COMMODUS TO DIOCLETIAN (continued 2)

So far we have been viewing agriculture as proceeding in times and under conditions assumed to be more or less normal, without taking account of the various disturbing elements in rustic life, by which both landlords and tenants were liable to suffer vexation and loss. Yet these were not a few. Even a lawyer could not ignore wild beasts. Wolves carried off some of A’s pigs. Dogs kept by B, colonus of a neighbouring villa, for protection of his own flocks, rescued the pigs. A legal question[1503] at once arises: are the rescued pigs regarded as wild game, and therefore belonging to the owner of the dogs? No, says the jurist. They were still within reach; A had not given them up for lost; if B tries to retain them, the law provides remedies to make him give them up. I presume that B would have a claim to some reward for his services. But the lawyer is silent, confining his opinion to the one question of property. References to depredations of robbers or brigands (latrones, grassatores,) occur often, and quite as a matter of course. The police of rural Italy, not to mention the Provinces, was an old scandal. Stock-thieves, who lifted a farmer’s cattle sheep or goats, and sometimes his crops, were important enough to have a descriptive name (abigei)[1504] and a title of the Digest to themselves. That bad neighbours made themselves unpleasant in many ways, and that their presence gave a bad name to properties near them, was an experience of all lands and all ages: but the jurists treat it gravely[1505] as a lawyer’s matter. Concealment of such a detrimental fact[1506] by the seller of an estate made the sale voidable. The rich (old offenders in this kind) were by a rescript of Hadrian[1507] awarded differential punishment for removing landmarks: in their case the purpose of encroachment was not a matter open to doubt.

In one connexion the use of force as an embarrassing feature of rustic life was a subject of peculiar interest to the jurists, and had long been so. This was in relation to questions of possession. In Roman law possessio held a very important place. All that need be said of it here is that the fact of possession, or lack of it, seriously affected the position of litigants in disputes as to property. Great ingenuity was exercised in definition and in laying down rules for ascertaining the fact. Now among the means employed in gaining or recovering possession none was more striking or more effective than the use of force. Special legal remedies had been provided to deal with such violence; interdicta issued by the praetor, to forbid it, or to reinstate a claimant dislodged by his rival, or simply to state the exact issue raised in a particular case. On conformity or disobedience to the praetor’s order the case was formally tried in court: the question of law mainly turned on questions of fact. What concerns us is that force was solemnly classified under two heads, vis and vis armata. Each of these had its own proper interdict at least as early as the time of Cicero, and they occupy a whole title[1508] in the Digest. Clearly the use of force was no negligible matter. That it was a danger or at least a nuisance to owners or claimants of property, is not less clear. But how did it touch the colonus? He was, as such, neither owner nor claimant of the property of his farm. He had in his own capacity[1509] no possession either. But, as tenant of a particular owner, his presence operated[1510] to secure the possession of his landlord. Hence to oust him by force broke the landlord’s possession; whether rightly or wrongly, the law had to decide. Now it is obvious that, in cases where serious affrays resulted from intrusion, a tenant might suffer grave damage to his goods and person. The intruders (often a gang of slaves) would seldom be so punctiliously gentle as to do no harm at all. Therefore, having regard to the amount of interest in this subject shewn by the lawyers, we cannot omit the use of force in matters of possession from the list of rustic embarrassments.

Another cause of annoyance was connected with servitudes, such as rights of way and water, which were frequent subjects of dispute in country districts. Whether regarded as rights or as burdens, the principles governing them were a topic that engaged the minute and laborious attention[1511] of the lawyers. Now it is evident that a right of way or water through an estate, though a material advantage to a neighbouring estate served by the convenience, might be a material disadvantage to the one over which the right extended. Also that the annoyance might be indefinitely increased or lessened by the cantankerous or considerate user of the right by the person or persons enjoying it. When we consider that servitudes were already an important department of jurisprudence in Republican days, and see how great a space they occupy in the Digest, we can hardly resist the conclusion that country proprietors found in them a fertile subject of quarrels. But surely the quarrels of landlords over a matter of this kind could not be carried on without occasional and perhaps frequent disturbances and injury to the tenants on the land. Even if the law provided means of getting compensation for any damage done to a tenant’s crops or other goods in the course of attempts to enforce or defeat a claimed servitude, was the average colonus a man readily to seek compensation in the law-courts? I think not. But, if not, he would depend solely on the goodwill of his own landlord, supposing the latter to have got the upper hand in the main dispute. On the whole, I strongly suspect that in practice these quarrels over rustic servitudes were a greater nuisance to farmers than might be supposed. So far as I know, we have no statement of the farmer’s point of view. Another intermittent but damaging occurrence was the occasional passage of soldiery, whose discipline was often lax. We might easily forget the depredations and general misconduct of these unruly ruffians, and imagine that such annoyances only became noticeable in a later period. But the jurists do not allow us to forget[1512] the military requisitions for supply of troops on the march, the payment for which is not clearly provided, and would at best be a cause of trouble; or the pilferings of the men, compensation for which was probably not to be had. It would be farmers in northern Italy and the frontier-provinces that were the chief sufferers.

