However this may be, the trifling duties imposed on freeholds lead to the inference that the agreement between lord and tenant had been made on the basis of the latter's independent right, and not on that of the lord's will and power. They testify to a subjection of free people and not to the liberation of serfs. And as they are found constantly allied with shareholding, we have to say that they imply manorial relations superimposed on a community which, if not entirely free, contained free elements within it. The manorial duties are more varied and capricious than are the shares just because they are a later growth. I should not like to leave this intricate inquiry without testing its results by yet another standard. I have been trying to prove two things: that some of the feudal freeholds are ancient freeholds, not liberated from servitude but originally based on the recognised right of the holders; that such ancient freeholds were included in the communal arrangement of ownership, although the assessment of their duties was not communal. To what extent are these propositions supported by an analysis of that admittedly ancient tenure, the tenure of the socmen? We must look chiefly to the 'free' socmen; but I may be allowed, on the strength of the chapter on Ancient Demesne, to take the bond socmen also into account. Let us take the manor of Chesterton, in Cambridgeshire[760]. It is royal, but let out in feefarm to the Prior of Barnwell, and its men make use of the _parvum breve de recto_. There is one free tenant of eighty-eight acres holding _de antiquitate_ and the Scholars of Merton hold forty-four acres freely. They have clearly taken the place of some freeman, whether by purchase or by gift I do not know; they are bound to perform ploughings and to carry corn. Both tenements are worthy of notice because charters are not mentioned and still the holdings are set apart from the rest. In the one case the tenure is expressly stated to be an ancient one, and presumably the title of the other tenement is of the same kind. The number of acres is peculiar and points to some agrarian division of which eighty-eight and forty-four were fractions or multiples. The bulk of the population are described as customers. They used to hold half-virgates, it is said, but some of them have sold part of their land according to the custom of the manor. And so their tenements have lost their original regularity of construction, although it seems possible to fix the average holdings at twelve or fifteen acres. Anyhow, it is impossible to reduce them to fractions of eighty-eight; for some reason or another, the reckoning is made on a different basis. The duties vary a good deal, and it would be even more difficult to conjecture what the original services may have been than to make out the size of the virgate.