The subject of this paper is a forgotten and neglected style of architecture represented to-day by only a handful of monuments scattered among the mountains of northern Ethiopia. This is one of the by-ways of architecture, linked only obscurely with its main high-roads, and leading nowhere, unless it leads at some future time to a revived Ethiopian style. But, like other by-ways, it offers a fascinating field of inquiry, and one so little studied that the new-comer to it is free—for want of any established theories—to interpret it in his own way.

Present knowledge of the so-called Alhambra Vase series suggests that a known total of a dozen or so examples falls short of the vases that exist. A fine though, as usual, imperfect addition to the group attributed to Malaga, excavated at the Carthusian monastery at Jerez de la Frontera in 1930, is now in the Archaeological Museum at Madrid. The Burgio Vase was almost unknown at the moment of its acquisition for the Instituto de Valencia de Don Juan, Madrid, in 1924, although recorded by Michele Amari in 1872, together with its fellow at Mazzara that entered the Palermo Museum in the eighties of the nineteenth century. The ‘Giarra di S. Ugo’ with an immemorial location at the Cistercian church at Novara, near Messina, but a precise état civil going back less than half a century, has more recently asserted itself. If the ‘Giarra’ is ill preserved, the details of its decoration that are available indicate with some likelihood that this is a Hispano-Moresque vase of Alhambra type; though, to judge from its reported dimension as to height, of the second, rather than the first rank. It may have been one of the plusieurs vases découvertes en Sidle in which Girault de Prangey discerned, perhaps not too correctly, the closest analogy of shape, execution, and material to the surviving specimen at Granada.

Shortly after the presentation of the Second Report to the Society, in March 1926, the gravel-workings from which the material for both this and the First Report had been obtained were abandoned, the site of House XX marking the southerly limit of the land leased for exploitation. On a tour, however, of the district south of the Thames between Drayton and Long Wittenham in 1930, I found that a wide drift had been cut leading southwards from the Drayton-Sutton Courtenay road immediately east of Council cottages newly erected, and that new workings had been for some time in progress under the management of the Berkshire County Council, along the eastern boundary of the older pit.

Some years ago the Wellcome Historical Medical Museum, London, acquired the prehistoric collections resulting from the late Dr. Joseph Bailleau's excavations and researches in the Allier, that département which forms part of the old province of Bourbonnais, Central France (fig. I). The most important series is the classic one from a multiple cave, La Grotte des Fées, in the commune of Châtelperron. Stone and bone artifacts from here have long been held to typify the first stage of French Upper Palaeolithic culture. Assuch, the character of the most outstanding part of this industrial output is now familiar from innumerable references based on the writings of the Abbé H. Breuil, who in 1911 drew a number of inferences from the artifacts. These views have been fully supported since by evidence from elsewhere. However, credit is due in the first place to Bailleau for having attracted attention to thearchaeology of Châtelperron. Lecturing in 1866 he first mentioned his researches, and some years later he published a more detailed account of the discoveries, but in the state of knowledge then prevailing the full significance of the objects could not be appreciated. Now, having examined virtually all the known partof the Bailleau collection, and having classified quantities of other material from Châtelperron, added in different ways since Breuil referred to some of the relics, the present writer thinks that new features and certain aspects may usefully be brought to notice. That he can supplement previous communications is due to the opportunity he has had to study the remarkable collection handed over bythe executors of Dr. Bailleau's estate and also to information given him on the Palaeolithic lots by the Abbé G.-H. Pépin, Curé of Neuvy-lès-Moulins (Allier) and M. R. Sadourny, Moulins (Allier).

The Roman town of Silchester, and the traditional site of Calleva Atrebatum, after a respite from excavation for some eighteen years, was again investigated during 1938-9. In 1937 the Duke of Wellington handed over to the Ministry of Works and Buildings for restoration and preservation the stretch of the north wall of the town from the amphitheatre gate to the north gate. After a preliminary clearing of undergrowth and timber, work was started on the masonry during 1938. As this involved trenching along the bank behind the wall, a procedure which inevitably destroys dating evidence, it was felt that before this was completed a series of trenches should be cut through this bank at right angles to the wall to determine what dating evidence and structure existed. Mr. P. K. Baillie-Reynolds, on behalf of the Ministry of Works and Buildings, asked me to supervise this work. Excavations were started in June 1938 and continued for seven weeks. During this time the work on the inner defences was completed, except for the ditches outside the wall which involved a further season's work in April 1939, lasting for five weeks.