The Battle of Fulford ... at York Racecourse

Introduction

The Battle of Fulford was fought on 20th September 1066 between an alliance of Norse and Flemish invaders against an alliance of Mercians and Northumbrians. The former were led by King Harald Hardrada and Earl Tostig Godwinson, the latter by Earls Edwin and Morcar. It is the poor relation of 1066 battles, covered in just two pages in Frank McLynn’s ‘1066: The Year of the Three Battles’, one page in A H Burne’s ‘Battlefields of England’, and less than a paragraph in Smurthwaite’s ‘Complete Guide to the Battlefields of Britain’. Yet it had a crucial impact. If the English had been victorious at Fulford, Harold would not have been distracted from the defence of the south coast, which would probably have turned the finely balanced result at Hastings.

Fulford was a resounding victory for the invaders. No physical evidence of the battlefield location has ever been found and the contemporary account clues are few and vague. Several battlefield locations have been proposed. We think they are wrong. In this paper, we will explain why we think the battle was fought on a strip of land between the Knavesmire and the River Ouse at York racecourse.

The traditional battlefield location

The two detailed accounts do not mention the battlefield’s location. It is known instead from two minor accounts: Symeon says the battle was ‘apud Fulford’ where Latin ‘apud’ means ‘at’, ‘by’, ‘near’, ‘towards’; Gaimar says: ‘A Fuleford se combatirent’, ‘at Fuleford they fought’.

Nineteenth and twentieth century historians usually say that the battle was at ‘Gate Fulford’ and refer to it as the ‘Battle of Gate Fulford’. There is no supporting evidence and Gate Fulford did not exist at the time of the battle. The notion can be traced back to Augustus Freeman, the renowned Norman Conquest scholar, who states without any supporting evidence: “On the spot known as Gate Fulford, about two miles from the city, the armies met.” Freeman’s towering reputation ensured that his battle narrative and his proposed battlefield location held sway for the next hundred years. To be fair, it does match some clues.

Heimskringla (Palsson) describes the initial Norse troop deployment: “One flank reaching down to the river and the other stretching inland towards a dyke where there was a deep and wide swamp full of water”. So, the armies faced each other between the Ouse and a ditch. There was marshland beyond the ditch. The adversaries were drawn up in parallel shield walls, Hardrada and Edwin leading the flanks towards the river, Tostig and Morcar leading the flanks towards the ditch. This is depicted (Figure 1) by Kelly DeVries in his 1999 book about the Norse invasion.

Figure 1: DeVries's engagement scenario at Gate Fulford

Heimskringla (Hollander) describes the initial engagement: “Then the earls proceeded down along the ditch. There the wing of the Norwegian army extending to the ditch gave way, and the English followed them up, thinking that the Norwegians were about to flee. That part of the English army was led by Morkere.” So, Morcar’s flank forced back Tostig’s flank.

Heimskringla (Hollander): “But when King Harald saw that the battle array of the English had come down along the ditch right opposite them, he had the trumpets blown and sharply urged on his men to the attack.” So, Morcar’s flank advanced level with Hardrada’s on the other side of the battlefield, allowing Hardrada to cross the battlefield and attack them from the side. Morcar’s men were squeezed into the ditch and the marsh, many to their death. The rest of the English army fled to Jorvik.

Figure 2: David Cooke’s battle scenario

These events are portrayed (Figure 2) by David Cooke in his 2006 book ‘Yorkshire Battlefields from the Romans to the Civil War’. He wrongly depicts Edwin’s flank being forced back instead of Tostig’s, but the consequence is the same: Hardrada’s flank was able to attack Morcar’s flank from the side and force them into perilous terrain.

But the battlefield descriptions do not match the geography at Gate Fulford. Figure 3 transposes DeVries’s diagram to a heat relief map with sea level regressed to 1066. For one thing, there would have been 200m or more of marshy estuary, locally known as ‘ings’, between the armies and the Ouse. For another, there is no ditch or marshland to the east of the traditional battlefield. Walmgate Stray – shown as fenland to the east of the modern barracks (B) – is flood prone but it was usually firm enough to graze cattle. If Walmgate Stray would usually support cattle, it would usually support fleeing men. Even if it was temporarily flooded at the time of the battle, the English could easily have fled around it on better ground to the east of the supposed contact zone.

Figure 3:  Traditional battlefield at modern Fulford with Margary 28c and relief regressed to 1066

One kilometre north of the traditional battlefield, around the modern barracks (B), the terrain is a better fit for the contemporary account battlefield descriptions with relatively narrow ings and Walmgate Stray immediately east. But there is still no ditch, and it is implausible that the English would have marched over the moraine ridge and then down onto the flood plain. They would clearly have made their shield wall on top of the ridge, near the modern cemetery (C). But if the English were at the modern cemetery, the battle would not have been anything like that described in the contemporary accounts.

The traditional ‘Gate Fulford’ battlefield location looks wrong. In addition to its lack of ditch and marsh, we list some other deficiencies below that apply equally to the new orthodox battlefield.

The new orthodox battlefield

Back in the 1970s, Kenneth Penn used the reasoning above to reject the traditional battlefield location. He proposed instead that the battle was fought 500m south of the traditional location at Germany Beck. His theory was developed into the currently accepted orthodox battle narrative by Professor Forrest Smyth Scott, then Charles ‘Chas’ Jones.

Their point is that that a ditch was a crucial feature of the battle landscape and the only significant ditch near modern Fulford was Germany Beck. Moreover, it seems to match one translation (Laing’s) of Heimskringla’s description of the initial Norse troop deployment: “The one arm of his line stood at the outer edge of the river, the other turned up towards the land along a ditch; and there was also a morass, deep, broad, and full of water.” Jones, not unreasonably, interprets ‘turned up towards the land’ to mean that the ditch was a tributary of the Ouse and roughly perpendicular to it. Germany Beck matches this description . Laing’s translation also says that the Norse army was deployed ‘along’ that ditch, giving the basis for Jones’s theory that the battle was fought across Germany Beck.

