Varro speaks of its being an open point, to be decided according to the circumstances of each farm, whether it were better to till the land by slaves or by freemen, or by _both_.[373] And Columella, speaking of the families or 'hands' upon a farm, says 'they are either slaves or _coloni_;'[374] and he goes on to say, 'It is pleasanter to deal with coloni, and easier to get out of them _work_ than _payments_. . . . They will sooner ask to be let off the one than the other. The best _coloni_,' he says, 'are those which are _indigeni_, born on the estate and bound by hereditary ties to it.' Especially distant corn farms, he considers, are cultivated with less trouble by free coloni than by slaves under a villicus, because slaves are dishonest and lazy, neglect the cattle, and waste the produce; [p267] whilst _coloni_, sharing in the produce, have a joint interest with their lord. «_Adscripti glebæ._» That the _coloni_ sometimes were _indigeni_ upon the estate, and were sometimes called _originarii_, shows the beginning at least of a tendency to treat them as _adscripti glebæ_, like the mediæval 'nativi.' Indeed, we find it laid down in the later laws of the empire that _coloni_ leaving their lord's estate could be reclaimed at any time within thirty years.[375] And nothing could more clearly indicate the growth of the semi-servile condition of the colonus, as time went on, than the declaration (A.D. 531) that the son of a colonus who had done no service to the 'dominus terræ' during his father's lifetime, and had been absent more than thirty or forty years, could be recalled upon his father's death and obliged to continue the services due from the holding.[376] We know from Tacitus that the typical colonus had his own homestead and land allotted to his use, and paid tribute to his lord in corn or cattle, or other produce. And there is a clause in the Justinian Code prohibiting the arbitrary increase of these tributes, another point in which the _coloni_ resembled the later _villani_.[377] «Likeness to a manor.» «Village round a villa.» A villa under a villicus, with servi under him living within the 'curtis' of the villa, and with a little group of coloni in their _vicus_ also upon the estate, but outside the court, would thus be very much like a later manor indeed. And Frontinus,[378] describing [p268] the great extent of the _latifundia_, especially of provincial landowners, expressly says that on some of these private estates there was quite a population of rustics, and that often there were _villages_ surrounding the _villa_ like fortifications. It would seem then that the villas in the provinces were still more like manors than those in Italy. «The villa becoming the prevalent type of estate.»