Here in fourteen vills we have an average of thirty-two non-servile households for every vill. Now even in our own day a parish with thirty-two houses, though small, is not extremely small. But we should form a wrong picture of the England of the eleventh century if we filled all parts of it with such vills as these. We will take at random fourteen vills in Staffordshire held by Earl Roger[49].
| Non-servile population | Servi | Total | |
| Claverlege | 4 | 0 | 45 |
| Nordlege | 9 | 0 | 9 |
| Alvidelege | 13 | 0 | 13 |
| Halas | 40 | 2 | 42 |
| Chenistelei | 11 | 0 | 11 |
| Otne | 7 | 1 | 8 |
| Nortberie | 20 | 1 | 21 |
| Erlide | 8 | 2 | 10 |
| Gaitone | 16 | 0 | 16 |
| Cressvale | 8 | 0 | 8 |
| Dodintone | 3 | 0 | 3 |
| Modreshale | 5 | 0 | 5 |
| Almentone | 8 | 0 | 8 |
| Metford | 7 | 1 | 8 |
| Total | 200 | 7 | 207 |
Here for fourteen vills we have an average of but fourteen non-servile households and the servi are so few that we may neglect them. We will next look at a page in the survey of Somersetshire which describes certain vills that have fallen to the lot of the bishop of Coutances[50].
| Non-servile population | Servi | Total | |
| Winemeresham | 8 | 3 | 11 |
| Chetenore | 3 | 1 | 4 |
| Widicumbe | 21 | 6 | 27 |
| Harpetrev | 10 | 2 | 12 |
| Hotune | 11 | 0 | 11 |
| Lilebere | 6 | 1 | 7 |
| Wintreth | 4 | 2 | 6 |
| Aisecome | 11 | 7 | 18 |
| Clutone | 22 | 1 | 23 |
| Temesbare | 7 | 3 | 10 |
| Nortone | 16 | 3 | 19 |
| Cliveham | 15 | 1 | 16 |
| Ferenberge | 13 | 6 | 19 |
| Cliveware | 6 | 0 | 6 |
| Total | 153 | 36 | 189 |
Here we have on the average but eleven non-servile households for each village, and even if we suppose each servus to represent a household, we have not fourteen households. Yet smaller vills will be found in Devonshire, many vills in which the total number of the persons mentioned does not exceed ten and near half of these are servi. In Cornwall the townships, if townships we ought to call them, are yet smaller; often we can attribute no more than five or six families to the vill even if we include the servi.
Population of the vills.
Contrast between east and west.
Unless our calculations mislead us, the density of the population in the average vill of a given county varies somewhat directly with the density of the population in that county; at all events we can not say that where vills are populous, vills will be few. As regards this matter no precise results are attainable; our document is full of snares for arithmeticians. Still if for a moment we have recourse to the crude method of dividing the number of acres comprised in a modern county by the number of the persons who are mentioned in the survey of that county, the outcome of our calculation will be remarkable and will point to some broad truth[51]. For Suffolk the quotient is 46 or thereabouts; for Norfolk but little larger[52]; for Essex 61, for Lincoln 67; for Bedford, Berkshire, Northampton, Leicester, Middlesex, Oxford, Kent and Somerset it lies between 70 and 80, for Buckingham, Warwick, Sussex, Wiltshire and Dorset it lies between 80 and 90; Devon, Gloucester, Worcester, Hereford are thinly peopled, Cornwall, Stafford, Shropshire very thinly. Some particular results that we should thus attain would be delusive. Thus we should say that men were sparse in Cambridgeshire, did we not remember that a large part of our modern Cambridgeshire was then a sheet of water. Permanent physical causes interfere with the operation of the general rule. Thus Surrey, with its wide heaths has, as we might expect, but few men to the square mile. Derbyshire has many vills lying waste; Yorkshire is so much wasted that it can give us no valuable result; and again, Yorkshire and Cheshire were larger than they are now, while Rutland and the adjacent counties had not their present boundaries. For all this however, we come to a very general rule:—the density of the population decreases as we pass from east to west. With this we may connect another rule:—land is much more valuable in the east than it is in the west. This matter is indeed hedged in by many thorny questions; still whatever hypothesis we may adopt as to the mode in which land was valued, one general truth comes out pretty plainly, namely, that, economic arrangements being what they were, it was far better to have a team-land in Essex than to have an equal area of arable land in Devon.