448 REVIEWS OF BOOKS July he investigate the exact degree of resemblance between the Flemish and the German castellan. He contents himself with quoting Rietschel's suggestion that the German Burggrafenamt may be to some extent of Flemish origin.^ Still on grounds of chronology alone it is impossible to suppose that the counts of Flanders adapted a German model. The judicial powers of the castellan are well illustrated by the docu- ments relating to Bruges and Ghent. Bruges was the administrative capital for the whole of the primitive Pagus Flandrensis ; and there existed in the castle at Bruges a court-house known as the domus scabinatus, which was used as late as 1127 for meetings of the court of the Pagus. The keure of Bruges, granted in 1190 — a keure to the inhabitants of the whole castelry — stipulates for the holding of three placita in each year ; at one of these, called the goudinga, the count or the castellan shall be present. From this and other passages in the keure, which is a custumal rather than a grant of new liberties, it appears that the castellan represented the count in matters of police and justice ; he was in fact as in name a vice-comes (pp. 23-9). Ghent was the capital of a group of old territorial divisions, the Pagus Gandensis, the Pagus Wasiensis, and the district known as the Quatuor Officia. The last of these three was a reclaimed area,2 which at first formed part of the Pagu^ Wasiensis, but was recognized as a separate entity, for jurisdictional purposes, in a keure of 1242 (pp. 84- 85). In the thirteenth century the castellan had lost all jurisdiction over the Pagus Wasiensis in its diminished extent, and retained very little authority in the Quatuor Officia. But even at that date it was the rule that certain sublimia negotia might be heard and determined, if the count or the castellan so desired, by a court of echevins, drawn from the whole of the castelry, which sat in the precinct of the old castle at Ghent, although the castle itself had been demolished and a new residence for the count had been constructed on another site (pp. 65-8, 84-5). But it must not be inferred that every castelry corresponded to a pagus or a group of pagi and that every castellan was the president of a goudinga. M. Blommaert conjectures that the casteliies of Ypres, Dixmude, and Courtrai, and those of the Pagus Mempiscus represented ancient centenae, though he gives no evidence in support of his hypothesis (p. 231). At Lille and at Saint- Omer we find traces of old popular courts which appear to be something more than hundred-courts. The Timallum of Lille was in the thirteenth century a court composed of all allodial proprietors in the castelry, and met three times in the year to deal exclusively with cases relating to allodial lands. The suitors of the court are called scabini and the name appears to be a corruption of thiotmaUi (i.e. the popular assembly). Of this court the castellan remained the president until about the year 1250 ; he was then superseded by the count's bailiff (pp. 148-50). M. Blommaert does not expressly state his views as to the original character of the TimaUum, ' S. Bietschel, Daa Burggrafenamt und die hohe Oerichtsbarkeit (1905).
- See the map of the castelry of Ghent in Havenith, Slvde aur la Rigion de la Basse-
plaine Flamande (Lidge, 1907). In constructing this map M. Havenith has been unduly influenced by the theory of Vanderkindere and Willems that Ghent was originally held by a line of imperial coimts. This theory, which rests on the statements of Thilrode of St. Bavon, a thirteenth-century chronicler, is rejected by M. Blommaert (p. 40).
