation. In 1848, WP Ormerod, writing about The Sanatory Condition of Oxford, said he would depart from the Registrar General’s units and include within ‘Oxford’ ‘that more or less continuous mass of building lying between Summertown, the toll-gates on the Witney and Abingdon roads, and the lower part of Headington Hill’: that is, an expanding eastern periphery of the city.

Yet the parish did not entirely urbanise. If anything, in the late eighteenth and early nineteenth century, the uphill part of the parish became increasingly distinct. Headington Hill Hall was built up on the hillside; those who bought out old manorial rights developed park and arable land; they and a few farmers dominated these stretches of open space.

St Clements 1830 and 1880-1914 compared: 1st and 2nd series OS . From University of Portsmouth/Vision of Britain (see end of post for full acknowledgement) and

When St Clement’s and Cowley’s common fields were both enclosed, by an act of 1853, this left open space higher up the hill intact: it was already largely owned in parcels. Down below, allocation of land to individual owners opened the way to yet more easterly urban spread. In 1868, a new parish was carved out of Cowley to cater to that urban population (alive and dead – a new churchyard was an important feature of the plan). It looks as if St Clement’s parish boundaries were rationalised at that point. A few years previously, Pembroke and Magdalen had sold ground in what seems, from the tithe map, to have been part of St Clements for the construction of a new Oxford Union workhouse. St Clements was not part of the relevant union, so it made sense for new parish boundaries to put that plot within Cowley St John. As an effect of that and other rationalisations, St Clement’s lost all but a few hundred yards along the Cowley Road. Its southern boundary instead tracked the edge of the open land– which extended further and more jaggedly to the south before the construction of Morrell Avenue. St Clements thenceforth consisted of its settlement round the Plain, along the Cherwell, then a stretch of parkland up to Headington Hill Hall, and the future South Park.

We might say that one continuity through all these changes was that the parish retained its hybrid urban/rural character. And the parish probably maintained some meaning in its residents’ lives, not just because of its (now relocated) church but also because it acquired its own national school in 1839 – though both were designed to serve Anglicans in a parish with many dissenters. Conversely, its powers in relation to its poor were reduced when the New Poor Law of 1834 placed it within a union centred on Headington (a union which covered the eastern perimeter of urbanity to which Ormerod had referred). Enclosure in the 1850s probably had little effect on how most residents made their living. But, whatever change in these respects did or did not mean, the larger relationship between the parish and the city clearly changed as, from being a small fringe settlement, it became a transition zone between the old town and urbanising east Oxford.

Nonetheless, to this day, it remains hybrid. When the wealthy on the hill sold out, in the twentieth century, the City Council bought Headington Hill Park and the Oxford Preservation Trust, South Park, so that both remain open. (In the turf of South Park it’s still easy to discern old ridge and furrow plough marks: ghosts of an older St Clements).

London Place and St Clement’s from South Park, with the spires and towers of the city beyond

In its latest city plan (looking forward to 2036) the City Council insists on the importance of retaining the open feel around the Marston Road and South Park, identified as historically significant inasmuch as they preserve a sense of old sightlines between parliamentary forces and royalist defenders. They suggest though that some low-rise development, insofar as possible out of sight, in the meadow land around the church would be compatible with this constraint.

Church Meadow: ripe for development?

Oriented as it long has been to the town across the bridge, to space on the fringe of the built up area, and to main routes heading variations on east, St Clement’s doesn’t, by contrast, seem in any major way to have been shaped by its relationship to the river – in some minor and indirect ways yes, but not directly and importantly. In what ways did the river impinge on it, and it on the river, and how did the relationship change over time?

The most mysterious and intriguing element in the story takes us back to the Danes. In 1886, some Viking stirrups were found on Angel and Greyhound island. They were extracted, so their precise original context can only be guessed. They have been suggested to derive from a Viking burial. But more recently it’s been suggested – in the context of the ArchEOx project – that they may have been votive offerings (as also possibly an Anglo-Saxon shield, found nearby). ‘Christianity was widespread by Æethelred II’s reign (978-1016), but some pagan practices recurred under Viking influence.… perhaps demonstrated by an increase in river offerings or ritual deposits… Common to both Britain and Scandinavia, these offerings were thought to have been a votive practice, giving thanks to the gods for success in battle or to ensure good luck in future battles.….A distribution of votive offerings seems to follow the movements of the Danish armies of 993-1016 along the Thames and Lower Severn valleys into East Anglia and Lincolnshire.’

An intriguing ritualistic use of the river. (Though some syncretism seems to be implied, if the same Vikings made votive offerings in rivers and dedicated a church to St Clement).

Thereafter, things become more prosaic.

If St Clements was – like Jericho and St Ebbe’s – a sponge for early nineteenth-century city growth, in contrast to those other districts, at no time, and least of all in the early nineteenth century (when all these suburbs were growing), were its fortunes importantly shaped by waterborne traffic. Unlike its western counterparts, it was not the site of wharves and warehouses, or of a significant water workforce.

