¹ “Stagnation” is taken in the general sense of an “inactive, sluggish, dull condition” rather than in the more restrictive meaning sometimes ascribed to it by economic historians of a condition in which opportunity for investment is so glutted as to make the movement of capital sluggish. The author of the other paper on the program chose the phrase “re-tardative factors” to avoid possible ambiguity.

² The following works have been particularly used in the preparation of this article:

Frank, T., ed., An Economic Survey of Ancient Rome (Baltimore: The Johns Hopkins Press): Vol. I (1933)Google Scholar, Frank, T., Rome and. Italy of the Republic; Vol. II (1936)Google Scholar, Johnson, A. C., Roman Egypt to the Reign of Diocletian; Vol. III (1937)Google Scholar, R. C. Collingwood, Roman Britain, J. J. Van Nostrand, Roman Spain, V. M. Scramuzza, Roman Sicily, Grenier, A., La Gaule romaine; Vol. IV (1938)Google Scholar, R. M. Haywood, Roman Africa, F. M. Heichel-heim, Roman Syria, J. A. O. Larsen, Roman Greece, Broughton, T. R. S., Roman Asia; Vol. V (1940)Google Scholar, T. Frank, Rome and Italy of the Empire. The Danubian provinces have not been covered. There is a General Index to Vols. I-V (1940). This work will be referred to hereafter as Economic Survey.

Rostovtzeff, M., The Social and Economic History of the Roman Empire (Oxford: Clarendon Press, 1926)Google Scholar, is referred to as , Rostovtzeff, Roman Empire. There is a German edition (2 vols.; Leipzig: Quelle and Meyer, 1931) and an Italian (Florence: La Nuova Italia, 1933).Google ScholarHeichelheim, F. M., Wirtschaftsgeschichte des Allertums (2 vols.; Leiden: Sijthoff, 1938)Google Scholar; Vol. I (text) is referred to as Heichelheim, I, and Vol. II (notes) as Heichelheim, II. Chap, viii deals with the period from Augustus to Diocletian.

The Cambridge Ancient History, Vols. IX-XI I (Cambridge: The University Press; New York: The Macmillan Company, 1932-1939) is referred to as C. A. H. Chap, vii of Vol. XII by F. Oertel deals with “The Economic Life of the Empire.”Google ScholarPauly's Real-Encyclopddie der classischen Altertumswissenschaft is referred to by series, volume, and half volume as: RE, I (1), etc., and RE 2,1 (1), etc. The first series begins with “A”; the second series with “R.”

³ The economic historian of the classical world has only scattered and inadequate evidence on which to base his conclusions so that his treatment must inevitably be more general and less statistical than that of the modern economic historian (see Heichelheim, I, 5-7).

⁴ The failure of the Greek genius in practical inventiveness lies outside the scope of an economic discussion. There is a notable distrust of innovation in all ancient thought (cf. the myth of Prometheus). The Romans showed little inventiveness along either cultural or technical lines and their literature adopted the attitude that inventions and economic expansion were responsible for the greed which led to wars and corruption and reduced the primitive age of gold to the contemporary one of iron (see, for example, Horace Odes i. 3; Epodes 7 and 16; Tibullus i. 3. 33-50; and similar passages in other authors). This distrust of inventions is well illustrated by the story which Pliny the Elder tells about Tiberius. When an artisan discovered a method of making flexible glass, Tiberius ordered his whole shop destroyed lest the value of bronze, silver, and gold be diminished.- Nat. Hist, xxxvi. 26 (27). Pliny says the story was better known than certain. Petronius (Satiricon 51) states that Tiberius executed the artisan, as does Dio Cassius (Ivii. 21. 7).

⁵ On inventions and introduction of new plants and products during the Hellenistic period, see Rostovtzeff, M., The Social and Economic History of the Hellenistic World (3 vols.; Oxford: Clarendon Press, 1941), I, 351–80 (Ptolemaic Egypt)Google Scholar; II, 1180-1238 (general summary). For Syria, see Heichelheim in Economic Survey, IV, 126; for Roman science, see Singer, C., “Science,” in Bailey, C., ed., The Legacy of Rome (Oxford: Clarendon Press, ) Pp. 265–324, with bibliographyGoogle Scholar.

⁶ See below, n. 104.