Jones describes the major battle events in the ‘Battle at the Ford’ section of his 2010 revisionist battle narrative ‘Finding Fulford: The search for the first battle of 1066’. He gives therein some other reasons to believe that the Battle of Fulford was fought at Germany Beck:

  1. It matches the battlefield geography descriptions in the contemporary accounts – see ‘New orthodox battlefield geography’ below.
  2. Remnants of medieval iron recycling have been found on the southern bank of Germany Beck. If that iron included weapons or armour, it could be evidence of a nearby battle.
  3. Five Roman sarcophagi have been found in the area. He says about the first two: “in 1835 and 1836 workmen quarrying for gravel located two stone coffins in an area of land known as the ‘Nunneries’ or ‘Nunners’ fields’. The early OS maps show several gravel workings along the Ings [i.e. east of the Ouse].” The next two (quoting MAP): “During excavations for York sewer in 1892, two Roman coffins were found near St Oswald’s church containing human skeletons”. The fifth: “In 1997 during work on the A19/A64 interchange: A gritstone sarcophagus was encountered during the machining of a drainage service trench (SE 6135 4790).” Sarcophagi are associated with high status Romans who often lived near communications highways, such as Roman roads. Jones reckons they infer that an unpaved Roman road ran near the route of the A19. He proposes this road crossed a tidal part of Germany Beck on a low tide ford, and that this ford was the origin of the name ‘Fulford’. If all this is right, it is consistent with the battle having been fought at Germany Beck.
  4. The English fled to Heslington. Jones says on battleoffulford.org.uk that: “one version of the Anglo-Saxon Chronicle records them [the English] as fighting their way to Heslington, a mile away along the Beck”. Heslington is northeast of Germany Beck, the direction in which they would have fled if the battlefield was at Germany Beck.

In our opinion, all Jones’s evidence is unreliable. We will return to the geography in the next section. As for the other evidence:

  1. Laing’s  translation of the initial Norse troop deployment is an aberration. For unknown reasons, he invented ‘turned up’. Heimskringla says nothing of the sort. His translation of the initial troop deployment is also faulty. The Old Norse phrase that Laing translates as ‘along a ditch’ is ‘at díki’. Old Norse ‘at’ means ‘to’, ‘towards’, ‘at’ or ‘by’, not ‘along’. The normal Old Norse word for ‘along’ is ‘með’. Snorri Sturluson, Heimskringla’s skald, knew the difference because he used ‘með’ in the second Heimskringla sentence above to describe the English army advancing ‘along the ditch’. Other translators are more accurate. This is how they handle Heimskringla’s sentence about the initial Norse troop deployment, and its equivalent in Morkinskinna:
    • Heimskringla (Smith): “One wing stood forth on the river bank, while the other went farther inland towards a dyke, where there was a deep wide marsh full of water”.
    • Heimskringla (Binns): “One wing by the river, the other by the dyke, a deep swamp full of water”.
    • Heimskringla (Palsson): “One flank reaching down to the river and the other stretching inland towards a dyke where there was a deep and wide swamp full of water”.
    • Heimskringla (Hollander): “One wing stood on the bank of the river, the other was arrayed further up on land, and extended to a ditch. There was a swamp, deep and broad and full of water.”
    • Heimskringla (Finlay): “One wing of his formation was stationed forward on the bank of the river, and the other extended up inland to a certain dyke. It was a deep fen, broad and full of water.”
    • Morkinskinna (Andersen & Gade): “one wing toward the river and the other toward the still water of a deep pool”.
    So, Heimskringla and Morkinskinna are saying that the Ouse and the ditch were roughly parallel, and that the armies were arrayed between and roughly perpendicular to them. This is quite unlike Jones’s proposed Germany Beck battlefield.
  2. There is no evidence that the iron being recycled at Germany Beck was weapons or armour. Even if it was, there is no reason to believe that it was being recycled at the battlefield. Evidence of a paved Roman Road (probably RR803x, also known as RCHME 1) has been found a few hundred metres upstream. It is quite plausible that weapons and armour from the battle, even if the battlefield was miles away, was being recycled at Germany Beck because the road made it easy to bring wood for charcoal.
  3. The first two sarcophagi were found at Masham, 30 miles northwest of York, so they are irrelevant. Roman coffins were found near Clementhorpe Nunnery, which might have been known as the ‘Nunneries’ to match Jones’s statement, but they were on the Ouse west bank. A Roman coffin lid was found near Old St Oswalds on the Ouse east bank, but this and the fifth sarcophagus were on the Ouse riverbank and were therefore more likely to have been associated with the river than with a Roman road.
  4. None of the Anglo-Saxon Chronicle recensions say that the English fought their way to Heslington.

As Jones himself points out: “absence of proof is not proof of absence”. Just because there is no evidence that the battle was fought at Germany Beck does not mean it was not fought there. We will return to this after looking at the new orthodox battlefield geography.  

The new orthodox battlefield geography

Jones believes that the major military events are a good match for the geography at Germany Beck. He depicts these events in a series of diagrams on his battleoffulford.org.uk website. They are reproduced below. Note that north is to the left on these diagrams. His battle narrative is this:

  1. The Norse army marched north from their camp at Riccall towards Jorvik on an unpaved Roman road. It is the black line appearing on the right of the diagrams below.
  2. The English army marched south from their garrison at Jorvik towards Riccall on the same road.
  3. The armies lined up on relatively high ground either side of the beck. Germany Beck is tidal, depicted by blue hatching. The tide was in at the initial engagement.












     
  4. The tide ebbed. Earl Morcar attacked near the ford. Most of his men came down off the higher ground into the marshy tidal valley.












     
  5. Hardrada sent his men across ‘the delta’ - Jones’s name for the boggy ground at the confluence of the Ouse and Germany Beck - trying to get behind Edwin’s flank.












     
  6. Edwin’s flank fell back: “Once enough of King Harald’s men had crossed the beck’s delta, Edwin had to fall back”.













     
  7. Hardrada’s men crossed the battlefield to attack Morcar’s flank from the side and behind. Jones says the first hint of their fate came: “when those behind them were felled by Norwegian warriors”. Morcar’s men fled southeast into the beck and the marsh. Many drowned. Edwin realised that the game was up, so the rest of the English army fled back to Jorvik.





     

 

In our opinion, the geography at Germany Beck is entirely inconsistent with the battlefield described in the contemporary accounts.

