While there is no mention of the comet in the saga (and neither is there any account of Tostig’s activities), other omens are described in such sinister detail as to cast the darkest shadow over Harald’s fleet being brought together in the shelter of the Solund Isles. Snorri mentions very many ‘dreams and portents’ reported at this time, but selects just three for inclusion in his saga. The first of these was a nightmare which came to one Gyrth, probably one of Harald’s own housecarls, who was sleeping aboard the king’s ship when he dreamed that he saw a huge troll-woman (one of the monstrously ugly and invariably hostile giant race of northern mythology) out on one of the islands, with a knife in one hand and a trough in the other. As his dreamscape widened, Gyrth could see eagles or ravens perched atop every prow in the fleet as the gruesome giantess sang (in skaldic verse, of course) of the king being enticed west-over-sea to fill graveyards and of birds of carrion following in his wake to feast on slain seamen.
A troll-woman also appeared in a dream to another man as he slept aboard a vessel lying alongside Harald’s ship, this one riding on a wolf with a dead warrior in its blood-streaming jaws and the English battle-array behind her against the skyline. As soon as the wolf had consumed the corpse, its grisly rider dropped another into its jaws while chanting a strophe foretelling reddened shields, fallen fighting-men and the doom awaiting Harald himself. Apparitions very much like these are found elsewhere in the saga literature – perhaps most vividly in its record of ominous supernatural experiences surrounding the battle of Clontarf where the Orkney jarl Sigurd was slain in Ireland in 1014 – and represent a legacy from the darker side of pagan antiquity still preserved in the literary Christian culture of thirteenth-century Iceland. The association of such traditions with Harald may not be entirely accidental, however, and especially in the light of Adam of Bremen’s claim that he ‘gave himself up to the magic arts’. In fact, Adam’s remark most probably refers to nothing more sinister than Harald’s accomplishment in the art of the skald which demanded of its practitioners an extensive knowledge of the ancient beliefs of the northlands to inform the imagery of their kennings. There is good reason – and, perhaps, on more than one count – to believe Harald well acquainted with Odin, the lord of battle among the old gods of the north. The last of the three apparitions described in Snorri’s saga (and also included in the other collections) has no such pagan associations, however, because it concerns a dream – or, perhaps more properly, a vision – said to have occurred to Harald himself and in which ‘his brother Saint Olaf’ brought him a warning: ‘Now I fear, great Harald, your death at last awaits you. . . .’
No ‘dreams and portents’, however disturbingly prophetic, could turn back the great enterprise now because the invasion fleet was ready and word had long since been sent through all the kingdom to summon up a ‘half-levy of the whole army’. These terms ‘levy’ and ‘half-levy’ occur throughout the kings’ sagas and yet there is still no full consensus in scholarly circles as to whether any such system of muster – of which there is no formal historical record until the twelfth century – was actually practised in the Scandinavian warfaring of Harald’s time. While the forces raised by Svein Forkbeard and his son Cnut for their invasions of England certainly appear to have been of the order of national armies as distinct from viking warbands, one school of thought has still recognised them as an effective coalition of the king’s own force (or lið) made up of his warrior retinue of housecarls (or hirð) with the semi-professional manpower of free farmers called bonders and those of his allied chieftains and client rulers, possibly drawn from a wider extent of the Scandinavian world.