Damage by natural disturbances or by fires may happen in any age or country. That Italy in particular was exposed to the effect of floods and earthquakes, we know. Accordingly the lawyers are seriously concerned with the legal and equitable questions arising out of such events. It was not merely the claim of tenants[1513] to abatement of rent that called for a statement of principles. Beside the sudden effects of earthquakes torrents or fires, there were the slower processes of streams changing their courses[1514] and gradual land-slides on the slopes of hills. These movements generally affected the proprietary relations of neighbouring landlords, taking away land from one, sometimes giving to another. Here was a fine opening for ingenious jurists, of which they took full advantage. The growth of estates by alluvion, and loss by erosion, was a favourite topic, the operation of which, and the questions thereby raised, are so earnestly treated as to shew their great importance in country life. Of fire-damage, due to malice or neglect, no more need be said; nor of many other minor matters.

But, when all the above drawbacks have been allowed for, it is still probably true that scarcity of labour was a far greater difficulty for farmers. We hear very little directly of this trouble, as it raised no point of law. Very significant[1515] however are the attempts of the Senate and certain emperors to put down an inveterate scandal which is surely good indirect evidence of the scarcity. It consisted in the harbouring[1516] of runaway slaves on the estates of other landlords. A runaway from one estate was of course not protected and fed on another estate from motives of philanthropy. The slave would be well aware that severe punishment awaited him if recovered by his owner, and therefore be willing to work for a new master who might, if displeased, surrender him any day. The landlords guilty of this treason to the interests of their class were probably the same as those who harboured[1517] brigands, another practice injurious to peaceful agriculture both in Italy and abroad. Another inconvenience, affecting all trades and all parts of the empire in various degrees, was the local difference in the money-value[1518] of commodities in different markets. This was sometimes great: and that it was troublesome to farmers may be inferred from the particular mention of wine oil and corn as cases in point. No doubt dealers had the advantage over producers, as they generally have, through possessing a more than local knowledge of necessary facts. These middlemen however could not be dispensed with, as experience shewed, and one of the later jurists[1519] openly recognized. Facilities for borrowing, and rates of interest, varied greatly in various centres. But all these market questions do not seem to have been so acute as to be a public danger until the ruinous debasement of the currency in the time of Gallienus. A few references may be found to peculiar usages of country life in particular Provinces. Thus we read that in Arabia[1520] farms were sometimes ‘boycotted,’ any person cultivating such a farm being threatened with assassination. In Egypt[1521] special care had to be taken to protect the dykes regulating the distribution of Nile water. Both these offences were summarily dealt with by the provincial governor, and the penalty was death. Here we have one more proof of the anxiety of the imperial government to insure the greatest possible production of food. The empire was always hungry,—and so were the barbarians. And the northern frontier provinces could not feed both themselves and the armies.

