The Roman writers on agriculture of the later republic and the imperial period reckon on an average five -modii- of wheat as sufficient to sow a -jugerum-, and the produce as fivefold. The produce of a -heredium- accordingly (even when, without taking into view the space occupied by the dwelling-house and farm-yard, we regard it as entirely arable land, and make no account of years of fallow) amounts to fifty, or deducting the seed forty, modii. For an adult hard-working slave Cato (c. 56) reckons fifty-one -modii-of wheat as the annual consumption. These data enable any one to answer for himself the question whether a Roman family could or could not subsist on the produce of a -heredium-. The attempted proof to the contrary is based on the ground that the slave of later times subsisted more exclusively on corn than the free farmer of the earlier epoch, and that the assumption of a fivefold return is one too low for this earlier epoch; both assumptions are probably correct, but for both there is a limit. Doubtless the subsidiary produce yielded by the arable land itself and by the common pasture, such as figs, vegetables, milk, flesh (especially as derived from the old and zealously pursued rearing of swine), and the like, are specially to be taken into account for the older period; but the older Roman pastoral husbandry, though not unimportant, was withal of subordinate importance, and the chief subsistence of the people was always notoriously grain. We may, moreover, on account of the thoroughness of the earlier cultivation obtain a very considerable increase, especially of the gross produce--and beyond doubt the farmers of this period drew a larger produce from their lands than the great landholders of the later republic and the empire obtained (iii. Latium); but moderation must be exercised in forming such estimates, because we have to deal with a question of averages and with a mode of husbandry conducted neither methodically nor with large capital. The assumption of a tenfold instead of a fivefold return will be the utmost limit, and yet it is far from sufficing. In no case can the enormous deficit, which is left even according to those estimates between the produce of the -heredium- and the requirements of the household, be covered by mere superiority of cultivation. In fact the counter-proof can only be regarded as successful, when it shall have produced a methodical calculation based on rural economics, according to which among a population chiefly subsisting on vegetables the produce of a piece of land of an acre and a quarter proves sufficient on an average for the subsistence of a family.
It is indeed asserted that instances occur even in historical times of colonies founded with allotments of two -jugera-; but the only instance of the kind (Liv. iv. 47) is that of the colony of Labici in the year 336--an instance, which will certainly not be reckoned (by such scholars as are worth the arguing with) to belong to the class of traditions that are trustworthy in their historical details, and which is beset by other very serious difficulties (see book ii. ch. 5, note). It is no doubt true that in the non-colonial assignation of land to the burgesses collectively (-adsignatio viritana-) sometimes only a few -jugera- were granted (as e. g. Liv. viii. ii, 21). In these cases however it was the intention not to create new farms with the allotments, but rather, as a rule, to add to the existing farms new parcels from the conquered lands (comp. C. I. L. i. p. 88). At any rate, any supposition is better than a hypothesis which requires us to believe as it were in a miraculous multiplication of the food of the Roman household. The Roman farmers were far less modest in their requirements than their historiographers; they themselves conceived that they could not subsist even on allotments of seven -jugera- or a produce of one hundred and forty -modii-.
7. I. VI. Time And Occasion Of The Reform
8. Perhaps the latest, although probably not the last, attempt to prove that a Latin farmer's family might have subsisted on two -jugera- of land, finds its chief support in the argument that Varro (de R. R. i. 44, i) reckons the seed requisite for the -jugerum- at five -modii- of wheat but ten -modii- of spelt, and estimates the produce as corresponding to this, whence it is inferred that the cultivation of spelt yielded a produce, if not double, at least considerably higher than that of wheat. But the converse is more correct, and the nominally higher quantity sown and reaped is simply to be explained by the fact that the Romans garnered and sowed the wheat already shelled, but the spelt still in the husk (Pliny, H. N. xviii. 7, 61), which in this case was not separated from the fruit by threshing. For the same reason spelt is at the present day sown twice as thickly as wheat, and gives a produce twice as great by measure, but less after deduction of the husks. According to Wurtemberg estimates furnished to me by G. Hanssen, the average produce of the Wurtemberg -morgen- is reckoned in the case of wheat (with a sowing of 1/4 to 1/2 -scheffel-) at 3 -scheffel- of the medium weight of 275 Ibs. (= 825 Ibs.); in the case of spelt (with a sowing of 1/2 to 1 1/2 -scheffel-) at least 7 -scheffel- of the medium weight of 150 lbs. ( = 1050 Ibs.), which are reduced by shelling to about 4 -scheffel-. Thus spelt compared with wheat yields in the gross more than double, with equally good soil perhaps triple the crop, but--by specific weight--before the shelling not much above, after shelling (as "kernel") less than, the half. It was not by mistake, as has been asserted, but because it was fitting in computations of this sort to start from estimates of a like nature handed down to us, that the calculation instituted above was based on wheat; it may stand, because, when transferred to spelt, it does not essentially differ and the produce rather falls than rises. Spelt is less nice as to soil and climate, and exposed to fewer risks than wheat; but the latter yields on the whole, especially when we take into account the not inconsiderable expenses of shelling, a higher net produce (on an average of fifty years in the district of Frankenthal in Rhenish Bavaria the -malter- of wheat stands at 11 -gulden- 3 krz., the -malter- of spelt at 4 -gulden-30 krz.), and, as in South Germany, where the soil admits, the growing of wheat is preferred and generally with the progress of cultivation comes to supersede that of spelt, so the analogous transition of Italian agriculture from the culture of spelt to that of wheat was undeniably a progress.