⁷ Rostovtzeff gives the best survey of general agricultural conditions during the early empire.- Roman Empire, chaps, vi and vii, pp. 180-305. See also the appropriate sections of the volumes of the Economic Survey, especially, for the decay of Greece, Vol. IV, 465-921 and, for the decay of Italy, Vol. V, 184, 297.

⁸ , Rostovtzeff (Roman Empire, pp. 329–30) disputes the view of J. Liebig and his followers (see p. 591, n. 30) that soil exhaustion was the cause of the decay of the Roman empire. On pp. 494-96, n. 25, Rostovtzeff denies soil exhaustion even in ItalyGoogle Scholar.

⁹ Heitland, W. E., Agricola (Cambridge: The University Press, 1921)Google Scholar, is still standard for agriculture in the classical world, particularly in Italy. He argued that, as the upper classes acquired wealth during the later republic, they gradually bought out the small farms to form large estates worked by slave gangs and that, when the supply of slaves diminished under the empire, the large landowners substituted tenants (see esp. pp. 205-12, and below, n. 69, for a similar thesis in Barrow, Slavery).

¹⁰ M. Rcstovtzeff (here spelled Rostowzew), Studien zur Geschichte des romischen Kolonates, erstes Beiheft zum Archiv fur Popyrusforschung (Leipzig and Berlin: Teubner, 1910), is still the fundamental study of tenant farming in the ancient world, although his conclusion that the Romans borrowed the system from the Near East is perhaps overdrawn. See Frank in Economic Survey, V, 300-2. See also Heichelheim, I, 744-49, and II, 1171-73, n. 42 (bibliography); and Rostovtzeff, Roman Empire, p. 646, index under Coloni.

¹¹ Rostovtzeff holds that free peasants in Italy survived the growth of large s'ave-worked estates only to become tenants of absentee landlords at a later date.- Roman Empire, pp. 192-93. Frank in Economic Survey (V, 168-75) accepts their survival in Italy during the first century. Scramuzza in Economic Survey (III, 366-67) argues that they were never eliminated in Sicily.

¹² Grenier in Economic Survey (III, 495) states that the domains established in Gaul at the opening of the Roman epoch reproduced to a large extent those of the Celtic aristocracy. Collingwood traces the contrasting villa and village systems in Roman Britain to native origins.- Ibid., pp. 73-87.

¹³ Heichelheim, I, 747, 749. Similarly (pp. 753-59), he sees in this period the beginning of a shift from the distinction between farmers and city dwellers to that between the great landowning officials, the honestio'es, and the tenant serfs, the humitiores, of the later empire, a distinction perpetuated in the feudal societies of both the eastern and tb. western Mediterranean. Oertel accepts this view.- CAM., XII, 281. Heichelheim (I, 671-82) blames Augustus for having failed to carry through the Caesarian program of a classless society and a unified economic system, and for having perpetuated the city-state concept of economic organization, which proved inadequate to maintain itself on an imperial scale (see below, n.85).

¹⁴ , Heitland, Agricola, pp. 432–59Google Scholar. He remarks (pp. 434, 443) on the scorn that the ancients felt for manual labor.

¹⁵ , Frank in Economic Survey, V, 175-82, 297Google Scholar., Broughton in Economic Survey (III, 690-92, 839–40)Google Scholar finds that the amount of agricultural slavery in the province of Asia is hard t o estimate.

¹⁶ , Rostovtzeff (Roman Empire, pp. 319–23)Google Scholar thinks that Hadrian and (p. 357) even Septimius were eager to promote a free peasantry.

¹⁷ Ibid., pp. 296-97. For the wealth of senators, Friedlander, L., Darstellungen aus der Sittengeschichte Roms (9th ed.; Leipzig: Hirzel, 1919-1921), I, 121–35Google Scholar; and for luxury, II, 263-379' Davis, W. S., The Influence of Wealth in Imperial Rome (New York: The Macmillan Company, 1910Google Scholar; reprint, New York: Peter Smith, 1933), is a vivid if popular study n i which the undue pursuit of wealth is blamed for the fall of Rome (see below, n. 104; also , Frank in Economic Survey, V, 22-26, 56–60)Google Scholar.