  1. Heimskringla (Palsson this time) describes the Norse initial troop deployment: “One flank reaching down to the river and the other stretching inland towards a dyke where there was a deep and wide swamp full of water”. So, the shield walls were between and perpendicular to the Ouse and a ditch, whereas Jones proposes they were parallel to and on both sides of the ditch.
  2. Heimskringla (Hollander) says that the “lines were thinnest by the ditch” and that the “the wing of the Norwegian army extending to the ditch gave way”. This corroborates that the flanks extended to the ditch rather than that they were arrayed along the ditch, thereby contradicting a battlefield at Germany Beck. The offset in the English line that led to their defeat was caused by Morcar’s flank advancing whereas Jones proposes that it was caused by Edwin’s flank falling back. Any credible reading of this second phrase contradicts a battlefield at Germany Beck.
  3. Heimskringla (Hollander) describes the initial engagement: “Then the earls proceeded down along the ditch. There the wing of the Norwegian army extending to the ditch gave way, and the English followed them up, thinking that the Norwegians were about to flee. That part of the English army was led by Morkere.” So, Morcar’s flank forced back Tostig’s flank, but that is implausible at Germany Beck. There is no possibility that Morcar would relinquish advantageous higher ground to cross a boggy ditch and fight uphill on horribly disadvantageous ground, and if they did there is no possibility that they would force back Tostig’s flank. Jones presumably realises this because he states that the battle started with Morcar’s flank advancing to the ford over Germany Beck without crossing and without forcing back Tostig’s flank, but this contradicts the contemporary accounts.
  4. Jones proposes that Hardrada’s flank crossed the ‘delta’ to attack and force back Edwin’s flank on the other side. The contemporary accounts give no indication that any of this happened. As Jones says, the 1m wide upper beck would have been ‘easy to defend’, so the 40m wide ‘delta’ would have been trivial to defend. There is no plausible reason that Hardrada would have tried to cross the ‘delta’, and no plausible reason that Edwin would have allowed Hardrada’s men to get out of the gloop alive.
  5. Heimskringla (Hollander) says of the English flight: “most leapt into the ditch. There the bodies of the fallen lay so thick that the Norwegians could walk dry-shod over the swamp”. Jones depicts this the only way it could have happened at Germany Beck, which is to say that Morcar’s men to fled southeast, but they would have clearly fled north on higher ground towards Jorvik.
  6. Heimskringla says that there was marshland ‘beyond the ditch’. This makes no sense at Jones’s Germany Beck battlefield because the armies are supposed to have been arrayed parallel to the ditch.
  7. Heimskringla says that the English advanced “down along the ditch” which implies they were moving downstream or downhill whereas they would be heading upstream and uphill at Germany Beck.
  8. John of Worcester and Symeon say: “Edwin and Morkar, with a great army joined battle with the Norwegians on the north bank of the River Ouse”. Germany Beck is on the east bank of the Ouse, not the north bank.
  9. Heimskringla says that the ditch was ‘upp á landit’, ‘inland’, of the Ouse, whereas Germany Beck was seaward of the Ouse.
  10. If the battlefield was at Germany Beck, the stream should have been a major factor in the battle, yet it seems to have played no part in the battle before the English flight.

Arguments against the new orthodox battlefield

1. There was no road across the tidal part of Germany Beck

Jones depicts two routes between Riccall and Fulford (Figure 4). He describes them thus: “From Riccall there are two routes on firm ground that converge at Fulford and a third route which passed through some wet land. The main route was along the moraine via Escrick. The other firm route followed the Ouse to the north west before striking north to Naburn and then towards the ford. Although they did not have the benefit of surfaced roads, the going would likely be good enough to cope with a marching army. The third, most direct route is indicated on some 18th century maps, but this had a limited capacity since in three places it had swampy ground on either side”.

Figure 4: Jones's diagram of roads between Riccall and Jorvik

Jones goes on to say: “There would doubtless be other tracks and routes leading from Riccall to the fording point”, then: “There was no acceptable by-pass or other route to reach York from the south so the ford was an excellent ‘choke point’ for the defenders to control because the invaders would be forced to pass that point.” Jones does not explain the source of his information about roads between Riccall and Fulford, and the LiDAR (Figure 5) makes it unlikely that there were any roads out of Riccall. We have regressed sea levels to 1066. Broadly, land shaded green, yellow or red on Figure 5 was permanently above water, land shaded blue areas was marshy, permanently underwater, land shaded cyan was fenny, sometimes underwater.

Figure 5: Humber Levels in 1066 based on LiDAR - Jorvik J; Riccall R; Germany Beck G

Figure 5 shows that in 1066 Riccall (R) was an island surrounded by tidal saltmarsh. The Escrick moraine ridge, to which Jones refers, was between Stillingfleet (S) and High Catton (HC), passing though Escrick (E) and Wheldrake (W) with a gap for the River Derwent. It is at least five miles from Germany Beck. The gap was marshland and fen. Jones recognises this, proposing that Riccall was connected to the Escrick moraine ridge by a causeway and/or levee. It seems unlikely. The distance is ten times longer than any other causeway in the region. Natural levees can develop on the banks of rivers that regularly flood but seldom in salt marshes because they get washed away by tidal scour. There is no documentary or physical evidence of causeways or levees on the Humber Levels, but plentiful evidence that people and freight moved by boat. Moreover, if there was a land link between Riccall and Jorvik, Hardrada had no reason to move his army to Stamford Bridge a week later. Even if there was a causeway or levee north out of Riccall, the Norse would have been jeopardising their lives to have used it because it would not have been designed hold an army.

We suspect that Jones was misled by Freeman’s unfounded claim that the Norse army arrived at the battlefield on foot. We have no reason to doubt the contemporary accounts are right that the Norse rowed upstream from their camp to the theatre of war. If so, they had no reason to form a shield wall south of Germany Beck.

2. Fulford ford is unlikely to have been over Germany Beck

Jones has found some evidence of an unpaved medieval track near modern Fulford. If it went from Jorvik to Water Fulford, it might have crossed Germany Beck at a ford, and that ford might have been the source of the manor’s name. We are sceptical.

Water Fulford had one family in Domesday. The average traffic between it and Jorvik would have been perhaps one cart a day. Tidal fords are maintenance intensive because they get scoured twice a day. Even the Romans balked at cost of building and maintaining them. It seems to us that there was nowhere near enough traffic to justify maintaining a tidal ford at the location Jones suggests. Romans usually diverted roads to cross estuaries at their head of tide. Briden reckons that the maximum tidal range at medieval York was 1.5m which implies that Germany Beck would have been tidal 600m upstream (Jones reckons 900m without explaining his calculation). In the unlikely event that there was a river-hugging east bank road, it would surely have been diverted to Germany Beck’s head of tide, where the fluvial stream was narrow enough to step across. Indeed, this is exactly where evidence of a paved Roman road (RR803x) has been found.

Another reason to think that there was no ford over the tidal part of Germany Beck is that the ‘Ful’ part of Fulford means ‘muddy’. Jones argues that Germany Beck would have been muddy, but it seems unlikely. Its tiny catchment area was too small for significant alluvium deposits, and it would have been scoured twice a day by the tide.

In summary, there is no evidence of a river hugging Ouse east bank road, no incentive to build such a road, and therefore no likelihood that there was a ford over the tidal part of Germany Beck. Even if there was such a road and ford, in our opinion, it would have been too small and too little used to have had a name that anyone would recognise. If Fulford ford was not over Germany Beck, the battlefield is elsewhere.

3. Absence of archaeology

No battle related archaeology has been found at Germany Beck. Chas Jones explained to us that anything valuable would have been scavenged and that the acid soil would have corroded away any weapons or armour. This seems unlikely because Romano-British iron archaeology has been found nearby. Moreover, hundreds of men died at the battlefield, over 1000 Englishmen alone, according to Marianus Scotus. There should be lots of copper alloy personal items, like strap ends, brooches, pins and buckles that were too small to be scavenged.

