farmers, this factor too would have tended to produce higher assessments north of the Downs. Yet despite these difficulties in using the subsidy to show the distribution of population, the contrast in the number of taxable people between the various regions of the county is great enough to suggest that it reflects differences in the density of the population.
   The population was smallest in the western Weald. Taking the hundreds of Washlingstone and Somerden, and the lowy of Tonbridge together, there were under four and a half taxable inhabitants for each one thousand acres. Settlement in this part of the Wealden forest began late, probably not before the tenth and eleventh centuries, and paled forests and other woodlands may have covered as much as half the area as late as the sixteenth century.
   In the remainder of the Weald between Brenchley in the east and Tenterden in the west settlement had begun as early as the seventh and eighth centuries, and forest clearance was probably far more extensive by the fourteenth century. Yet the number of taxable persons for every one thousand acres, in eight hundreds taken together, was only eight, a considerably smaller figure than in any region in the county outside the Weald.¹  It was in this area that the next two hundred years were to see the greatest change: in the middle of the sixteenth century, when the cloth industry was it its height, the Cranbrook area in the 

central Weald had the highest population density of any region of rural Kent.
   One of the most interesting features of the subsidy is the large number of taxable people listed under the hundreds lying either on the border of or within Romney Marsh. All the later hundreds were included, and none had fewer than fifty-five names, suggesting that the enclosed area of the marshland was probably little less than its modern extent. In eight hundreds taken together there were approximately ten taxable persons for every one thousand acres, reflecting a population density considerably higher than in the Weald and hardly smaller than in northwest Kent or on the central and eastern downlands.²   By the sixteenth century the contrast with the Weald was reversed, and the Marsh had the sparsest population of any area in the county.³  If, as seems almost certain, the change reflects an actual decline in the number of inhabitants of the region, one is led to speculate as to the cause of the
  ¹  The hundreds were Barkley, Tenterden, Rolvenden, Blackburn, Cranbrook, East Barnfield, Seibrittenden, and Brenchley.
  ²  The hundreds were St. Martin, Langport, Aloesbridge, Newchurch, Oxney, Street, Ham and Worth.
  ³  The comparisons with sixteenth century population figures are based on: Bodleian Library, Tanner MS. 240: numbers of households and communicants in the parishes of the dioceses of Canterbury, 1~57 and 1569 ; British Museum, Han. MS. 280 : total number of communicants in the dioceses of Canterbury and Rochester, 1603.