¹⁸ , Frank in Economic Survey (V, 306–2)Google Scholar doubts that the growth of large estates and tenant farming had serious economic effects before the extensive confiscations of Septimius, which vastly increased the crown property (res priuata). He feels that this caused rivalry for the throne, loss of initiative in agriculture, and increased pressure for production. For Asia, , Broughton in Economic Survey (IV, 587, 648-63, 905–6)Google Scholar reaches the same conclusion. Nostrand, Van in Economic Survey (III, 216)Google Scholar also thinks that imperial properties were not large in Spain until the time of the Severi. It is possible that the imperial properties in Africa were extensive at an earlier date, if there is any truth in Pliny the Elder's remark that Nero put to death six men who owned half of Africa (presumably the old proconsular province, namely, Tunis).- Nat. Hist., xviii. 6 (7). 35Google Scholar. The conclusion is generally drawn that Nero confiscated their estates, though Pliny does not say so and only cites the fact to illustrate the spread in the provinces of large estates, which he condemns as the ruin of Italy. From just outside proconsular Africa come the much discussed inscriptions concerning the management of the Saltus Burunitanus under Trajan, Hadrian, and Commodus. , Hay-wood in Economic Survey (IV, 83–102Google Scholar, esp. pp. 85-86) thinks that, whatever the previous imperial holdings in Africa were, Septimius greatly increased them.

¹⁹ , Rostovtzeff (Roman Empire, pp. 266, 269)Google Scholar thought that concern with abandoned or unreclaimed land became a government policy when Augustus took over the management of Egypt (see his Kolonates, pp. 351). The policy was later applied by Hadrian to the African estates.- Idem, Roman Empire, pp. 321, 330, and 591, n. 31; Kolonates, pp. 391-93. Herodian (ii. 4. 6), quoted in Kolonates (p. 391, n. 1), states that Pertinax, in 193 A.D., generalized the rule that those who would work uncultivated land could possess it. It is, in fact, at about this time, under Marcus, that the settlement of barbarians within the empire begins to become common. , Rostovtzeff (Roman Empire, p. 374)Google Scholar connects this settlement of barbarians with the general policy under the Severi of converting the army into a farming militia, settled on farms around strongholds which served as a rallying point for defensive operations (see pp. 375-79 and below, n. 75).

²⁰ See above, n. 16.

²¹ , Rostovtzeff (Roman Empire, pp. 349, 597, n. 5)Google Scholar compares the threat of the peasants of the Saltus Burunitanus under Coramodus “to flee to some place where we can live as free men” to the traditional “strikes” of Egyptian peasants (see p. 256), who would stage a “secession” to a temple if they felt that some wrong was not being properly redressed. The cases are, however, somewhat different: the Egyptian peasant did not plan to settle somewhere else if he did not get justice; the Africans definitely threatened to depart. The growth of desertion, both of peasants after the time of the Severi and, under the later empire, of the oppressed tax-paying bourgeoisie (see below, n. 92), is a well-known phenomenon. Rostovtzeff connects various “peasant uprisings” of the late second century, notably that of Maternus in Gaul and Spain, with the oppression of tenants (see pp. 327, 357, 424, and 620, n. 16, 436, 438). The government of the fourth century tried to combat the tendency toward “desertion” by binding all classes of the population more strictly to their appointed tasks (see pp. 465-69; see also Johnson in Economic Survey, II, 245-46, for Egypt, and Heichelheim in Ibid., IV, 234, for Syria; Broughton in Ibid., IV, 658-60, for Asia; Oertel in C.A.H., XII, 254-59).

²² For Roman relaxation of Ptolemaic monopolies, see , Rostovtzeff, Roman Empire, pp. 159, 169Google Scholar; Heichelheim, I, 736-37. , Johnson in Economic Survey (II, 325–35)Google Scholar discusses the monopolies but draws no conclusions concerning Roman policies.

²³ For glass, see Heichelheim, II, 1139, n. 14.

²⁴ For Arretine ware, see Heichelheim, II, 1163, n. 32; Comfort in a special section on terra sigillata in Economic Survey, V, 188-94.

²⁵ , Rostovtzeff, Roman Empire, pp. 161–69Google Scholar; Heichelheim, I, 732-40; , Frank in Economic Survey, V, 185–217 (Italy)Google Scholar, 222-29 (Rome).

