The same group appears in manorial documents under the name of _censuarii_[387]. Both terms interchange, and we find the same fluctuation between free and servile condition in regard to the _censuarii_ as in regard to _molmen_. The thirteenth-century extent of the manor of Broughton, belonging to the Abbey of Ramsey in Huntingdonshire, when compared with Domesday, shows clearly the origin of the group and the progress which the peasantry had made in two hundred years. The Domesday description mentions ten sokemen and twenty villains; the thirteenth-century Cartulary speaks in one place of _liberi_ and _villani_, sets out the services due from the latter, but says that the Abbot can 'ponere omnia opera ad censum;' while in another place it speaks as though the whole were held by _liberi et censuarii_[388]. A similar condition is indicated by the term _gavelmanni_, which occurs sometimes, although not so often as either of the designations just mentioned[389]. It comes evidently from _gafol_ or _gafel_, and applies to rent-paying people. It ought to be noticed, however, that if we follow the distinction suggested by the Kentish documents, there would be an important difference in the meaning. Rent need not always appear as a result of commutation; it may be an original incident of the tenure, and there are facts enough to show that lands were held by rent in opposition to service even in early Saxon time. Should _mal_ be taken as a commutation rent, and _gafol_ strictly in the sense of original rent, the gavelmen would present an interesting variation of social grouping as the progeny of ancient rent-holding peasantry. I do not think, however, that we are entitled to press terminological distinctions so closely in the feudal period, and I should never enter a protest against the assumption that most gavelmen were distinguished from molmen only by name, and in fact originated in the same process of commutation. But, granting this, we have to grant something else. _Vice versa_, it is very probable indeed that the groups of _censuarii_ and _molmen_ are not to be taken exclusively as the outcome of commutation. If _gafol_ gets to be rather indistinct in its meaning, so does _mal_, and as to _census_, there is nothing to show whether it arises in consequence of commutation or of original agreement. And so the Kentish distinction, even if not carried out systematically, opens a prospect which may modify considerably the characteristic of the status on which I have been insisting till now. Commutation was undoubtedly a most powerful agency in the process of emancipation; our authorities are very ready to supply us with material in regard to its working, and I do not think that anybody will dispute the intimate connexion between the social divisions under discussion and the transition from labour services to rent. Yet a money rent need not be in every case the result of a commutation of labour services, although such may be its origin in most cases.