English: Incomplete 9th century steel-bladed Viking sword of Petersen type K (or Wheeler type IV). Examples can be seen in Peirce, 2002, Swords of the Viking Age, The Boydell Press, Woodbridge. The sword is incomplete and heavily corroded, but enough survives to make it an interesting example. It weighs 442g. The upper hilt consists of an upper guard which is sub rectangular in plan, having slightly curved corners, and rectangular in section. It is 62.94mm long, 26mm wide and 11.74mm thick. It is pierced to either end with a circular perforation, approximately 6mm in diameter, where the pommel was affixed to the guard by rivets. Both the pommel and rivets are entirely missing. The grip of the sword is missing, as this was probably made of bone or another organic material that has since rotted away. What is exposed is the rectangular tang of the sword, which tapers towards the upper guard. It is 87.28mm long, 31.24mm wide tapering to 21.16mm wide and is 4.44mm thick. The lower guard or cross guard is an enlarged version of the upper guard, with the same curved corners, but minus any perforations. It is 112.32mm long, 19mm wide and 12.78mm thick. Corrosion has obscured the slots on both the upper and lower guards, so it is not possible to see with definition where the tang has passed through. However, though the hilt is missing, there is an oblong impressed line around the tang to the under side of the upper guard and upper side of the lower guard where the tubular hilt was once positioned.

The blade of the sword is very damaged and truncated, the point of the blade being entirely missing. It has a shallow bend in the blade of approximately 10 degrees, approximately 75mm down from the lower guard. Its surviving reconstructed length is 320mm. It is 52.68mm wide just below the upper guard (its widest point) the blade is 5.1mm at its thickest point. Other blades of Petersen type K swords have a blade length of up to 800mm, suggesting that just over half of this blade may be missing (Peirce, 2002, page 72). .The blade is wasted irregularly along its width and is also of irregular thickness. The fuller is just about visible, this being the broad shallow groove along the blade edge to reduce the weight of the sword without foregoing strength. It appears to be of uniform width, approximately 17mm, though the state of the blade makes this impossible to confirm with certainty. The cutting edges are missing, but it presumably had a double cutting edge which is common for swords of this type.

The blade seems to be parallel edged, but it may have had a gentle taper which would be consistent with swords of this date. Under Geibig's classification, this would make it a blade type 2 or 3 (Peirce,2002, page 22).

The original surface of the sword is entirely missing and the exposed iron is flaky and brittle. The sword is dark orangey brown with corrosion. There is a longitudinal 'tear' for about 33mm in the centre of the blade just beyond the point of the bend. X-rays of the sword reveal that there is no inscription on the guards, which sometimes occurs, and that the blade is pattern-welded as would be expected.

Barry Ager of the British Museum kindly informed the recorder that the sword is probably of Rhenish manufacture (i.e. Germanic) as Rhenish smiths were known for producing the best blades (Peirce, 2002, page 3). However it is likely that the Scandinavians copied the form and this may be one such imitation. Whatever its manufacturing provenance, it was most likely lost by a Viking warrior from Norway, as this seems to be a favoured Norwegian rather than Danish type (Peirce, 2002, page 4). M. Müller-Wille in Hoops (ed), 2006, Reallexikon der Germanischen Altertumskunde, volume 31, page 398, illustrates a distribution table for type K swords which clearly shows a Norwegian biased distribution and interestingly no known British examples, which could make this find quite unusual.

Examples of swords from England where the type is questionable, but they are possibly of type K, are a sword dredged from the River Kennet at Reading, Berkshire, ), described as 'possibly Wheeler type IV' (Grove, 1938, Five Viking-Period Swords in The Antiquaries Journal, volume XVIII, page 252, figure 1) and a sword from excavations at the Roman fort at Richborough, Kent, described as 'the Saxon sword', but recognised to be possibly type K or M (Evison, 1968, in Cunliffe (ed) Fifth report on the excavations of the Roman Fort at Richborough, Kent, page 114-116). Here Evison discusses the Rhenish connection and suggests that find may have been an Anglo-Saxon copy, but Barry Ager thinks this is probably not the case (pers.com.).

During the 9th century, Colchester was part of the 'Danelaw', that part of England under Danish administration. In AD 917 the king of England, Edward the Elder, the son of Alfred the Great, attacked Colchester and recaptured it from the Danes for the English. This 10th century event in Colchester would post date the manufacture of the sword found there, but we do know that the Vikings passed down their swords from generation to generation and Norse literature gives some swords the name 'Ancient Heirloom'. It is possible that some of these swords were used over a couple of centuries (Peirce, 2002, page 1). It would be tempting to think that this sword had found its way into the river at Colchester either during a skirmish as part of that 917 event, or because its owner deliberately deposited it in the water as a gift to the gods to secure a Viking victory that ultimately did not come. Other, earlier 8th century Saxon weapons were recovered from the river in 1916 and it does seem to have been a place of deliberate deposition in the Early Medieval period (Crummy, 1981, Aspects of Anglo-Saxon and Norman Colchester, Colchester Archaeological Report 1, CBA Research report 39, page 19), however it must remain open to speculation why this particular weapon was lost to the water.