²⁶ , Rostovtzeff (Roman Empire, p. 167) connects the decline in artistic skill with standardization and decentralization. Standardization should not have caused a decline in artistic skill, because standardization did not really cut the workman off from directly shaping his product. Even in mold-made pottery decorated with reliefs, the mold was handmade and new ones had frequently to be made. Athenian pottery of the sixth to the fourth century B.C. was produced in quantity for export on fairly standard models, but the individual potters and painters achieved a high level of artistry. More properly, Rostovtzeff later (p. 479) connects the decline in industrial art with the general decline of classical civilization. An interesting, but noneconomic, question is that of how far the decline was due to “barbarization,” that is, the spread of classical culture to peoples who, not temperamentally inclined to accept it, therefore handled it without inspiration, whatever their native abilities along the lines of their own native cultures.Google ScholarCollingwood, R. C., in his and Myers', J. N. L.Roman Britain (Oxford: Clarendon Press, 1936Google Scholar; chap, xv, pp. 245-60), maintains that this happened in Britain; that the normal development of native Celtic art was interrupted by the Roman conquest and that the British during the Roman period produced nothing but dull provincial imitations of classical art, but that the native Celtic genius reasserted itself as the Roman hold weakened.

²⁷ For overland trade, see , Rostovtzeff, Roman Empire, pp. 142–61Google Scholar; Heichelheim, I, 690-722; appropriate sections of the Economic Survey; Charlesworth, M. P., Trade-routes and Commerce of the Roman Empire (Cambridge: The University Press, 1924)Google Scholar.

²⁸ For piracy, see Ormerod, H. A., Piracy in the Ancient World (Liverpool: The University Press of Liverpool, 1924)Google Scholar. Piracy was largely suppressed during the later republic, chiefly by -Frank, Pompey in Economic Survey, I, 301–3Google Scholar. For geographic and climatic factors in navigation, see Semple, E. C., The Geography of the Mediterranean Region (New York: Henry Holt and Company, 1931), Part IV, pp. 579–707Google Scholar. Caesar's pursuit of Pompey across the Adriatic in the winter of 49/48 B.C. was seriously, and almost disastrously, interrupted by storms.- Bell. Ciu. iii. 25. According to Tacitus (Ann. xv. 46. 3), Nero in 64 A.D. ordered the Italian squadron to return (presumably from Ostia) to its station at Misenum without regard for the weather. It was caught off Cumae in a southwest (Africus) gale and most of the triremes and lesser craft were blown ashore.

²⁹ Twenty days from Puteoli to Alexandria is the average given by Warmington, E. H., The Commerce between the Roman Empire and India (Cambridge: The University Press, 1928)Google Scholar, in his table on p. 50. Thompson, J. W., An Economic and Social History of the Middle Ages (New York: Century Company, 1928), p. 3Google Scholar, gives twelve days, presumably for a fast voyage.

³⁰ Acts of the Apostles 27-28; Heichelheim, I, 730. Paul's journeys in the east are discussed by , Broughton in Economic Survey (IV, 858–60)Google Scholar.

³¹ According to , Tacitus (Ann. xii. 43)Google Scholar, a shortage of grain in Rome during the winter of 51 A.D. led to a riot in which the mob surrounded Claudius and had to be driven away by troops. Only ten days' supply remained in the city but a fortunate spell of good weather (as well as the kindness of the gods) permitted the ships to come from Africa. Tacitus moralizes that formerly Italy had fed not only itself but distant legions; now it was dependent on Africa and Egypt and the life of the Roman people was subject to the chances of shipping. In 189 A.D., on the occasion of a severe famine, the prefect of the grain supply artificially aggravated it in such a way that the pernicious minister of Commodus, Cleander, should seem responsible. Thereupon, the people besieged Commodus in his villa and, despite the Praetorian Guard (the police), forced him to have Cleander and his son executed.-Dio Cassius lxxii (lxxiii in Boissevain and the Loeb ed.). 13; Scriptores Historiae Augustae, Comm. 14. 1-3; Herodian i. 12. 3-13. 6.

³² , Frank in Economic Survey, V, 267–95Google Scholar.

³³ The general supplying of grain to Rome was called annona (a term applied also to other aspects of the supplying of grain); the distribution of grain free to the poorer people of Rome was called frumentatio (see , Rostovtzefl's article on “Frumentum” in RE, VII (13), 126–87Google Scholar, and the articles in E. de Ruggiero, Dizionario Epigrafico di Antichha Romane, on “Annona” (I, 474-87) and on “Frumentatio” (III, 225-315). The dole, as well as the free gifts of money to the people (congiaria or liberalitates, as against donatiua to troops) are frequently alluded to in the imperial coinage.- Mattingly, H., Roman Coins (London: Methuen, 1928), p. 151Google Scholar. See generally Mattingly, H. and Sydenham, E. A., The Imperial Roman Coins (London: Spink and Sons, 1923) andGoogle ScholarMattingly, H., Coins of the Roman Empire in the British Museum (London: For the Trustees, 1923)Google Scholar.

