The middest of this shire [Wiltshire] which for the most part lieth plain and even, is divided athwart from East to West, with a dyke of wonderful1 work, cast up for many miles together in length: the people dwelling thereabout call it Wansdyke, which upon an error generally received, they talk and tell to have been made by the Divell upon a Wednesday.’ So wrote John Aubrey in 1663 in his Monumenta Britannica and, since his day, Wansdyke, one of the two mightiest linear earthworks of ancient Britain—the other is Offa’s Dyke—has rightly attracted the admiration and speculation of innumerable antiquaries (PLATE X (a) ).

Dr gordon has, in my opinion, not begun to prove his case and it is difficult to know where to begin its refutation. In at least one major respect the two articles he has written in ANTIQUITY disagree with each other. In the first article (ANTIQUITY XXXI, 124) the frequent sign L 98 was taken (on the basis of the similarity of the Linear B sign) to represent ku or gu. With the help of this value three words have been recognized as Semitic. But in the second article (ANTIQUITY XXXI, 237) the value of the sign is stated as an unknown ‘X’, and then, in order to make it possible to identify a word for ‘cumulative total’ as the Akkadian kitmuru, it is allotted the sound mu. We are not told what we must now think of two other words previously identified on the assumption that the sign stood for ku. Nor is it made clear on what grounds one may pick and choose among the phonetic values of the Linear B signs. L 98 is certainly the same sign as the Linear B 81 ku; L 22 is the same sign as the Linear B 2 ro. Why are we justified in assuming that the one phonetic value is transferable and not the other? This point is quite fundamental in any attempt to decipher Linear A by using the Linear B phonetic values, but Gordon ignores it completely. And not only here. By identifying the supposed word for ‘cumulative total℉ as kitmuru, he commits himself to the value ki for the first sign L 21. As this sign hardly, if ever, occurs elsewhere he cannot check his assumption, though it compels him to alter the value of L 103 (= Linear B 67 ki) to ku (para. 6, p. 239). Allotting the value tu to L 39 in the same word leads to a similar difficulty, since the value tu is already occupied by L 6 (accepted by Gordon in the first article, para. 10).

The significance of the find lies both in its size and in its spanning of the late 9th and early 10th centuries, a period earlier than that embraced by the very few other Late Saxon hoards from this area. Its importance is further enhanced by the fact that it is believed to have been recovered in its entirety, whereas a majority of hoards of this period are feared to have suffered some dispersal at least before they could be recorded and assessed by numismatists.

The new hoard has come to light in the grounds of Wymondham College in the parish of Morley St Peter some twelve miles south-west of Norwich (Nat. Grid. 62/076984). During January 1958 trenches were dug for new foundations, but heavy snow impeded progress, and its melting caused the side of one trench to collapse. When two bricklayers began to clean up the trench on the morning of 27 January, fragments of a broken pot and a shower of silver ‘discs’ fell into the trench between them at a depth of 2 ft. 3 in. To the finders the coins looked like ‘mineral-water bottle caps’, but the men noticed that some bore a portrait, and sought the help of one of the masters at the College. He at once appreciated their importance, and, suspecting that they were Saxon coins, conveyed 874 of them to the headmaster, Mr R. V. Metcalfe, who promptly telephoned news of the discovery to the Castle Museum, Norwich.

The long barrow at Nutbane, Penton Grafton, Andover, Hants, was excavated during the autumn of 1957 by the writer on behalf of the Andover History Group. The barrow (map reference SU 330495) was not included in the map of Neolithic Wessex and had only been discovered two or three years previously, when the attention of Ordnance Survey officers was drawn to its presence.

The barrow had been considerably denuded by ploughing over many years. It is built along the contour of a gentle slope, part of the gradual incline from the chalk ridge five miles to the north down to the valley of the Anton River two miles to the south. It is orientated East-West, with the higher and broader end to the east.

Ghirza is the site of a Roman frontier settlement in the pre-desert zone of Tripolitania in North Africa, 5 miles south-west of the confluence of the Wadi Ghirza and the Wadi Zemzem, 80 miles from the coast of the Gulf of Sidra and 150 miles (in a straight line) south-east of Tripoli. It is best known for its two groups of monumental tombs decorated with rough sculptured reliefs showing the life of the inhabitants; but it is also notable for other antiquities, and especially for its ‘town’ of fortified farmhouses, clustered together on the edge of the wadi.