This because the Cherwell hasn’t, or hasn’t in recent centuries, ever borne much river traffic. Charles I supposedly thought of developing it when Oxford was his HQ, but once the parliamentary forces were hovering on his eastern flank, there wasn’t much scope for that. There was probably into the recent past some local traffic: some portering of goods, cattle and people, in small boats or punts, from here to there across and along the river. In a 1907 novel, a bad punter was rescued by ‘An opportune waterman rowing down the stream [who] picked me out of the mud and set me on board again’. As this incident illustrates, when leisure punting became a thing, the Cherwell became, as it remains, a haunt of amateur punters.

Punt slipway, alongside Magdalen. Magdalen Water Meadows and Angel and Greyhound Meadow separate this braid of the Cherwell from St Clement’s

But if punts entered around Magdalen bridge, the entry point was probably always on the Magdalen side, a river braid or two away from St Clements. A 1900 guide to punting judged the stretch of water that ran along the back of St Clement’s ‘not particularly attractive, going by the back walls of houses’.

Furthermore, because the parish’s field system didn’t extend across the river, the river wasn’t an artery through it. The meadow across the river, now called Angel and Greyhound Meadow (after two St Clement’s pubs) belonged (and belongs) to Magdalen, who used to lease it out (as now, but now to the City Council). Perhaps their tenant needed his animals, or his workers, to have access from St Clement’s, across the few bridges that link the parish to the meadow. Perhaps locals used the meadow as a place to walk and take the air. So possibly some dawdling along the river banks, but not obviously much congregating around them.

There are one or two nineteenth century indications that those with property overlooking the river that location as potentially an asset. The proprietor of baths in what came to be known as Bath St (of which more in a moment), fantasised about a grand entrance for customers, from the river side, via a ‘water gate’. A library/reading room on the premises was to overlook the river. But it’s not clear that the baths as they operated included these amenities. Graeme Salmon, in his book Beyond Magdalen Bridge, points out that an 1850 map made by the City Engineer, Hoggar, shows a housing block in York Place bearing the name ‘Magdalen Prospect’, suggesting that the view to the meadow and beyond was a selling point. But the building doesn’t seem to have attracted a fashionable class of tenant. New early nineteenth century streets in St Clements were mostly constructed to lead down to the river, but don’t connect up with anything once there. The river terminates a series of cul de sacs. No effort was invested in making it a feature.

Boulter street, like many others, dead-ends at the river

Presumably the river was the source of the water in Bath Street’s baths – established (in 1827) at a point when the district was developing. These were Oxford’s first public baths. The envisaged clientele were adult males, not women or family bathing parties. Later, when the City had more confidence in the cleanliness of its rivers, and the social function of bathing had developed and changed, the river behind the church (that is, running past Magdalen’s ‘Church Meadow’) was developed as one of the City Council’s public bathing places. It remained in use (though probably, as with most such sites, increasingly unsupported use) until 1984.

One problem with use of water from the river – for bathing people, washing things, or indeed drinking – was that it was unclean, and, as a consequence of urban development, increasingly polluted, like the watercourses in Oxford’s other expanding nineteenth-century poor districts. Epidemiological mappers Ormerod and Acland agreed in identifying St Clements as one of the city’s fever/cholera black spots. Ormerod struggled to relate the prevalence of a range of infectious diseases to the dirtiness or cleanliness, and especially the smelliness, of particular streets and houses – a correlation he expected to find. Bath St he said was clean, but still instances of fever could be found there. He suggested that in Oxford overall it was disadvantageous to live in low-lying land near the river, possibly because, as flooded meadows dried, inhabitants were exposed to vapours arising in the drying process. Acland, more alert to the dangers of contaminated water, mapped (in dark triangles) the places at which sewers discharged into the river. He noted too that by the time it reached St Clement’s, the Cherwell had already been contaminated by ‘the drainage of parts of St. Mary Magdalen, Holywell [and] St. Peter in the East.’ The case for water supply being crucial was particularly well demonstrated by the case of St Clements, he thought, because after 1832, when it was badly hit by cholera and other infectious diseases, water for its pumps was supplied from up Headington Hill – and in later epidemics,  it was less hard hit.

St Clement’s on Acland’s map

Both water supply and drainage are better (if not impeccably) managed these days, and the cheapest housing has gone, though some of the more substantial older houses survive in St Clement’s dead-end streets.

More modern, still relatively modest, sometimes apartment housing has replaced most of it. Much of the area closest to Magdalen bridge has been developed as student housing: thus by Magdalen, Queen’s, Oxford University (graduate accommodation in Alan Bullock Close) and on a purely commercial basis by iQ.

I’m told that the 1960s City Council master plan for this part of St Clements envisaged a riverside path with landing points for punts etc., and, presumably in at least partial fulfilment of this vision, some of the student buildings, and ordinary housing, have been equipped with desultory river terraces, but these are badly maintained. Some lead nowhere, and seem unused. The district’s relationship to the river remains ambivalent, even abortive.

I would have made my life simpler had I started by reading Graeme Salmon’s chapter on St Clement’s in his Beyond Magdalen Bridge, which also includes a good selection of (post-enclosure) maps of the parish. But sometimes it’s good to learn the hard way. I also found Daniel MacCannell’s Oxford: Mapping the City useful once again. He is my source for the 1789 map referred to above.

This post incorporates [in the form of part of the 1830 OS map] historical material provided by the Great Britain Historical GIS Project and the University of Portsmouth through their web site A Vision of Britain through Time ( It’s used under this licence