DOMESDAY SURVEY the exception made in favour of the abbot's demesne — presumably to the abbot. In dealing with the fiscal side of the survey in Sussex — and Domesday is of course essentially a fiscal record — one is met at once vi^ith an exceedmgly remarkable difficulty. No axiom of Domesday research is more elementary than that four virgates make one hide, and yet in Sussex we are compelled to suspect and almost to assert that eight virgates went to the hide. The question has been debated by Professor Tait and the present writer in the English Historical Review,^ and here it will be sufficient just to indicate the nature of the evidence on either side. Briefly, then, the existence of the eight-virgate hide is supported by : (i) a number of entries of holdings of ' half a hide and 2 virgates,' ' 4 hides and 5 virgates,' ' 3 hides less 2 virgates ' ; and the case of Brightling which was assessed at one hide, and where ' of this hide Robert holds 4 virgates ' ; all of which suggest a hide of more than four virgates : (ii) the clear statement of the chronicler of Battle Abbey ^ that ' eight virgates make one hide,' supported by the assessment of the liberty of Battle Abbey, which is stated in the chronicle to be 6 hides and half a virgate, and is found to consist of 3I hides and 2o| virgates : (iii) the case of' Francwelle,' which is assessed at 2 hides, and was composed of i hide and 8 virgates ; also several other cases pointing in the same direction but not clear and indisputable. Against this must be set : (i) the cases of ' Werste,' where 4 hides and 3 virgates and 5 virgates appear to make 6 hides ; and Boxgrove, where 6 hides seem to be composed of 5^ hides and 2 virgates : (ii) the probability that the fiscal units of hide and virgate would bear the same relation to one another in Sussex as in other counties. This last argu- ment is indeed so strong that, although a good case can be made out for the 8-virgate hide, I hesitate to go further than to suggest its probable existence." Professor Tait's suggestion that the difficulty may be due to ' beneficial hidation ' — the total assessment of the manor being reduced but that of the subtenures retained — is strengthened by the evidence we have produced for the geld having been paid by manors and not by vills, and is further supported by the evidence of pre-Conquest ' beneficial hidation ' which has now to be considered. In the case of several midland counties — Cambridge, Northants, Bedfordshire — Mr. Round has, by means of some very pretty little addition sums, proved the existence of the ' five-hide unit ' in the assessment of vills. No such unit is observable in Sussex " — nor indeed • Sept. 1903 ; Jan. and July, 1904. 2 Dugdale, Mon. iii. 24.2. 3 Domesday Book says 6^ Iiides. < At the same time it is clear that the existence of a hide containing double the usual number of virgates would not cause any confusion in the collecting of the geld, as the assessors — working upon the unit of the hide — would assess the fractions of the hide at twice the usual number of virgates ; thus half a hide would equate with 4 virgates in Sussex instead of with 2 as in other counties, and so on. 6 Mr. Round however holds that such assessments as those of Alciston (50), Lyminster (20), Mailing (80), Stanmer (20), Pagham (50), Tangmere (10), Bishopstone (25), Henfield (25), Selsey (10), Preston (20), Donnington (5), BexhiU (20), Wilcsham (15), and Sherrington (5), show distinct traces of the five- hide unit. 359