The alternative view proposes a systematic levy (or leiðang) of ships and men, together with their weaponry and provisioning, called up by the king and supplied by his subjects on a proportional basis – such as that estimated for the early eleventh century in terms of three farmsteads required to supply one man with his war-gear and provisions. The terms of such a system, known only from later sources, provide for the mustering of a full levy for national defence, while only the half-levy was to be called up for a campaign of aggression such as Harald intended in 1066 and so Snorri’s statement of his summoning ‘a half-levy of the full army’ would correspond to what is known of the leiðang in the twelfth century. Indeed, one Norwegian historian has even proposed this same leiðang as Harald’s own innovation – to which Kelly DeVries adds an observation of particular relevance here when he suggests that ‘it seems ludicrous to believe that someone like Harald Hardrada, who had served in what was probably the most organised army in the world at the time, the Byzantine army, would abandon such a logical notion once he had returned to Scandinavia’.5
None of which, unfortunately, is of very great assistance in attempting to estimate either the size of fleet or the numbers of fighting-men which Harald brought to England in the autumn of 1066. As to ship numbers, Snorri records it being ‘said that the king had over two hundred warships as well as supply ships and smaller craft’ assembled in the Solund Isles and his estimate is comparable (especially if reckoned in ‘long hundreds’) to the round figure of three hundred vessels in the Anglo-Saxon Chronicles, a total which must also include the reinforcements acquired on route, principally those known to have joined the expedition by the time it set sail from Orkney. Other estimates supplied in the English sources tend to range upwards from that figure and even so far as the total of ‘five hundred great ships’ claimed by the historian John (formerly known as ‘Florent’ or ‘Florence’) of Worcester writing in the early twelfth century, but the figure most widely accepted by modern historians would be around the three hundred recorded in the Chronicles and plausibly supported by the saga.
In the absence of any numerical estimate of manpower in the earlier sources, it has been upon a base-line of ship numbers that historians have attempted to calculate the strength of Harald’s army of 1066 – and with an extraordinarily wide range of results. Such calculations are subject to many variables, of course, and not least that of the average number of warriors aboard each ship. Harald’s capital ship at the battle on the Nissa was said to have been fitted with thirty-five benches which would have accommodated a crew of at least seventy oarsmen, in addition to a shipmaster or steersman and, presumably also, the king accompanied by his attendants. Snorri’s claim for its quite exceptional proportions must be taken to mean that there would have been no other vessel of such size in the fleet which sailed from Norway in 1066 (if, indeed, the ‘great dragon’ was still in service at that time, because the sagas make no further reference to it after 1062).
In view of such variables, it is perhaps unsurprising that the estimated sizes of Harald’s forces vary so very widely from as few as seven and a half thousand to the most ambitious estimate (at least, of which I am aware) placing it at eighteen thousand fighting men. While a figure in the region of nine thousand seems to be most often found in the general currency of modern accounts, a more detailed analysis has suggested a figure somewhere between eleven and twelve thousand, presumably including an essentially non-combatant component such as serving-men and boys.
As to the course followed by the fleet between its departure from the Solund Isles and its arrival off the Northumbrian coast, history depends almost entirely upon the evidence of the saga record, which is preserved in greatest detail by Snorri’s Harald’s saga and Orkneyinga saga. The first landfall was in Shetland – as was customary for Norse voyagers bound west-over-sea – whence the fleet sailed south, through the turbulent Sumburgh Roosts towards the Fair Isle and on to Orkney where it assuredly found moorings in the famous broad haven of Scapa Flow.
The jarls Paul and Erlend, sons of the mighty Thorfinn, had recently succeeded their father, who had died only a year or two earlier. They are said by Orkneyinga saga to have been good-natured and well regarded by their people, ruling as joint jarls rather than dividing up the islands and mainland territories between them. Having come awry with King Magnus by reason of a blood-feu
d arising out of the killing of Rognvald Brusason, Thorfinn is said by the saga to have returned again to Norway after Magnus’ death and there made his peace with Harald, effectively acknowledging him as his overlord. So, when Harald’s fleet put into Orkney on voyage to England in 1066, Thorfinn’s sons and successor jarls would have been obliged to join their ships to his expedition – as, indeed, they appear to have done quite willingly. Neither would their Orkney contingent have been the only reinforcement waiting in Scapa Flow, because a Hebridean Norse adventurer by the name of Godred Haraldsson – but more usually remembered as Godred Crovan (‘of the white hand’) – is known to have fought at Stamford Bridge and would almost certainly have joined the Norwegian forces in Orkney, presumably in company with some number of other ships and warriors from the Isles and from Man.