³⁴ For the view that Rome could not be supported by grain transported overland but must draw from overseas, see H. M. Last's discussion of the grain bill passed by Gaius Gracchus in 122 B.C. in C.A.H., IX, 4, 57-60.

³⁵ , Frank in Economic Survey (V, 218)Google Scholar compares Rome's unfavorable trade balance, met by government expenditures or paid for by income from provincial investments, with those of such modern capitals as Washington or Rome.

³⁶ Even Athens, in the fifth century B.C., routed the import of grain from the Black Sea through the Piraeus, where her citizens had first claim on it. The Gracchan grain law is derived by Last (C.A.H., IX, 57) from Hellenistic precedents.

³⁷ But Rostovtzeff indicates that the administrations of all large cities had to assist and control the food supply.- Roman Empire, pp. 148-53, esp. p. 149, and p. 532, n. 22).

³⁸ For trade outside the empire, see Frank in Economic Survey, V, 283-90; Johnson in Ibid., II, 344-46 (part played by Egypt); Heichelheim in Ibid., IV, 198-201, 203 (part played by Syria). The trade with India and China was very large and is well discussed by Warmington, Commerce. The Chinese sources are translated and discussed by F. Hirth, China and the Roman Orient (Leipzig: G. Hirth, 1885). Hirth tries to identify the countries and products mentioned in the Chinese sources. He thinks (pp. 173-78) that the famous embassy mentioned as having reached the Chinese court in 166 A.D. from “Antun, ruler of Ta-Ts'in in [Syria?]” was not an official embassy from Marcus Aurelius Antoninus but comprised traders who, having reached Annam, pushed into China in the guise of an official embassy. He points out that their gifts were not typically Syrian products but objects that they could easily have procured in Annam. Oertel, however, in C. A. H. (XII, 235) accepts a view that they were an official embassy which had started from Ctesiphon [Ta Ts'in?]. Hirth (pp. 221-41; see below, n. 67) remarks on the high esteem enjoyed by Syrian glass in China. The New York Times (April 26, 1946, p. 23) reports the finding in Indo-China of a hoard of coins and precious objects, partly Greco-Roman and partly Hindu, among which was a coin of Antoninus or Marcus.

³⁹ Diz. Epigraftco, I, 485 (left col.) under “Annona,” citing Codex Theodosianus xiv. 26. 1 with the comment of Gothofredus (see Heichelheim, I, 799).

⁴⁰ Heichelheim (I, 796-808) contrasts the continuance of the luxury trade after the barbarian invasions with the decentralization of trade in cheaper goods and thinks that the Byzantine emperors fostered the former.

⁴¹ Pirenne's, H. view is briefly and recently set forth in his Economic and Social History of Mediaeval Europe (New York: Harcourt, Brace and Company, 1937)Google Scholar, Introduction, pp. 1-15. His view is accepted by , Thompson, Middle Ages, pp. 217–18Google Scholar.

⁴² Perhaps the most famous description of the prosperity of the early empire is Gibbon's, in the first three chapters of his Decline and Fall of the Roman Empire. In the third chapter, after discussing the reign of Marcus, he makes the famous remark: “If a man were called to fix the period in the history of the world during which the condition of the human race was most happy and prosperous, he would, without hesitation, name that which elapsed from the death of Domitian to the accession of Commodus” (see the edition by Bury, London: Methuen, 1896,1, 78; see also Thompson, Middle Ages, chap. i).

⁴³ For Rome, Ostia, and Pompeii, see , Frank in Economic Survey, V, 218–66Google Scholar; for Rome, Carcopino, J., Daily Life in Ancient Rome at the Height of the Empire, ed. Rowell, H. T. (New Haven: Yale University Press, 1940)Google Scholar. In general, see Oertel in C. A. H., XII, 249-50; Rostovtzeff, Roman Empire, passim; Friedlander, Sittengeschichte, passim; Dill, S., Roman Society from Nero to Marcus Aurelius (2d ed.; London: Macmillan and Company, 1905), passimGoogle Scholar.

