Nearly forty years ago Sir Charles Peers read a paper of outstanding importance on Caernarvon Castle. The description he then gave of the various buildings of which the castle is composed is printed in the Transactions of the Cymmrodorion Society. It will never be bettered, combining as it does a mastery of detail with a simple and lucid presentation of a most complicated structure. It is prefaced by an analysis of contemporary documents bearing on the original work of construction, on the gradual progress of building and on the castle's subsequent history. On the strength of these records a concise building sequence is established, and this has in its turn been adopted as the basis of all subsequent writings on the castle. A dated ground-plan, setting out this sequence in graphic form, accompanied the printed account, and has since been widely reproduced.

The obvious English-language literature on the early civilizations of the Ancient East gives less attention than it might to the northern aspects, towards central Asia. Here, one of the leading researchers in Soviet Central Asia reviews an important new French book on the Iranian plateau and its inter-relations.

Profound books are always worthwhile. If we agree with them, their ideas become a direct contribution to our knowledge. If we disagree, they impel us to examine the basis of disagreement and to develop our own counter-ideas, and thus they make an indirect contribution to our knowledge. R. G. Collingwood's The Idea of History (Oxford, Clarendon Press, 1946) is the most penetrating work on the foundations of historical method known to me, but I find it unacceptable. My major difficulties result from his ontological idealism and his dichotomy between mind and nature.

Anyone who tries for long to penetrate the fascinating world of the prehistoric Aegean becomes aware very quickly of a certain malaise in the works of modern scholars. It is now a century since Schliemann first dug at Troy, and the mass of archaeological evidence continues to increase at an almost alarming rate. I say ‘alarming’ because the pace of excavation threatens to outrun our ability to assimilate and control the evidence.

Many years ago I was put in charge of the excavation of an ancient mound-site at Abu Geili on the Blue Nile near Sennar. Careful records were kept, but for reasons wholly outside my control it is only recently that I have been enabled to write a full account for publication. In the course of doing so I have constantly been brought up against first principles; in the present article I propose to consider how these principles apply, as I see it, to Sudanese archaeology at the present time.

We appreciate the emerging discussion on the identification of the recently discovered Chalcolithic artefacts from the Judean Desert, and the careful attention to detail by Langgut et al. (above) that helps to further clarify our current understanding of spinning and metallurgical technologies in this period. Yet while we agree that by its nature archaeology is full of surprises and exceptional discoveries, we argue, however, that especially in such cases as this, the supporting evidence should be robust. This is clearly not the situation here; regardless of Langgut et al.’s nuanced argumentation (above) on specific contextual observations, our interpretation of the lead artefact as a metallic macehead that happened to be found with its wooden shaft still attached (a rare find but paralleled in the hoard from the Cave of the Treasure) remains much simpler and more straightforward than the interpretation that this extremely rare metal was used as part of a mundane spinning implement (which has no parallels anywhere). As the limitation on space does not allow us to address each of the points raised by Langgut et al., we leave it to the reader to assess the accuracy and relevance of their claims. In any case, their detailed response is only tangential to the essential line of our argument, which is related to weighing the available data and contextual information properly. The conclusion remains that while the observations that ostensibly connect the newly discovered artefacts to textile production are feeble and can be simply related to Ghassulian prestige metal objects (e.g. the use of wooden shafts and textiles in their carrying and maintenance), other observations make this connection difficult (as admitted also by Langgut et al. regarding the weight of the lead ‘whorl’), if not impossible.

Brunanburh was fought in the late summer of 937 between king Athelstan and all the might of England on one side, and the viking Olaf Guthfrithson, king of Dublin and claimant to the throne of York, and his ally Constantine II, king of Scots, upon the other. The thousandth anniversary of this British battle of the nations is an occasion for reviewing critically the evidence upon the disputed question of its site. The well-known poem in the Anglo-Saxon Chronicle, being impressionist rather than narrative, provides only two or three slight clues; the Latin poem preserved by William of Malmesbury,l of early but uncertain date, helps us only a little further; and the Northumbrian annals used by Symeon of Durham seem to have supplied him with an alternative name for the battlefield but no other indication of its locality. Judgment must therefore be based largely upon the names by which the battle was known and upon traditions preserved by writers who had no claim to be contemporary; and the reliability of such evidence must be tested.