Summary

Jones was unlucky. His theory was based on Laing’s translation of Heimskringla which has multiple translation errors that make it a plausible match with Germany Beck. Jones updated his theory in 2020. He acknowledges that the other translations contradict the geography at Germany Beck, but he stands by his theory because of his other evidence. Alas, as we explain above, we think that is faulty too.

Germany Beck has a ditch, but it has the wrong orientation for the battlefield. Gate Fulford has the right orientation for the battlefield, but no ditch. Neither of them has any supporting evidence, other than proximity to modern Fulford. Germany Beck has a dozen inconsistencies with the contemporary accounts. Some of them apply equally, or more so, to the traditional Gate Fulford battlefield.

Battlefield location clues

The established Fulford battlefield candidates have no supporting evidence and are inconsistent with the contemporary account descriptions. It is therefore worth considering other candidates. The only geographic and military details appear in Heimskringla (see Appendix A). It is discouraging then that it has a couple of major errors, namely that it says Morcar was killed in the battle and that his co-commander was named Walthiof. Some historians interpret these errors as evidence that Heimskringla is an unreliable source. In practice, Norse Sagas are prone to these types of incidental errors through having been handed down by word of mouth for two hundred years. In our opinion, they do not undermine the veracity of their core events, in which case there is a coherent pre-battle narrative:

  1. The English army was in the medieval walled city of Jorvik.
  2. The invaders entered the Humber in some 300 ships, then sailed up the Ouse to camp at Riccall.
  3. The following day the Norse fleet sailed/rowed up the Ouse mooring midstream within sight of Jorvik’s ramparts.
  4. The English army marched south from Jorvik along the riverbank.
  5. The Norse fleet landed on the north bank of the Ouse.
  6. The Norse army disembarked and marched north.
  7. The battlefield was between the River Ouse and a watery ditch.

There is also a credible and consistent battle narrative:

  1. The armies were arrayed in shield walls between the River Ouse and a ditch that was parallel to it, with Hardrada and Edwin facing each other on the flanks near the river, Tostig and Morcar facing each other on the flanks near the ditch.
  2. Morcar’s flank pushed back Tostig’s flank.
  3. Morcar’s flank advanced beyond the front rank of Hardrada’s flank, allowing Hardrada’s men to cross the battlefield behind Morcar’s men, surrounding them on three sides.
  4. Morcar’s men were squeezed sideways into the ditch and the marsh beyond where many perished.
  5. Seeing their compatriots capitulate, Edwin’s men fled along the Ouse back to Jorvik, some tried to escape into the river.

A handful of disappointingly vague battlefield location clues can be extracted from the contemporary accounts:

  1. The battlefield’s nearest named settlement was ‘Fulford’.
  2. The battlefield was adjacent to a stretch of the Ouse between Riccall and Jorvik.
  3. The battlefield was between the Ouse and a watery ditch that was parallel to it.
  4. Beyond the ditch was marshland that was sometimes flooded, perhaps after heavy rain or perhaps at high tide.
  5. The battlefield was near to and north of where the Ouse had a north bank.

Anglo-Saxon Fulford

We are about to propose that the Battle of Fulford was fought on the Ouse west bank. The glaring apparent inconsistency is that modern Fulford is on the east bank. It is not the smoking gun that most experts think.

Fulford manor was presumably named after a ford. To be well known enough to be referenced by Symeon and Gaimar, it seems likely that this ford carried a relatively important road over a relatively important river. The only river candidates in the target vicinity are the Ouse and the Fosse. Both have a huge flat catchment area, so both carry a lot of alluvial sediment. A ford over either of them may have been known as ‘Fulford’, meaning ‘muddy ford’. The only major road in the target vicinity was the RR810 Roman road that ran on the Fulford glacial moraine ridge, the known part of which is depicted in black dots on Figure 6. It probably had a south branch, RR803(x), from Heslington that crossed Germany Beck at the modern golf course.

The main Roman road east of Jorvik was the RR810. The RRRA have found evidence that it forked at Dunnington, with a southern branch running along the crest of the Fulford moraine ridge (the known part of which is depicted as a solid black line on Figure 6). That branch might have forked at Heslington with a minor branch, the RR803(x), that crossed Germany Beck at the golf course. The RRRA propose that the Fulford moraine road terminated at Colonia, Jorvik’s vicus outside its southwest gate. If so, it crossed the Ouse on a ford south of Jorvik. They suggest that ford was at modern Clementhorpe near the Ouse/Foss confluence (see Figure 6).

Figure 6: Fulford district relief in 1066 with RRRA's proposed route for Margary 810
Figure 7: Fulford district relief in 1066 with our proposed route for Margary 810
Figure 8: OS Map of the battle theatre

It seems likely to us that the road dipped south to cross the Ouse adjacent to the Millennium Fields (Figure 7). Here are our main reasons:

  1. The only stretches of the Ouse with no marshy ‘ings’ to south of York – and therefore the only places that might have had a ford - are Clementhorpe and the Millennium Fields (see Figure ). But Clementhorpe is close to the Ouse/Foss confluence where the river was fast enough to wash away a ford after heavy rain. The ford is therefore likely to have been adjacent to the Millennium Fields.
  2. A section of north-south paved Roman road has been found at Nunthorpe on the Ouse west bank (see below), which suggests the ford was south of it, consistent a crossing near Old St Oswalds.
  3. Old St Oswalds church, on the east bank close to where we depict the ford, is thought to have been built on the site of an earlier Anglo-Saxon church. That church was presumably associated with the only known settlement in the vicinity, namely Fulford. Fulford got its name from a ford, so the ford was probably adjacent to Old St Oswalds church.
  4. Christopher Rainger found a road mirroring St Oswalds Road on the opposite bank, suggesting that ford crossed immediately west of Old St Oswalds church.
  5. There is a medieval stone cross in modern Fulford, now outside ALDI’s carpark. Its information plaque says it might be a 1484 boundary marker, although it could be much older. It might have marked the corner of RR810 as it turned towards the ford. This would be consistent with a Victorian railway causeway that once crossed Millennium Fields. Its location can still be seen on LiDAR (circled on image below), on the route from Fulford Cross to the Ouse. We wonder whether it was built on an existing Roman causeway or was built over Roman road foundations
  6. According to Heimskringla, the English advanced along the river, which is only likely if they were on a road.

In summary, there is no evidence that Fulford ford crossed Germany Beck, and it seems more likely to us that it crossed the Ouse. It is possible that the battle was named after the ford rather than the settlement, in which case it is just as likely to have been on the west bank as on the east bank. There is a more likely explanation. Domesday lists two Fulfords near York, one spelled ‘Fuleford’, the other ‘Foleforde’. Historians think they were north and south of modern Fulford, but the entries are in different sections of Domesday. Fuleford is in a section along with Ouse east bank manors. Foleforde is in the Westriding section along with west bank manors. It implies to us that the two Fulfords straddled the Ouse at the ford. If so, the battle might have been named after a manor on either side of the Ouse, and it is just as likely to have been on the west bank as on the east bank.