While speaking of landlords and tenants we must not forget that all over the empire considerable areas of land were owned by municipalities, and dealt with at the discretion of the local authorities. Variety of systems was no doubt dictated by variety of local circumstances: but one characteristic was so general as to deserve special attention on the part of jurists. This was the system of perpetual leaseholds[1522] at a fixed (and undoubtedly beneficial) rent, heritable and transferable to assigns. So long as the tenant regularly paid the vectigal, his occupation was not to be disturbed. It was evidently the desire of the municipal authorities to have a certain income to reckon with: for the sake of certainty they would put up with something less than a rack-rent. There were also other lands owned by these civitates that were let on the system[1523] in use by private landlords; the normal term probably being five years. Of these no more need be said here. Beneficial leases under a municipality were liable to corrupt management. It had been found necessary[1524] to disqualify members of the local Senate (decuriones) from holding such leases, that they might not share out the common lands among themselves on beneficial terms. But this prohibition was not enough. The town worthies put in men of straw[1525] as nominal tenants, through whom they enjoyed the benefits of the leases. So this evasion also had to be met by revoking the ill-gotten privilege. But disturbance of tenancies was not to be lightly allowed, so it appears that a reference to the emperor[1526] was necessary before such revocation could take place. This system of perpetual leases is of interest, not as indicating different methods of cultivation from those practised on private estates, but as betraying a tendency to fixity[1527] already existing, destined to spread and to take other forms, and to become the fatal characteristic of the later Empire. Another striking piece of evidence in the same direction occurs in connexion with the lessees (publicani) of various state dues (vectigalia publica) farmed out in the usual way. In the first half of the third century the jurist Paulus attests[1528] the fact that, in case it was found that the right of collecting such dues, hitherto very profitable to the lessees, could only be let at a lower lump sum than hitherto, the old lessees were held bound to continue their contract at the old price. But Callistratus, contemporary or nearly so, tells us that this was not so, and quotes[1529] a rescript of Hadrian (117-138 AD) condemning the practice as tyrannical and likely to deter men from entering into so treacherous a bargain. It appears that other[1530] emperors had forbidden it, but there is no proof that they succeeded in stopping it. At all events the resort to coercion in a matter of contract like this reveals the presence of a belief in compulsory fixity, ominous of the coming imperial paralysis, though of course not so understood at the time. It did not directly affect agriculture as yet; but its application to agriculture was destined to be a symptom and a cause of the empire’s decline and fall.

Another group of tenancies, the number and importance of which was quietly increasing, was that known as praedia Caesaris[1531], fundi fiscales, and so forth. We need not discuss the departmental differences and various names of these estates. The tenants, whether small men or conductores on a large scale who sublet in parcels[1532] to coloni, held either directly or indirectly from the emperor. We have seen specimens in Africa, the Province in which the crown-properties were exceptionally large. What chiefly concerns us here is the imperial land-policy. It seems clear that its first aim was to keep these estates permanently occupied by good solvent tenants. The surest means to this end was to give these estates a good name, to create a general impression that on imperial farms a man had a better chance of thriving than on those of average private landlords. Now the ‘state,’ that is the emperor or his departmental chiefs, could favour crown-tenants in various ways without making a material sacrifice of a financial kind. In particular, the treatment of crown-estates as what we call ‘peculiars,’ in which local disputes were settled, not by resort to the courts of ordinary law, but administratively[1533] by the emperor’s procuratores, was probably a great relief; above all to the humbler coloni, whom we may surely assume to have been a class averse to litigation. No doubt a procurator might be corrupted and unjust. But he was probably far more effectually watched than ordinary magistrates; and, if the worst came to the worst, there was as we have seen the hope of a successful appeal to the emperor. Another favour consisted in the exemption of Caesar’s tenants from various burdensome official duties in municipalities, the so-called munera, which often entailed great expense. This is mentioned by a jurist[1534] near the end of the second century: they are only to perform such duties so far as not to cause loss to the treasury. Another[1535], somewhat later, says that their exemption is granted in order that they may be more suitable tenants of treasury-farms. This exemption is one more evidence of the well-known fact that in this age municipal offices were beginning to be evaded[1536] as ruinous, and no longer sought as an honour. We must note that, if this immunitas relieved the crown-tenants, it left all the more burdens to be borne by those who enjoyed no such relief. And this cannot have been good for agriculture in general.

It is not to be supposed that the fiscus[1537] was a slack and easy landlord. Goods of debtors were promptly seized to cover liabilities: attempts to evade payment of tributa by a private agreement[1538] between mortgagor and mortgagee were quashed: a rescript[1539] of Marcus and Verus insisted on the treasury share (½) of treasure trove: and so on. But there are signs of a reasonable and considerate policy, in not pressing demands so as to inflict hardship. Trajan[1540] had set a good example, and good emperors followed it. We may fairly guess that this moderation in financial dealings was not wholly laid aside in the management of imperial estates. Nor is it to be imagined that the advantages of imperial tenants were exactly the same in all parts of the empire. In Provinces through which armies had to move it is probable that coloni Caesaris would suffer less[1541] than ordinary farmers from military annoyances. But on the routes to and from a seat of war it is obvious that the imperial post-service would be subjected to exceptional strain. Now this service was at the best of times[1542] a cause of vexations and losses to the farmers along the line of traffic. The staff made good all deficiencies in their requirements by taking beasts fodder vehicles etc wherever they could find them: what they restored was much the worse for wear, and compensation, if ever got, was tardy and inadequate. The repair of roads was another pretext for exaction. It is hardly to be doubted that in these respects imperial tenants suffered less than others. Some emperors[1543] took steps to ease the burden, which had been found too oppressive to the roadside estates. But this seems to have been no more than relief from official requisitions: irregular ‘commandeering’ was the worst evil, and we have no reason to think that it was effectually suppressed. It appears in the next period as a rampant abuse, vainly forbidden by the laws of the Theodosian code.