When the fleet sailed out of the Flow to cross the Pentland Firth on voyage for England, Harald’s queen Ellisif remained in Orkney, probably at Kirkwall, with her daughters and there awaited a husband and father whom they were never again to see alive. Presumably Harald had brought his queen and their daughters thus far west-over-sea so as to be nearer at hand to join him when he had won his new kingdom, and yet he had left his elder son Magnus with his mother, Thora, in Norway where he was to act as regent while his father was otherwise engaged upon the conquest of England.
Across the Pentland Firth now and rounding Duncansby Head, the fleet sailed down the east coast of the Scottish mainland until it reached the Forth. Two versions of the Anglo-Saxon Chronicle have been taken to indicate the Scottish court of Malcolm Canmore at Dunfermline as the place of Harald’s first meeting with Tostig, who there ‘gave him allegiance and became his man’, but the evidence of the Abingdon ‘C’ Chronicle is probably to be preferred when it places their meeting at ‘the mouth of the Tyne as they had previously arranged’. If so, then there is no reason to assume that Harald might have put into the Forth on an impromptu state visit to Scotland. In fact, the saga does state quite plainly that ‘he sailed down the coast of Scotland and down the English coast’ and there is a very practical reason why he should have done so, because the wind had thus far been with him, giving his fleet a good speed southwards from Orkney. That same wind was keeping William the Norman’s invasion fleet bottled up at the mouth of the Somme and winds can change or simply drop away altogether, so there was everything to be gained by pressing on to the target zone in northern England.
If the same thought had also been in Tostig’s mind, then he would have had reason to expect the timely arrival of the Norwegian fleet from Orkney and brought his own few ships out of the Forth to meet the Norwegian fleet at Tynemouth ‘as they had previously arranged’. Harald cannot have been greatly impressed by the first sight of his ally’s contribution to the enterprise, although the saga makes no mention of any such opinion and, indeed, makes no mention of Tostig at all until the expedition had sailed up the Humber and won its first battle. Only at that point is there entered a hasty reassurance – framed as if it were a mere detail which had slipped Snorri’s mind – of Tostig having travelled north from Flanders ‘to join King Harald as soon as he arrived in England’.
Snorri does have a full account, and one more detailed than in any other version of the saga, of the Norwegian progress down the Northumbrian coast to make its first English landfall in the district of Cleveland – probably disembarking at the mouth of the Tees – where a landing party is said to have ravaged the countryside without resistance. While this might be recognised as a ‘run ashore’ to replenish supplies after days at sea and in the way of the old viking strandhögg, the saga reference to Harald having ‘subjugated the whole district’ might equally be read to indicate it as an exercise in intimidation akin to his regular summer-raiding of the Danes. Unless the same raiding band stayed on shore to advance southwards overland, the fleet’s next landfall would seem to have been made at Scarborough – a place-name originating as ‘Skarthi’s burg’ and said to commemorate one Thorgil called Skarthi (‘the hare-lipped’) who had established his viking fortress there a full century before Harald’s arrival – and this one is still more reminiscent of that earlier campaigning in Denmark. Meeting with determined resistance on the part of the townsfolk, Harald’s weapon of choice on this occasion, as so often in the past, was fire and a great pyre was lit on the higher ground above the town, from which flaming logs were pitchforked down on to the roofs. Houses went afire one after another until Skarthi’s old burg was so greatly damaged – even ‘destroyed’ according to the saga – that the raiders were able to descend upon it for the usual plunder and slaughter. Curiously, nothing of this is recorded in the early English sources and yet it is said still to survive as a horror story in local tradition.
‘In this manner,’ says the saga, ‘King Harald subdued the country wherever he went.’ So he was to do further down the coast in the district of Holderness, a place-name indicating the district as the domain of a hold, a title reserved for an important local magnate and evidently one with forces ready to meet the invader in arms when Snorri tells of Harald’s men having defeated them in battle. The landing parties would have been back aboard their ships before the fleet rounded Ravenspur (now Spurn Head) to sail up the Humber estuary until it entered the River Ouse and there put into land at a place identified – but only by John of Worcester and his very early twelfth-century contemporary, Simeon of Durham – as Richale (now Riccall).