Our proposed battlefield

There is nothing clever about our Fulford battlefield location theory. As we explain above, the geography east of the Ouse does not match the contemporary accounts battlefield descriptions. The only other place near to medieval ‘Fulford’ and the Ouse was the west bank opposite modern Fulford, and it matches all the geographic battlefield clues: It was between Riccall and Jorvik; It was near the Ouse’s only north bank; It had a watery ditch, namely Knavesmire Beck, that was parallel to the Ouse; It had marshland beyond the ditch.

Knavesmire Beck is not obvious today because it was culverted in the 19th century. It is only visible where it discharges into the Ouse, but it is shown as a normal beck on the 1842 Ordnance Survey map with racehorses crossing on bridges at the 14 and 5-furlong posts. Jones questions whether Knavesmire was wetland in the 11th century. Its Old Norse name suggests it must have been. MYO4287 HER explains: “In the early 18th century the Knavesmire was drained and levelled. The first race meet took place there in 1731 having moved from Clifton Ings.” It sounds like it tidal marshland in the 11th century.

Figure 9: Fulford initial deployments with out proposed route for Margary 810

The starting point for our proposed battle narrative is John of Worcester who says: “Edwin and Morkar, with a great army joined battle with the Norwegians on the north bank of the River Ouse, near York”. The Norse fleet was coming upstream, so this means that they landed on the north bank of the Ouse and that the battlefield was on the north bank of the Ouse near York. This is only possible if they landed at Middlethorpe Ings.

Next, Heimskringla says: “Then the king went on land and began to array his army for battle. One wing stood on the bank of the river, the other was arrayed further up on land, and extended to a ditch.” This means that the Norse army lined up across Middlethorpe Manor (cyan on Figure 9).

Meanwhile Edwin and Morcar were leading the English army south from Jorvik. Heimskringla again: “The earls deployed their army down along the river with the whole body of their men.” So, the English headed east on what is now Nunnery Lane to join the Roman predecessor of Bishopthorpe Road, the cobbled remains of which were found 2m below the current surface in a 1981 excavation (EYO786). We think this road crossed the Ouse on Fulford ford, then followed the York moraine ridge to Derventio at modern Low Catton. Regardless, if there was a Roman road along the Ouse, the English would have used it. We think it crossed the Ouse at Nunthorpe, so the English army would have been as depicted in magenta on Figure 9.

Figure 10: Engagement and battlefield; our suggested position of the ford shown in black

The Norse would not have wanted to fight with a swamp to their rear, so they would have advanced at least a few hundred metres north. After Morcar’s flank fled into the ditch, John of Worcester says: “Not without some small loss they [the English] turned to flee, and many more of them were drowned in the river than had fallen in battle.” The English are only likely to have jumped into the river if it was closer to them than the Roman road. This means that the English line was probably at least as far south as the Terry’s chocolate factory, which is also alongside the only stretch of the Ouse where the river flows close to the riverbank with no ings. On balance, we think that the contact zone was between the one-furlong pole and the finish at York racecourse, alongside the car park entrance, as shown on Figure 10.

Figure 14: Tostig’s flank is pushed back

Figure 11 shows Tostig’s flank being pushed back. Morcar’s men think they are fleeing and chase. They come alongside Hardrada’s men (Heimskringla): “King Harald saw that the battle array of the English had come down along the ditch right opposite them”.

Figure 15: Morcar’s men are sandwiched

Figure 12 shows that Hardrada’s men crossing the battlefield to get behind Morcar’s cohort: “King Haraldr led the attack with his troops and joined battle so fiercely that the enemy was split” (Morkinskinna), and “he turns with his men at the back of Morcar's battle” (Hemings Tale). Morcar’s men are sandwiched with nowhere to fall back.

Figure 13: The English flee

Heimskringla: “Some fled up or down along the river, but most leapt into the ditch.” Morkinskinna: “the local army began to flee. They retreated to a place where there was no armed opposition, in the swampy ground around the pool.” So, Morcar’s men fled into the ditch (Figure 13).

The east bank of Knavesmire Beck was lined by 50m of marsh. Morcar’s men are forced into the marsh. So many Englishmen died in this ditch and marsh that the Sagas reckon the Norse could walk over their bodies without getting their feet wet (Heimskringla). Meanwhile, seeing Morcar’s men capitulate, Edwin and his men fled back to Jorvik along the Roman road. Some stragglers jumped into the river to avoid getting caught by the chasing Norse.

So many Englishmen died in the river that it was blocked with their corpses according to Vita Edwardi Regis. This implies to us that they probably died crossing Fulford ford, because the bodies would otherwise have been washed downstream.

There is one apparent minor discrepancy between the contemporary account geography and the current Knavesmire. Most of the accounts say that one side of the battlefield was bounded by a ditch with marshland beyond. But the Knavesmire is level. If there was marsh on one side of the ditch, there would be marsh on the other side. When it floods these days – which it did as recently as 2020 - the entire racecourse is under water. It doesn’t sound quite right. It is known that the racecourse was levelled in the 18th century. There are no records of the details, but perhaps a metre or so of topsoil was taken from the finishing straight and dumped on the back straight. If not, we can only imagine that Knavesmire Beck ran 150m east of its current culvert.

We cannot definitively prove that the battle was fought on the Chocolate Factory Peninsula between the Ouse and Knavesmire Beck, but it uniquely fits all the geographic clues, and it is not inconsistent with any of the contemporary accounts. There are only two other candidates, both on the east bank and both of which fail to match the geography or events described in the contemporary accounts.

Arguments against an Ouse west-bank battlefield

Jones considers other possible battlefields and gives a list of reasons for why he thinks that the battlefield was not west of the Ouse. Here are his reasons and our responses.

C J: “It lacks any ditch even though augering was undertaken to locate any ‘lost’ ditch. There are no equivalent becks entering this stretch of land that can be identified, even using LIDAR or on old maps.”

MB: This is wrong. Knavesmire Beck runs through Knavesmire north to south. It has been culverted, but at the time of the battle, it would have been just like Germany Beck on the other bank.

C J: “Some early historians place the battle on the northern bank of the Ouse while Middlethorpe is west of the Ouse.”

MB: Upstream of its confluence with the Wharf, the Ouse mainly flows north-south. Its banks would normally be referred to as the east and west. There is only one place in the battle theatre that could be described as the north bank of the Ouse, which is Middlethorpe to the north of an ENE to WSW flowing stretch of the river.