L. THE LATER COLONATE, ITS PLACE IN ROMAN HISTORY.

In the endeavour to extract from scattered and fragmentary evidence some notion of agricultural conditions in the Roman empire before and after Diocletian we are left with two imperfect pictures, so strongly contrasted as to suggest a suspicion of their truth. We can hardly believe that the system known as the later Colonate appeared in full force as a sudden phenomenon. Nor indeed are we compelled to fly so directly in the face of historical experience. That we have no narrative of the steps that led to this momentous change, is surely due to the inability of contemporaries to discern the future effect of tendencies operating silently[1544] and piecemeal. What seems at the moment insignificant, even if observed, is seldom recorded, and very seldom intentionally. Hence after generations, seeking to trace effects to causes, are puzzled by defects of record. Their only resource is to supplement, so far as possible, defective record by general consideration of the history of the time in question and cautious inference therefrom: in fact to get at the true meaning of fragmentary admissions in relation to their historical setting. The chief topic to be dealt with here from this point of view is the character of the Roman Empire in several aspects. For among all the anxieties of the government during these troubled centuries the one that never ceased was the fear of failure in supplies of food.

The character of the Roman Empire had been largely determined by the fact that it arose from the overthrow of a government that had long been practically aristocratic. The popular movements that contributed to this result only revealed the impossibility of establishing anything like a democracy, and the unreality of any power save the power of the sword. The great dissembler Augustus concealed a virtual autocracy by conciliatory handling of the remains of the nobility. But the Senate, to which he left or gave many powers, was never capable of bearing a vital part in the administration, and its influence continued to dwindle under his successors. The master of the army was the master of the empire, and influence was more and more vested in those who were able to guide his policy. That these might be, and sometimes were, not born Romans at all, but imperial freedmen generally of Greek or mixed-Greek origin, was a very significant fact. In particular, it marked and encouraged the growth of departmental bureaus, permanent and efficient beyond the standard of previous Roman experience. But the price of this efficiency was centralization, a condition that carried with it inevitable dangers, owing to the vast extent of the empire. In modern times the fashionable remedy suggested for over-centralization is devolution of powers to local governments controlling areas of considerable size. Or, in cases of aggregation, the existing powers left to states merged in a confederation are considerable. In any case, the subordinate units are free to act within their several limited spheres, and the central government respects their ‘autonomy,’ only interfering in emergencies to enforce the fulfilment of definite common obligations.

But, if it had been desired to gain any such relief by a system of devolution within the Roman empire, this would have meant the recognition of ‘autonomy’ in the Provinces. And this was inconceivable. The extension of Roman dominion had been achieved by dividing Rome’s adversaries. Once conquered, it was the interest or policy of the central power to keep them in hand by preventing the growth of self-conscious cohesion in the several units. Each Province was, as the word implied, a department of the Roman system, ruled by a succession of Roman governors. It looked to Rome for orders, for redress of grievances, for protection at need. If the advance of Rome destroyed no true nations, her government at least made the development of truly national characteristics impossible, while she herself formed no Roman nation. Thus, for better or worse, the empire was non-national. But, as we have already seen, the decline of Italy made it more and more clear that the strength of the empire lay in the Provinces. Now, having no share in initiative and no responsibility, the Provinces steadily lost vitality under Roman civilization, and became more and more helplessly dependent on the central power. As the strain on the empire became greater, the possibility of relief by devolution grew less: but more centralization was no cure for what was already a disease.