Harald was now within reach of his principal objective, because the Ouse flows through York, a city of some fifteen thousand households defended by such fortified walls as befitted the capital fortress of the vast Northumbrian earldom and yet still very much a trading centre of similar Scandinavian character to that of Dublin beyond the Irish Sea. Capture of York must have been the first major objective of the expedition, and certainly on the part of Tostig who would have known the city, its people and its surrounding shires very well since it had been his official power base for a full decade, thus raising the question as to why the invasion fleet should have come to rest at Riccall rather than pressing on up the Ouse, even to York itself. The most likely explanation would seem to be the presence of an element of the English fleet, or possibly the naval component of the Northumbrian earl’s forces, which would appear – from a reference in the Anglo-Saxon Chronicle – to have been up the River Wharfe, where it might well have taken refuge on learning of the appearance of so great a Norwegian force approaching the Humber. It would seem, then, that Harald probably placed his own fleet at Riccall to trap these English ships in the Wharfe, while still enabling disembarkation of his troops at a point just some seven miles’ march from the city.
It is possible, had he been so optimistic, that Tostig might have hoped to gain immediate entry to his former capital and the submission of its townsfolk on his arrival with an impressive Norwegian army. It is certainly no less likely that he would have given Harald just such an impression, and yet his own arrival off the Tyne with a mere dozen ships would not have inspired confidence in such assurances, especially when Morcar and Edwin had their combined forces within the city walls, presumably inspired with confidence after their success against Tostig in Lincolnshire. While the alarming news of this quite unexpected invasion had been despatched to Harold Godwinson as soon as the Norwegian Harald landed at Riccall, the great forces he had mustered against Tostig’s earlier attacks on the south coast are known to have been disbanded on 8 September (and thus, according to John of Worcester and Simeon of Durham, only shortly before the Norwegian fleet reached Tynemouth).
Probably thinking that York was unable to hold out against the expected onslaught or siege long enough for a royal relief force to be assembled and to accomplish the long march north, Edwin and Morcar appear to have decided to march their forces out of the city and confront the enemy in open battle, possibly even in the belief that just such a strategy had more often favoured English than Scandinavian forces in the past. The saga seems to indicate Harald’s forces disembarking from th
eir ships at Riccall and almost immediately advancing along the road to York, but it is more likely that the landing of so many troops with their equipment and supplies from some three hundred craft would have taken rather longer, and that the advance along the Ouse would not have begun until at least a day or two after the disembarkation. Edwin and Morcar, meanwhile, had assuredly stationed a watch along the obvious route of approach while holding their forces in readiness for warning of the approaching enemy.
While so much of this preliminary detail is left obscure in all the sources, the account of the conflict itself as set out by Snorri in his Harald’s saga is possessed of an unusual clarity. He is even able to agree with the English sources on the precise date of the battle when he assigns it to the ‘Wednesday before Saint Matthew’s Day’ or Wednesday 20 September in the modern calendar. It is to just one English authority however – namely, the locally well-informed Simeon of Durham – that history is indebted for the specific location of the site of battle ‘at Fulford near York on the northern bank of the Ouse’. The name Fulford apparently does mean ‘foul ford’ and is said by a local historian to refer to ‘the foul or muddy beck which feeds into the river here’,6 yet in older times there were two Fulfords along this bank of the river, one of them called Water Fulford and the other Fulford Gate (or sometimes ‘Gate Fulford’) which has given its name to the battle fought along the stretch of riverbank known as Fulford Ings.
Simeon’s use of the term ‘northern bank’ is somewhat misleading because it can only apply to the city of York rather than to the battlefield, which was clearly on the southern bank of the river as it flowed along the left of Harald’s advance from Riccall. By the time his army reached the vicinity of the Fulford and had thus come within a couple of miles of the city, the Northumbrian and Mercian armies were formed up behind their shield-wall to block the invaders’ path. Snorri tells how Harald drew up his troops across a broad front, with its left flank reaching down to the river and the other towards a dike (or ditch) which appears to have extended into a deep, wide swamp. Harald’s standard was raised on the riverside flank ‘where his forces stood thickest’ while the thinner ranks of his less reliable troops (presumably those with Tostig) stood over by the dike.