C J: “Other sources mention ‘south of York’ and Middlethorpe moves the battle to the western bank and it is not sensible to regard it as ‘south’.”

MB: It is not obvious to us what Jones is trying to say, but our proposed battlefield on the Chocolate Factory Peninsula is south of York.

C J: “Both armies would have to cross to the west bank of the Ouse to fight. This might have been possible for the Norse if sandbars and low tides had allowed this, but the evidence for low-tide fording places across the Ouse is extremely limited. The tides at the time of the battle were extremely high with a corresponding very low tidal level to follow. The onrushing tidal bore would threaten a repeat of the closing of the Red Sea, with its fatal consequences if an army got the timing wrong. There is no mention or hint of the armies crossing the Ouse in any writing and it is hard to image that such a ‘parting of the waves’ would have gone unmentioned.”

MB: This arbitrarily presupposes that the Norse arrived at the battlefield on foot and that the English left York through the east gate. It has no basis. The Norse accounts clearly state that the Norse army arrived at the battle theatre by ship. They could have disembarked on either bank. If they landed on the west bank, the English would have left through the southwest gate on the Roman road to Tadcaster.

C J: “The presence of elements of the Mercian army to the west of the Ouse is however implied by the evidence of tanged arrows on the only possible mooring place, which is beside Water Fulford Hall. This accommodation would provide a further incentive for the Norse army to stay on the opposite bank. But physical evidence in the form of tanged arrows however does imply some action on the Middlethorpe Ings although confined to the southern limit where the Ouse turns sharply west towards Bishopthorpe.”

MB: The tanged arrows were found close to where we think the Norse fleet moored and close to where we think the battle was fought. They might be evidence of a west bank landing and a west bank battlefield.

C J: “Even had the opposing armies come face to face on these Ings …”

MB: As far as we know, no one is suggesting that the armies met on the Ings. They would have met on the relatively high ground north of the Ings, to the east of York racecourse.

C J: “… there is no obvious place in this landscape for either of the commanders to anchor their flanks except by stretching their shieldwalls to fill a gap of 2100 metres between the river bank and the Knavesmire, which might or might not have been wetland in 1066.”

MB: ‘mire’ is Old Norse for ‘marsh’, so Knavesmire was almost certainly wetland in 1066. The maximum lateral distance from the Ouse to Knavesmire Beck is 900m, where it crosses Middlethorpe Ings. We think the battle was further north, around the York racecourse carparks and the former Terry’s chocolate factory, where the peninsula was roughly 600m wide.

C J: “The Norse ships posed a threat to which the Mercians were bound to respond. The boats might threaten an attack on York, make a landing on the west bank, or a move to outflank Morcar. A bridge of boats might have been feasible and that would have moved part of the conflict to Middlethorpe. However, the possibility of a D-Day style landing by King Harald to allow full battle on these ings, didn’t happen. It would be many centuries before an opposed, mass landing from ships would be successful.”

MB: This arbitrarily presupposes that the English were waiting at the battlefield for the Norse army/fleet to arrive. It has no basis. The Norse accounts unambiguously say that the Norse fleet was ‘lying in the Ouse’ when Morcar and Edwin leave York, they unambiguously go on to say that Harald then lands and arrays his troops before the engagement. Therefore, the English were not at already at the battlefield waiting for the Norse army to disembark.

C J: “This was one of the areas chosen by archaeologists for a comparative metal detecting survey but it yielded only modern debris.”

MB: It sounds like the survey was in Middlethorpe Ings, and therefore unlikely to produce any Saxon era finds. Even if it was near the prospective battle theatre, absence of evidence is not evidence of absence.

C J: “The area was also subject to some soil survey work as a way to understand the Ouse. The possibility that the Fulford Ings had at one time extended further west to include the Middlethorpe Ings (or vice versa) could not be excluded until the evidence demonstrated the stability of the course of the Ouse.”

MB: Once again, this arbitrarily presupposes that the Norse army arrived at the battle theatre on foot when the Norse accounts unambiguously say they arrived by ship. If they arrived by ship, they could have landed at Middlethorpe Ings.

C J: “Finally, if one respects the tradition that has named this the battle of Fulford, this name cannot sensibly be attached to a battle on the Middlethorpe Ings.”  

MB: Middlethorpe was probably known as Torp when Symeon and Gaimar named the battle, but it had no inhabitants according to Domesday. We think the battle took its name from the ford, or from a settlement adjacent to the ford, because it was the closest named place that Symeon and Gaimar knew. As we explain above, there are two ‘Fulfords’ in Domesday, and we think that they were either side of the ford. If so, the battlefield could have been on either side of the Ouse.

Conclusion

The Battle of Fulford narrative is about as straightforward as any medieval battle can be. There are only two main accounts - one English, one Norse - and they are not inconsistent. All the others are based on one or the other, albeit with some added information. They form a coherent engagement narrative in which the armies face each other across a narrow peninsula. They form a coherent battle narrative, in which Morcar’s flank advances beyond Hardrada’s flank, allowing Hardrada to cross the battlefield and sandwich Morcar’s men. They form a coherent flight narrative, in which Morcar’s men flee into the beck and marsh beyond, while Edwin’s men flee along the riverbank. They all point to a battlefield on the west bank of the Ouse, probably either side of the County Stand on York racecourse.

Appendix A - The contemporary accounts

ASC-C (Whitlock):

Harald, king of Norway, came by surprise north into the Tyne with a very large naval force - no small one: it could be [500] or more. And Earl Tosti came to him with all those he had mustered, just as they had agreed beforehand, and they both went with all the fleet up the Ouse towards York. Then King Harold in the south was informed when he disembarked that Harold, king of Norway, and Earl Tosti were come ashore near York. Then he went northwards day and night as quickly as he could assemble his force. Then before Harold could get there Earl Edwin and Earl Morcar assembled from their earldom as large a force as they could muster, and fought against the invaders and caused them heavy casualties, and many of the English host were killed and drowned and put to flight, and the Norwegians remained masters of the field. And this fight was on the eve of St. Matthew the Apostle, and that was a Wednesday. And then after the fight Harold, king of Norway, and Earl Tosti went into York with as large a force as suited them, and they were given hostages from the city and also helped with provisions, and so went from there on board ship and settled a complete peace, arranging that they should all go with him southwards and subdue this country.

ASC-D & ASC-E (Whitlock):

Meanwhile Earl Tosti came into the Humber with sixty ships and Earl Edwin came with a land force and drove him out, and the sailors deserted him. And he went to Scotland with twelve small vessels, and there Harald, king of Norway, met him with three hundred ships, and Tosti submitted to him and became his vassal; and they both went up the Humber until they reached York. And there Earl Edwin and Morcar his brother fought against them; but the Norwegians had the victory.