That local government of a kind existed in the empire is true enough; also that it was one of the most striking and important features of the system. But it was municipal, and tended rather to subdivide than to unite. It was the outcome of a civilization profoundly urban in its origins and ideas. The notion that a city was a state was by no means confined to the independent cities of early Greece. Whether it voluntarily merged itself in a League or lived on as a subordinate unit in the system of a dominant power, the city and its territory were politically one. Within their several boundaries the townsmen and rustic citizens of each city were subject to the authorities of that community. Beyond their own boundary they were aliens under the authorities of another city. It is no wonder that jealousies between neighbour cities were often extreme, and that Roman intervention was often needed to keep the peace between rivals. But the system suited Roman policy. In the East and wherever cities existed they were taken over as administrative units and as convenient centres of taxation: in the West it was found useful and practicable to introduce urban centres into tribes and cantons, and even in certain districts to attach[1545] local populations to existing cities as dependent hamlets. And, so long as the imperial government was able to guard the frontiers and avert the shock of disturbances of the Roman peace, the empire held its own in apparent prosperity. To some historians the period of the ‘Antonines’ (say about 100-170 AD) has seemed a sort of Golden Age. But signs are not lacking that the municipal system had seen its best days. The severe strain on imperial resources in the time of Marcus left behind it general exhaustion. The decay of local patriotism marked the pressure of poverty and loss of vitality in the cities. More and more their importance became that of mere taxation-centres, in which the evasion of duty was the chief preoccupation: they could not reinvigorate the empire, nor the empire them.

Another characteristic of the empire, not less significant than those mentioned above, was this: taken as a whole, it was non-industrial. Manufactures existed here and there, and products of various kinds were exchanged between various parts of the empire. So far as the ordinary population was concerned, the Roman world might well have supplied its own needs. But this was not enough. The armies, though perilously small for the work they had to do, were a heavy burden. The imperial civil service as it became more elaborate did not become less costly. The waste of resources on unremunerative buildings and shows in cities, above all in Rome, and the ceaseless expense of feeding a worthless rabble, were a serious drain: ordained by established custom, maintained by vanity, to economize on these follies would seem a confession of weakness. Nor should the extravagance of the rich, and of many emperors, be forgotten: this created a demand for luxuries chiefly imported from the East; precious stones, delicate fabrics, spices, perfumes, rare woods, ivory, and so forth. Rome had no goods to export in payment for such things, and the scarcity of return-cargoes must have added heavily to the cost of carriage. There was on this account a steady drain of specie to the East, and this had to be met by a corresponding drain of specie to Rome. In one form or another this meant money drawn from the Provinces, for which the Provinces received hardly the bare pretence of an equivalent, or a better security for peace.

Thus the empire, created by conquest and absorption, administered by bureaucratic centralization, rested on force; a force partly real and still present, partly traditional, derived from a victorious past. The belief in Rome as the eternal city went for much, and we hear of no misgivings as to the soundness of a civilization which expressed itself in a constant excess of consumption over production. Naturally enough, under such conditions, the imperial system became more and more what it really was from the first, a vast machine. It was not a league of cooperating units, each containing a vital principle of growth, and furnishing the power of recovery from disaster. Its apathetic parts looked passively to the centre for guidance or relief, depending on the perfection of a government whose imperfection was assured by attempting a task beyond the reach of human faculty and virtue. The exposure of the empire’s weakness came about through collision with the forces of northern barbarism. What a machine could do, that it did, and its final failure was due to maladies that made vain all efforts to renew its internal strength.

The wars with the northern barbarians brought out with singular clearness two important facts, already known but not sufficiently taken into account. First, that the enemy were increasing in numbers while the people of the empire were in most parts stationary or even declining. Bloody victories, when gained, did practically nothing to redress the balance. Secondly, that at the back of this embarrassing situation lay a food-question of extreme seriousness and complexity. More and more food was needed for the armies, and the rustics of the empire, even when fitted for military service, could not be spared from the farms without danger to the food-supply. The demands of the commissariat were probably far greater than we might on the face of it suppose; for an advance into the enemy’s territory did not ease matters. Little or nothing was found to eat: indeed it was the pressure of a growing population on the means of subsistence that drove the hungry German tribes to face the Roman sword in quest of abundant food and the wine and oil of the South and West. The attempt of Marcus and others after him, to solve the problems of the moment by enlisting barbarians in Roman armies, was no permanent solution. The aliens too had to be fed, and their pay in money could not be deferred. Meanwhile the taxation of the empire inevitably grew, and the productive industries had to stagger along under heavier burdens. The progressive increase of these is sufficiently illustrated in the history of indictiones. At first an indictio was no more than an occasional[1546] impost of so much corn levied by imperial proclamation on landed properties in order to meet exceptional scarcity in Rome. But it was in addition to the regular tributum, and was of course most likely to occur in years when scarcity prevailed. No wonder it was already felt onerous[1547] in the time of Trajan. Pressure on imperial resources caused it not only to become more frequent, and eventually normal: it was extended[1548] to include other products, and became a regular burden of almost universal application, and ended by furnishing a new chronological unit, the Indiction-period of 15 years.

