We may perhaps state the problem thus. In the mediæval manor there were two elements, the _seigneurial_--the relations of the tenants to the lord; and the _communal_--the relations of the tenants to one another. The mark theory taught that the seigneurial was grafted on to the communal. The value of the work of M. Fustel de Coulanges and of Mr. Seebohm is in showing that we cannot find a time when the seigneurial element was absent; and also in pointing to reasons, in my opinion conclusive, for connecting that element with the Roman villa. But the communal element is still an unsolved mystery. Among the difficulties which lie on the surface in M. Fustel’s treatment of the question, it may be worth while to mention two. He insists that the _villa_ itself, from the earliest time at which it appears, has a unity which it retains throughout.[49] This seems to suggest some earlier economic formation out of which it arose; for if the villas were originally nothing more than private estates, like the estates formed in a new country in our own day, they would hardly have had such a fixity of outline. Then, again, nothing is more characteristic of the later manor than the _week-works_, the labour performed by each villein for two or three days every week on the lord’s demesne. But such week-works do not appear in mediæval documents until A.D. 622.[50] M. Fustel hardly realises that a fact like this requires explanation; or, indeed, that such services were far more onerous than any he describes in the case of the earlier _coloni_. Difficulties such as these can only be satisfactorily overcome by taking into account both sides of the subject--the economic as well as the constitutional or legal. Side by side with a development which combined together gangs of slaves and the households of dependent coloni into the homogeneous class of serfs, and then went on to make out of the mediæval serf the modern freeman, another series of changes was going on of which M. Fustel de Coulanges says nothing. It was the development from a “wild field grass husbandry,” where a different part of the area in occupation was broken up for cultivation from time to time, to the “three-field system” with its permanent arable land pasture, and then again from that to the “convertible husbandry” and the “rotation of crops” of more recent times. The task for the economic historian is to put these two developments into their due relation the one to the other.