Symeon (Stevenson):

After these events, Harold Harvager, king of the Norwegians, brother of king Olave the saint, came unexpectedly to the mouth of the river Tyne with a very powerful fleet, to wit, more than fifty great ships. Earl Tosti, with his fleet, met him there as he had before agreed, and with a quick voyage they entered the mouth of the river Humber, and so sailing up the river Ouse they landed at the place called Richale, and took York after a hard struggle. When king Harold learnt this, he rapidly marched his troops towards Northumbria. But before the king arrived, on Wednesday the vigil of St. Matthew the Apostle, the brother earls Edwin and Morkar, with a large army, joined battle with the Norwegians at Fulford, near York, on the northern bank of the river Ouse, and at the first onset of the fight they overthrew many; but after a long continuance of the contest, the Angles, unable to resist the force of the Norwegians, turned their backs not without some loss of their men, and many more of them were drowned in the river than fell in the field. The Norwegians were masters of the field of slaughter, and taking five hundred hostages from York and leaving there a hundred and fifty hostages of their own men, they retired to their vessels.

John of Worcester (McGurk):

When these things had been done, Harald Fairhair [should be Hardrada], king of the Norwegians, brother of St Olaf the king, landed unexpectedly at the mouth of the River Tyne with an extremely strong fleet; that is more than 500 great ships. Earl Tostig joined him with his fleet as he had previously promised, and on a swift course they entered the mouth of the River Humber; sailing thus up the River Ouse, they landed at a place called Richale [Riccall]. When King Harold learnt of this, he speedily undertook an expedition to Northumbria. But before the king arrived there, the two brother earls, Edwin and Morkar, with a great army joined battle with the Norwegians on the north bank of the River Ouse, near York, on Wednesday, the eve of St Matthew the Apostle's day, and fighting manfully in the first thrust of the battle, they laid many low. But after the struggle had continued for a long time, the English were unable to withstand the Norwegian attack. Not without some small loss they turned to flee, and many more of them were drowned in the river than had fallen in battle. But the Norwegians gained the mastery in that place of death and, having taken 150 hostages from York, they returned to their own ships, having left in York 150 hostages of their own.

Gaimar (Stevenson):

He and Tosti had spoken so much together, that each pledged his faith to the other, that whatever they conquered together, they would divide all equally. Now they wished first by their warfare to divide all England between them. The two had a great fleet, four hundred ships and seventy sailed forward. They steered and sailed a great way until they entered the river Humber; from the Humber they went to the Ouse, and disembarked at Saint Wilfrid's. On the morrow they set sail for York, and arrived there in the evening. But the two earls met, and brought the people of six counties; and they fought at Fulford, the Norwegians were masters of the field; but, on both sides, there were many killed. Afterwards the Norwegians took the land; they desolated all the country and seized many spoils. Whoever knows not this, let him remember that it was twelve days within September.

Henry of Huntingdon (Forester):

He [Tostig] escaped to Scotland, where he fell in with Harald, king of Norway with 300 ships. Tosti was overjoyed, and tendered him his allegiance. Then they joined their forces and came up the Humber, as far as York, near which they were encountered by the Earls Edwin and Morcar; the place where the battle was fought is still shown on the south side of the city. Here Harold, king of Norway, and Tosti, his ally, gained the day.

Morkinskinna (Andersson & Gade)

He [Harald] subjected the whole north of England, then sailed south to the Humber and sailed up the Ouse. There the jarls of Northumbria, the sons of Jarl Godwin, Morkere and Waltheof of Huntingdon, marched against him with an enormous army that had been gathered throughout the summer and autumn. King Haraldr elected to give battle. He landed and drew up his troops on the riverbank, with one wing toward the river and the other toward the still water of a deep pool. The jarls led their formation with all the rank and file down along the river. The king's standard was close to the river, where his battle array was densest, while it was thinnest by the pool. There too the troops were least reliable, and when the attack was made, that wing gave way. The English launched their attack down along the river, then toward the pool, and they thought that the Norwegians would flee. King Haraldr led the attack with his troops and joined battle so fiercely that the enemy was split and the local army began to flee. They retreated to a place where there was no armed opposition, in the swampy ground around the pool. Jarl Morkere had followed the standard closest to the pool, and Jarl Waltheof fought the king more toward the river. He fled up along the river, and the troops with him were the only ones to escape. Jarl Morkere fell, and together with him so many men that the pool was full of corpses where the fleeing men had congregated. Some were driven into the water and killed there together with some who jumped into the pool and were speared, so that it was filled with corpses, as Steinn Herdisarson says:

Many died in the river; submerged men drowned;
soon countless warriors lay dead around young Morkere.
The proud lord of the Filir pursued the fleeing enemies;
the army rushed madly before the ready king.
Proud-minded knows beneath [the sun].

He composed the praise poem that includes this stanza for King Haraldr's son Olafr. It was the evening before Saint Matthew's Day, which fell on a Wednesday. After this battle all the troops in the neighbouring districts submitted to King Haraldr, but some fled, as is told in this poem:

Young lord [Olafr], you let farmers fall close to the River Ouse;
the army attacked where the reckless king risked his life.
The arrival of that king must undoubtedly appear to those
Englishmen who were left alive as if they are still pursued.

Fagrskinna (Finlay)

From there he sailed the fleet south to the Humber and went up along the river to the Ouse. There the jarls of Northumbria, Morukári (Morkere) and Jarl Valbjófr of Hundatún (Waltheof of Huntingdon) came against them with an overwhelming force, which had been gathering together all the autumn. King Haraldr offered them resistance, went up onto the river bank and drew up his troops there with one flank of the army facing towards the river and a second towards a certain ditch. And that was deep and full of water. The jarls sent their army with all its men slowly down along the river above the king’s standard. The ranks were thickest there, but sparsest beside the ditch. When the fighting had been going on for a short time, the flank of Haraldr’s army lining the ditch gave way, and the Englishmen pressed them hard at that moment, expecting that the Norwegians would flee. King Haraldr with his troop advanced so hard that (the enemy) gave way before him in two directions. Then flight broke out in the land army, and the flank that was nearest to the ditch jumped out into it. And Morukári had been moving with the standard that was nearest to the ditch. Jarl Valbjófr fled from the battle up along the river, and only those who went with him got away. There Morukári fell, and men so thickly around him that the ditch was full of dead men. So says Steinn:

Men drowned, submerged
many died in the river.
No few soldiers soon lay
around young Morukári.
The men’s lord forced on, fearless,
the flight; before the brave leader
headlong the army hastened.
Headed off, strong-minded, from under . . .

He composed this about Óláfr, and that was on the eve of St Matthew’s Day (21st September), which fell on a Thursday that year. Morukári and Valbjófr were brothers of Haraldr Guäinason.

