Besides these, there are two other forms of tombs which, so far as is yet known, are quite peculiar to the Scandinavian province. The first of these are the so-called ship graves, from their form. They consist of two segments of a circle joined together at the ends, so as to represent the deck of a vessel, and are of all sizes, from 20 or 30 feet to 200 or 300 feet. They are generally found on the sea-shore, and it seems hardly to be doubted that they mark the graves of Vikings.
The other form is quite as peculiar, but more difficult to explain. It is marked by a range of stones forming an equilateral triangle, sometimes straight-lined, but as frequently the lines curve inwards so as to restrict the internal space considerably. It is by no means clear what suggested this form, or what it was intended to represent. It is, however, found on battle-fields ([woodcut No. 118]), and solitary examples are frequent in Sjöborg's plates, sometimes with a Bauta stone in the centre. The one hypothesis that seems to account for this form, is that it is the "Cuneatus ordo" of Olaus Magnus, and that it marked a spot where a combined phalanx of horse and foot fought and conquered.[368] The probability is that where single it marks the grave of a particular rank either in the army or in civil life.
All these forms are shown in the next woodcut, from a group found in the peninsula of Hjortehammer, in Bleking, in the south of Sweden, but others are found in the island at Amrom, and in many other places.[369] It has been disputed whether these represent battle-fields or are the ordinary graves of the inhabitants of the district in which they are found. That those found on the shore at Freyrsö ([woodcut No. 101]) mark the graves of those who fell in Blodoxe's battle there in the tenth century seems quite certain, but whether this was always the case may be open to doubt; but certainly a sandy peninsula, like that of Hjortehammer, seems a most unlikely place for peaceful men to bury their dead, especially at a time when not one-tenth part of the land around could have been under cultivation.
118. Graves at Hjortehammer. From Worsaae.
For our present purposes it is of no great consequence which opinion prevails, as these forms have no bearing on those of other countries, especially as their date does not seem to be doubted. Worsaae places them all between the years 700 and 1000,[370] or in the second and latest Iron age, and as no one seems to dispute this, it may be accepted as an established fact. Their peculiarities of form, and the smallness of the stones of which most of them are composed, are such that the date here ascribed to them does not necessarily bring down that of the true megalithic remains to anything like the same age. It takes away, however, all improbability from the assertion that these may be much more modern than was supposed, and this much is certain that there was no break between the great English and Irish circles and the Viking graves; or, in other words, men did not cease to mark their sepulchres with circles and cairns, and then after a lapse of centuries revive the custom, and begin it again on a smaller scale. There may be a descent, but there was no solution of continuity, and any one can consequently form an idea how long a time must have elapsed before the great Wiltshire circles could have degenerated into those of Hjortehammer.
119. Circles at Aschenrade. From Bähr.