⁴⁴ The best impression of the character, content, and distribution of Latin inscriptions can be obtained from an examination of Dessau, H., Inscriptiones Latinae Selectae (3 vols. in 5; Berlin: Weidemann, 1892-1916)Google Scholar.

⁴⁵ , Bayiies in C. A. H. (XII, 713)Google Scholar comments on the lack of money for memorials in the third century.

⁴⁶ For Augustus' budget, see , Frank in Economic Survey, V, 4–18Google Scholar. See also in Ibid., pp. 36-56, for the Julio-Claudians and Flavians, and pp. 65-90, for the Antonines and Severi.

⁴⁷ For the flow of accumulated wealth from the east to Italy, see , Rostovtzeff, Roman Empire, pp. 15–22Google Scholar; , Frank in Economic Survey, I, 296-98, 324-26, 338–41Google Scholar; V, 18, 25-26; Johnson in Ibid., II, 481 (Egypt). For the wealth of Gaul, see Grenier in Ibid., III, 455-64. For the wealth of the west generally, see Heichelheim, I, 695-96.

⁴⁸ Augustus set forth his contributions to the state in his own memorial, the Res Cestae (see the summary in Economic Survey, V, 14-15).

⁴⁹ For defeat of Varus, see Parker, H. M. D., The Roman Legions (Oxford: Clarendon Press, 1928), pp. 89–90 (see also below, n. 101). According to Suetonius, Augustus was so upset by the loss of Varus' three legions that for months on end he let his hair and beard grow (a sign of mourning) and occasionally beat his head against doors, crying “Quinctilius Varus, give me back my legions,” and kept the anniversary of the defeat as a day of gloom and mourning.- Aug. 23.2. When the Pannonian legions revolted on the accession of Tiberius in 14 A.D., demanding a reduction of service from twenty to sixteen years (the term for the praetorians) and increased pay, Tiberius refused their demands because the treasury could not stand the expense.-Google Scholar, Parker, Legions, p. 77Google Scholar, on the basis of Tacitus Ann. i.17.6, 78.2.

⁵⁰ The reasons for Claudius' conquest of Britain are much disputed.- , Collingwood and , Myers, Roman Britain, pp. 76–78Google Scholar. Wealth, exaggerated by report, was probably one of the reasons. No estimate is possible of the income and expenses of the province but it is probable that the increasing cost of the military establishment was a drain on the central government.-Collingwood in Economic Survey, III, 14-16. Yet in the third century,. Britain escaped the general anarchy of the empire.- , Rostovtzeff, Roman Empire, p. 422Google Scholar. Ros-tovtzeff blames the financial difficulties of the mid-second century on the cost of Trajan's wars against Parthia and Dacia.- Ibid., pp. 307-15. , Frank in Economic Survey (V, 65, 67)Google Scholar regards Trajan's finances more favorably and cites a lightening of taxation and the booty from Dacia.- Ibid., p. 65, n. 13.

⁵¹ Heichelheim (I, 762-65) regards it as impossible to estimate the Roman budget. The various authors of the Economic Survey have attempted to do so (see esp. Frank in Vol. V, as cited above in n. 46, and, for Egypt, Johnson in Vol. II, 481-90).

⁵² New taxes for the aerarium militare were established in 6 A.D. to pay the veterans a cash bonus in place of allotments of land.- , Frank in Economic Survey, V, 7Google Scholar. Veteran colonies continued to be founded, particularly in the colonies.- , Parker, Legions, pp. 246–47Google Scholar; , Broughton in Economic Survey, IV, 702–3Google Scholar, for Asia; Frank in Ibid., V, 30-32, 62-65, for Italy. But the legionaries tended to settle themselves in the provincial towns near which they had seen service.

⁵³ , Parker, Legions (pp. 92–117)Google Scholar traces the history of the legions from the death of Augustus to 193 A.D. The basic statement is given by Tacitus (Ann. iv. 5) on the disposition of the Roman forces in 23 A.D. The fundamental discussion is , Ritterling's article, “Legio,” in RE, XII (23–24)Google Scholar, 1186-1837, of which the first half, to col. 1376, is a general discussion, and the second half treats the individual legions. Septimius' three new legions, I-III Par-thicae, are discussed in Half Vol. 23, cols. 1308-9. The normal imperial legion should have had 5,600 legionaries, but is generally figured at 5,000, which would give 125,000 legionaries for twenty-five legions. It is usually estimated that there were about an equal number of auxiliaries, making a 250,000 minimum for the whole army. Thirty-three legions with auxiliaries would give a minimum of 330,000 men.- Sandys, J. E., ed., A Companion to Latin Studies (3d ed.; Cambridge: The University Press, 1925), pp. 464–68Google Scholar; Nilsson, M. P., Imperial Rome (London: Bell and Sons, 1926), p. 285Google Scholar.