Heimskringla (Hollander):

Thereupon King Harald sailed to the Humber and up the river, and anchored there. At that time there were two earls in York, Morkere and his brother Wæltheow, with a huge army. King Harald lay in the Ouse when the army of the earls came down from the land to oppose him. Then the king went on land and began to array his army for battle. One wing stood on the bank of the river, the other was arrayed further up on land, and extended to a ditch. There was a swamp, deep and broad and full of water. The earls deployed their army down along the river with the whole body of their men. The royal banner was close by the river. There the king’s men stood thickest, and the lines were thinnest by the ditch, with the troops he could least rely on. Then the earls proceeded down along the ditch. There the wing of the Norwegian army extending to the ditch gave way, and the English followed them up, thinking that the Norwegians were about to flee. That part of the English army was led by Morkere. But when King Harald saw that the battle array of the English had come down along the ditch right opposite them, he had the trumpets blown and sharply urged on his men to the attack, raising his banner called Landwaster. And there so strong an attack was made by him that nothing held against it. Then there was a great slaughter among the earls’ men. Soon their army took to flight. Some fled up or down along the river, but most leapt into the ditch. There the bodies of the fallen lay so thick that the Norwegians could walk dry-shod over the swamp. There Earl Morkere lost his life. As says Stein Herdisarson:

Their lives lost there many, left this world by drowning;
Mired in the marsh, lay by Morkere young a legion.
Pursued the sea-king this smitten host.
They madly fled before the brave king.
Foremost under heaven

This dråpa Stein Herdisarson composed about Óláf, the son of King Harald; and we are told here that Óláf took part in the battle with King Harald, his father. This is mentioned also in the poem called Haraldsstikki:

Lay the fallen in fen thickly,
Wæltheow’s men, by weapons slaughtered;
so that walk could warlike Northmen
on dead bodies dryshod across.

Earl Wæltheow and those who managed to escape fled to the 1066 fortified town of York. It had been a murderous battle. It took place on Wednesday, the day before Saint Matthew’s Day.

Sagas of Olaf Tryggvason and Harold the Tyrant (Hearn & Storm):

Now having come thus far on his journey King Harald fared south to the Humber and went up that river and lay in it beside the banks. At that time there were up in Jerirk [York] Earl Morcar and his brother Earl Waltheof and with them was a vast host. King Harald was lying in the Ouse when the host of the Earls swooped down against him.

And King Harald went ashore and set to arraying his host, and one arm of the array was ranked on the banks of the river, whereas the other stretched up inland over towards a certain dyke, and a deep marsh was there, both broad, and full of water. The Earls bade the whole multitude of their array slink down alongside the river.

Now the banner to the King was near the river and there the ranks were serried, but near the dyke were they more scattered, and the men thereof also the least trustworthy.

The Earls then came down along by the dyke, and that arm of the battle-array of the Norwegians which faced the dyke gave way, and thereon the English pushed forward after them and as it seemed that the Norwegians would flee. Therefore did the banner of Morcar progress forward.

But when King Harald saw that the array of the English had descended alongside the dyke and was coming right toward them, then commanded he the war-blast to be sounded, and eagerly encouraged his men, and let the banner ‘Land-waster’ be carried forward; and even so fierce was their advance on the English, that all were repulsed and there fell a many men in the host of the Earls.

This host was even soon routed, and some fled up beside the river and some down, but the most of the folk ran right out into the dyke, and there the fallen lay so thick that the Norwegians could walk dry-shod across the marsh. There too fell Earl Morcar. Thus said Stein Herdisason:

Many in the river sank (The sunken men were drowned);
All round about young Morcar of yore lay many a lad.
To flight the chieftain put them;
The host to swiftest running Olaf the Mighty is.

The song that follows was wrought by Stein Herdason about Olaf the son to King Harald, and he said, by which we know that Olaf was in the battle with his father. This is told likewise in ‘Haraldsstikka’:

There the dead lay Down in the marsh
Walthiof’s fighters Weapon-bitten,
So that they might The war-wonted horsemen
There wend their way On corpses only.

Earl Walthiof and those men that contrived to make their escape from out the battle fled even up to the town of York, and there it was that the greatest slaughter took place. This battle was on the Wednesday or ever St. Matthew’s Day.

Heming’s Tale – Appendix to the Orkneyinga Saga (Dasent):

They [Hardrada and Tostig] go on land with the host, but some watch the ships.  Those brothers Morcar and earl Voltheof, and Aki their brother in law, gather a host together as soon as they hear of the Norwegian host.  They met at that river which is hight [named] Ouse, and there the hardest fight arises, and is kept up till nones [mid-afternoon].  Then Eystein had gone through the array of the English and slain Aki the tall.  Then he sees that Morcar has got at the back of Tosti's battle; and so he turns with his men at the back of Morcar's battle. And when earl Morcar sees that he bids his men face about and defend themselves well and manfully. And at last flight broke out among his men, and they flee out into the river [stream] and earl Morcar is there slain and the most part of his folk.  Many too sunk beneath the stream.  By that time king Harald had taken earl Valtheof prisoner. 

Bibliography

Contemporary sources

Anglo-Saxon Chronicle (of which three recensions cover the invasion, known as C, D and E); reasonably contemporary with events

Vita Edwardi Regis; c1067

Chronica Clara;  Marianus Scotus; c1070

Libellus de Exordio; Symeon of Durham; c1110;

Chronicon ex Chronicis; John of Worcester; c1125

Historia Anglorum; Henry of Huntingdon; c1125

Gesta regum anglorum; William of Malmesbury; c1135

L’Estoire des Engleis; Geoffrey Gaimar; 1138

Morkinskinna; translation Andersson & Gade; c1220

Fagrskinna; translation Alison Finlay; c1220 (after Morkinskinna)

Heimskringla; translations Laing, Binns, Palsson, Hollander, Finlay ; c1230

Modern sources

David Cooke – Battlefield Yorkshire from the Romans to the Civil Wars; Pen & Sword; 2006

Kelly DeVries - The Norwegian Invasion of England in 1066; Boydell Press; 1999

Charles Jones - The Forgotten Battle of 1066: Fulford; Tempus; 2006

Charles Jones - Finding Fulford: The search for the first battle of 1066; WPS; 2010

Charles Jones – The Battle of Fulford, property development submission; 2017

Chas Jones - The Literary Sources for the 1066 Battle of Fulford; Academia; 2020

Frank McLynn - 1066: The Year of the Three Battles; Pimlico; 1999

Christopher Rainger - St Oswald’s Old Church, Fulford; 2017

James Ramsay - The Foundations of England; Humphrey Milford; 1898

Michael Rayner - English battlefields : an illustrated encyclopaedia; Tempus; 2004

Guy Schofield - The Third Battle of 1066; History Today, October 1966