A HISTORY OF WORCESTERSHIRE driven by poverty to take the step. Birth, however, was probably its cause in most cases, for the servile status of the serf's children was rigidly enforced. The recognised existence, at Bristol, of a slave mart was only typical of a traffic that must have prevailed in other places also ; at Lewes, which, it may be pointed out, must then have been a port, the toll on the sale of a man was fourpence. Men thus sold as slaves may obviously have come from anywhere, and there was nothing to prevent slaves from Bristol being brought up the Severn to Worcester. We are told of the Domesday ' servus ' that ' earlier and later documents oblige us to think of him as a slave, one who in the main has no legal rights ; he is the theow of the Anglo-Saxon dooms.' ^ The density of the servile population in Devon and Cornwall supports the obvious presumption that the conquered Britons supplied, throughout the West of England, the bulk of the original serfs ; and, in his valuable chapter on the sub- ject. Dr. Andrews pointed out that the earliest gloss for ancilla ' is wyln, and it is also the most frequent, thus showing the use to which the Welsh women were put who were captured in the conquest.'^ But this reduction of the conquered race, in the West of England, to slavery needs, of course, to be carefully distinguished from the subsequent acquisition of serfs by purchase or capture in war. Above all is caution needed in dealing with the bondwomen ('ancillae') of Domesday, of whom Worces- tershire is said to contain the largest number, Mr. Eyton, a great Domesday student, argues, possibly with good reason, that the absence of 'ancills,' on the pages of the Survey, is no evidence of their non- existence : The Ancilla, or female serf, is never spoken of in the Somerset Survey, only once in the Dorset Survey, only once in the Survey of StafiFordshire. What follows ? Surely . . . that in certain counties the serf-wfife w^as hardly ever reckoned among the agricultural staff of an estate.^ It must be remembered that the instructions given, so far as we know them, to the Domesday commissioners* directed a return of the ' servi,' but not of the ' ancillee.' There may therefore have been uncertainty as to whether they ought to be entered or not, and a consequent diversity in practice. Professor Maitland even hints that the serfs themselves may in some districts have been omitted rather than non- existent,' while in others their numbers may be swollen by embracing a wider class." ' Maitland's Domesday Book and Beyond, p. 27.
- The Old English Manor, p. 198.
^ Staffordshire Domesday, p. 6.
- See p. 272 above.
® Domesday Book and Beyond, pp. 23-4. ® ' Nor can we be sure that the enumeration of the servi is always governed by one consistent principle. In the shires of Gloucester, Hereford, and Worcester we read of numerous ancilla — in Worcestershire of 677 servi and loi ancilla — and this may make us think that in this district all the able-bodied serfs are enumerated, whether or no they have cottages to themselves ' [Ibid. p. 34). The Professor's figures, as explained above, are those given in error by Ellis. 278