That agriculture, already none too prosperous, suffered heavily under this capricious impost in the second century, seems to me a fact beyond all doubt. And, not being then a general imperial tax, it fell upon those provinces that were still flourishing producers of corn. Debasement of currency already lowered the value of money-taxes, and tempted emperors to extend the system of dues in kind. Under Diocletian and Galerius things came to a head. Vast increase of taxation was called for under the new system, and it was mainly taxation in kind. Already the failure of agriculture was notorious, and attempts had been made to enforce cultivation of derelict lands. The new taxation only aggravated present evils, and in despair of milder measures Constantine attached the coloni to the soil. Important as the legal foundation of the later serf-colonate, this law is historically still more important as a recognition of past failure which nothing had availed to check. He saw no way of preventing a general stampede from the farms save to forbid it as illegal, and to employ the whole machinery of the empire in enforcing the new law. This policy was only a part of the general tendency to fix everything in a rigid framework, to make all occupations hereditary, that became normal in the later Empire. The Codes are a standing record of the principle that the remedy for failure of legislation was more legislation of the same kind. Hard-pressed emperors needed all the resources they could muster, particularly food. They had no breathing-space to try whether more freedom might not promote enterprise and increase production, even had such a policy come within their view. Hence the cramping crystallizing process went on with the certainty of fate. The government, unable to develope existing industry, simply squeezed it to exhaustion.

How came it that the government was able to do this? How came it that agricultural tenants could be converted into stationary serfs without causing a general upheaval[1549] and immediate dissolution of the empire? Mainly, I think, because the act of Constantine was no more than a recognition de iure of a condition already created de facto by a long course of servilizing influences. Also because it was the apparent interest, not only of the imperial treasury but of the great proprietors generally, to tie down to the soil[1550] the cultivators of their estates. Labour was now more valuable than land. In corn-growing Africa the importance attached to the task-work of sub-tenants was a confession of this. And, law or no law, things had to move in one or other direction. Either the landlord and head-lessee had to win further control of the tenants, or the tenants must become less dependent. Only the former alternative was possible in the circumstances; and the full meaning of the change that turned de facto dependence into legal constraint may be stated as a recognition of the colonus as labourer rather than tenant. Whether the settlement of barbarians as domiciled aliens in some Provinces under strict conditions of farm-labour had anything to do with the creation of this new semi-servile status, seems hardly to be decided on defective evidence. At all events it cannot have hindered it. And we must make full allowance for the effect of various conditions in various Provinces. If we rightly suppose that the position of coloni had been growing weaker for some time before the act of Constantine, this does not imply that the process was due to the same causes operating alike in all parts of the empire in the same degree. The evidence of the Theodosian Code shews many local differences of phenomena in the fourth and fifth centuries; and it is not credible that there was a greater uniformity in the conditions of the preceding age. Laws might aim at uniformity, but they could not alter facts.

My conclusion therefore is that the general character of the imperial system was the main cause of the later serf-colonate. However much the degradation of free farm-tenants, or the admission of slaves to tenancies, or the settlement of barbarians under conditions of service, may have contributed to the result, it was the mechanical nature of the system as a whole that gave effect to them all. After Trajan the rulers of the empire became more and more conscious that the problem before them was one of conservation, and that extension was at an end. Hadrian saw this, and strove to perfect the internal organization. By the time of Aurelian it was found necessary to surrender territory as a further measure of security. We can hardly doubt that under such conditions the machine of internal administration operated more mechanically than ever. Then, when the reforms of Diocletian made fresh taxation necessary to defray their cost, an agricultural crisis was produced by the turning of the imperial screw. The hierarchy of officials justified their existence by squeezing an assured revenue out of a population unable to resist but able to remove. There was no other source of revenue to take the place of the land: moreover, it was agricultural produce in kind that was required. Therefore the central bureaucracy, unchecked by any public opinion, did after its wont. In that selfish and servile world each one took care of his own skin. Compulsion was the rule: the coloni must be made to produce food: therefore they must be bound fast to the soil, or the empire would starve—and the officials with it.