The title from Rembrandt's inventory (Note 2), which was first linked with this problematical painting (Fig. I, Note I) by Smith, has hitherto never been questioned, but, broadly speaking, two different interpretations have been suggested for it. Schmidt-Degener connected it with the glorification of the illustrious past of Amsterdam in uniting with other towns to fight for the country's rights and prosperity (Note 4), an idea later amplified by Hellinga on the basis of Carel van Mander and Cesare Ripa (Note 5). The historian Cornelissen, on the other hand, saw it as a reflection of the tensions between Amsterdam and the Stadholder's court around 1640 and an illustration of the significance of Union, Religion, the Militia and Justice for the concord of the state (Note 6). Schmidt-Degener regarded the work as a sketch for a large painting, possibly a militia piece, but others have seen it as a design for a print (Notes 7 -9), on the analogy of some other oil sketches by Rembrandt which are models for etchings. A completely different suggestion made recently is that the painting is at the dead-coloured stage and thus unfinished (Note II). Since the last figure of the date is almost illegible, various dates in the 1640's have been put forward, most scholars opting for Schmidt-Degener's suggestion of 1641. Examinations in connection with the recent restoration of the painting have revealed a number of things which are presented briefly here as a new guide to possible lines of research. The restoration was necessitated by the fact that the varnish had become opaque in places and the excessive retouching had darkened (Figs. 2-4). The cradling on the back had also distorted the panel, which had been planed down to only about 3 mm. The cradling has now been removed and replaced by a lighter system of small oak blocks, which hook on to an aluminium grid (Fig. 5, Note 14). Investigation of the panel showed that the strip about 6 cm wide at the bottom, which Schmidt-Degener thought to have been added during painting, was an integral part of the original panel, with small blocks of wood let in by later hands to prevent the join from opening (Fig. 6) . A dendrochronological examination of the panel showed it to have come from the same oak tree as those of Rembrandt's River Valley with Ruins at Cassel (Br.454) and the Portrait of a Man in Polish Costume dated 1637 in Washington (Br.211). The latest date for the felling of the tree in question is 1634 ±5, so that our painting could have originated earlier than is indicated by the date under the signature. The edges of the painting appear to have been pared of when the panel was planed down, so that the last figure is probably missing altogether (there is no trace of it on the infra-red photograph, Figs. 14 and 15). A further discovery on the panel are the notches at regular intervals round the edges for securing it with pins and nails in a frame, a phenomenon known from other 17th-century panels (Note 18), although not previously encountered in any by Rembrandt. Investigation of the painting technique, produced a surprise in respect of the most recent theories on Rembrandt's technique (Note 19), viz. that alongside and underneath the highlights containing lead white there proved to exist an earlier stage in which the highlights were applied in a thin paint composed of a watery binding medium (animal glue) with chalk. These light touches, scarcely visible in the X-ray photograph (Fig. 7), are clearly part of the first laying-in of the composition, to which lightly sketched brushstrokes in a darker tone may also belong. Thus this largely monochrome painting also has a monochrome sketch underneath, which implies that what we now have is a finished picture. The meticulous detailing and local, use of colour support this theory, as does the extensive use of scratching in the wet paint to create modelling. The sketchy syle was evidently considered sufficientfor the picture's purpose. A further notable phenomenon is that a thin layer of varnish is found between some of the paint layers in those areas which Rembrandt repainted at a later stage, which could indicate that the picture was already regarded as finished before those changes were made. As the restoration proceeded, it became ever clearer that the picture was painted in various states. Strips about 6 cm wide at top and bottom proved to have been left unpainted originally (Fig. 9), for reasons that can only be guessed at. The two main stages comprised first the laying-in of the composition on a brown to pink ground in a light watery paint, with draughtsmanlike lines of brown to black and scratching in various places. The greenish-black shadows, very thick white highlights and the major part of the sky to the left of the tree also belong to this stage. At the second stage Rembrandt returned to the picture, possibly after a considerable period, and worked it up again with broad strokes of a heavier paint. This relates to the strips at top and bottom, the shadows under the battle and above the chain and the area to the right of the tree above the cavalry procession. Repentirs are found in both stages, notably the elimination of a row of escutcheons continuing the series now visible (see Figs. II and 7), while the light cloth to the left of the arms of Amsterdam and the text Soli Deo Gloria appear to be relatively late additions during the first stage. As to the date of the painting, the style alone indicates the earliest possible date in the 1640's, while the first stage could well date from before 1640. Stylistic links between it and the Landscape with the Good Samaritan of 1638 in Cracow (Br.442, Note 27) suggest that clues to the interpretation must be looked for earlier than 1640 and raise the question of a possible connection with a historical event like the entry of Maria de Medici in 1638 (Note 28). As a result of the restoration numerous details in the picture have now become more clearly legible, while the prominence of the arms of Amsterdam is even more apparent, suggesting that Schmidt-Degener's interpretation is the most likely. If the composition is regarded as an allegory on Concord, the battle in the background could be seen as Discord, so that the Concord here might be Ripa's Unione Civile (Note 29), in this case that of Amsterdam, rather than a united campaign against a historical enemy. This does not explain problematical details, such as Justice on the far left, but it might be a fruitful line to follow. While it cannot be ruled out that the picture is a study for a print, its large size, the fact that it is not on paper, like most of the sketches Rembrandt made for etchings, and its relatively rich palette point in another direction. It still seems closest to the type of modello made by Rubens, Lievens and Bol for large decorative projects (Note 30).