DOMESDAY SURVEY 104 assigned to it by Ellis, even after deducting eight on the manors he wrongly includes.^^' The rigidity with which the proportion of the bovarii to the demesne ploughs is preserved is very remarkable. Under Shropshire we have under Mortimer's fief bovarii mentioned on nine estates in the Domesday Hundred of Leintwardine. On five there are two bovarii to one demesne plough, on one three to one and a half, on two four to two, on one six to three. As examples of a joint body of oxmen, we have the king's three demesne ploughs at Marden with two bovarii and four serfs, or at Cleeve, where he has four with one bovarius and seven serfs."^ On the other hand, of his two manors at Kingsland, one has three demesne ploughs with six serfs, and the other five with ten bovarii}^^ At Stanford again he has three with six bovarii}^^ It is needless to multiply instances. One must note, however, that the proportion of the serfs to the demesne ploughs is nothing like so con- stant as that of the oxmen proper (bovarii). Thus, on the fief of Alvred de Marlborough, bovarii are mentioned on three estates, and serfs alone on four. On the former the proportion is four to two, five to two and a half, six to three ; ^^ on the latter it is four to two in two instances, but uneven in the other three. Whether there was a real distinction between bovarii and serfs, and if so, what it was, is a question discussed by Professor Tait.^" It is complicated by the occurrence in our county of eleven bovarii liberi and of one in Shrop- shire. As the former are found within the compass of two columns,^'^ the liber may be only one of the Domesday scribe's pleonastic vagaries."' On one point, however, there need be no hesitation. When Professor Tait endea- vours to found a distinction on ' the fact ' stated by Professor VinogradofF ' that the ancillae, or women slaves, are never associated with the bovarii^ "* one has only to turn to two entries in his own county of Shropshire "^ or to one in Worcestershire "° to find the bovarii associated with ancillae in the same way as the servi. But even more destructive to this statement is the evidence found in Herefordshire, where the two are associated in four,"' if not in eight, entries."^ There remain the serfs and the ancillae. As already observed, the pro- portion of serfs to the ploughs on the lord's demesne, although implying that their chief employment was to act as oxmen to these ploughs, is so frequently above or below the ratio of two to one — to say nothing of the cases in which no serfs are entered — that they must, where there was excess, have been '" See note 146, above. '" Fol. 179^. '"Ibid. '"Fol. 180. '™ It is important to observe that in this manor there vyere also six serfs. '" V.C.H. Shrops. i, 302-3. His criticism should, however, be compared with what I have actually said in F.C.H. Wore, i, 274-6, where I held it probable, of the bovarii, ' that some were still of servile status, though others were free and paid chevage.' "' One of them and three serfs to two demesne ploughs ; three and one serf to two ; six to three • one and one serf to one. "^ Prof. Tait appears to take this view and to think that the liber may perhaps make no distinction. '" I have not succeeded in finding this statement in the latter's work, but his meaning is at times obscure. '" ' In dominio sunt vi car. et xii bovarii et una ancilla.' ' In dominio sunt iiii car. et viii bovarii et ii ancillae ' (fol. 25 3^). '^* ' Ibi iiii bovarii et una ancilla ' (fol. 177,^). '" 'Ibi iiii bovarii et viii ancillae' (fol. 180). 'In dominio sunt ii car. et iiii bovarii et una ancilla' (fol. 181). 'Ibi iiii bovarii et iii ancillae' (fol. 183). 'Ibi iiii bovarii et una ancilla' (fol. 185). "* In these additional four entries (of which one relates to Eldersfield and another to Hartley in Wor- cestershire) serfs, ancillae, and bovarii are grouped together (fol. 179^ bis, 180, i8o3). I 289 27