⁵⁴ For the pay and extras of legionary soldiers and officers, see , Parker, Legions, pp. 214–24 (up to 193 A.D.)Google Scholar. For increases under the Severi, see , Lammert's article, “Stipendium” in RE2, III (6), 2537Google Scholar. A good table is in Durry, M., Les Cohortes pritoriennes (Paris: Boccard, 1938), p. 267Google Scholar. The fundamental work is still Domaszewski, A. von, “Der Truppensold der Kaiserzeit,” in Neue Heidelberger Jahrbiicher, X (1900), 218–41Google Scholar., Frank in Economic Survey (V, 301)Google Scholar argues that the confiscations and innovations of Septimius, b y reducing a large number of tenants to the position of imperial coloni, made them unavailable for military service because their labor on the imperial estates was “essential,” and therefore encouraged the hiring of barbarian mercenaries, which both increased the cost of defense and tended to draw money out of the empire (see below, n. 64).

⁵⁵ For difficulties under Marcus, see , Rostovtzeff, Roman Empire, pp. 325–27Google Scholar; see also below, n. 106. , Frank in Economic Survey (V, 76–77)Google Scholar points out that the imperial treasury (fiscus) contained two billion seven hundred million sesterces at the accession of Marcus in 161 A.D. and was empty at his death in 180 A.D. Marcus was generous in gifts to the troops (donatiua) and to the people (congiaria).

⁵⁶ The best discussion of the Roman civil service is still Hirschfeld, O., Die kaiserlichen Verwaltungsbeamten bis auf Diocletian (2d ed.; Berlin: Weidmann, 1905)Google Scholar; more summary is Mattingly, H., The Imperial Civil Service of Rome (Cambridge: The University Press, 1910)Google Scholar. For salaries, see , Rosenberg's article “Salarium” in RE2, I (2), 1847Google Scholar. Hirschfeld (I, 477-80) thinks that the thoroughgoing reorganization of the administrative machine under Hadrian, with its sharp differentiation between the imperial household and the imperial officials, meant a considerable increase of personnel. Septimius introduced further fundamental reforms, chiefly to the disadvantage of the senatorial class (see pp. 480-82).

⁵⁷ For the court, , Hirschfeld, Verwaltungsbeamten, pp. 307–17Google Scholar; , Friedlander, Sittenge-schichte, II, 32–102Google Scholar.

⁵⁸ For extravagances of “bad” emperors, see Frank, in Economic Survey, V, 39–40 (Gaius Caligula), pp. 43, 45Google Scholar (Nero), pp. 55-56 (Titus and Domitian). He does not give details for Commodus, Caracalla, or Elagabalus. Thompson agrees that their faults had little adverse effect on the empire as a whole.- Middle Ages, pp. 6-7.

⁵⁹ Cancellation of delinquent taxes by the emperors was much praised during the second century but it proves either that people were increasingly in default or that the government collection agencies were inadequate. See , Frank in Economic Survey (V, 65 and 70, n. 23) for the burning of records under Trajan as portrayed on one of two balustrades now in the Forum (photograph in Rostovtzeff, Roman Empire, p. 314)Google Scholar, Ibid., p. 70 for Hadrian (from , Dessau, Ins. Lai. Sel. no. 309), p. 77Google Scholar for Marcus (from Dio Cassius Ixxi [lxxii]. 32.2, of 178 A.D.). Egypt had difficulty meeting the burden of taxes in the third century.- , Johnson in Economic Survey, II, 354Google Scholar.

⁶⁰ The content of the coins is conveniently presented by , Frank in Economic Survey, V, 90–93Google Scholar; see also Heichelheim, I, 682-89.

⁶¹ This interpretation of Nero's changes is Mattingly's in his Roman Coins (pp. 121-32). Frank, in Economic Survey (V, 35Google Scholar) justified Nero even for silver. Nero reduced the aureus from 1/42 to 1/46 of a Roman pound and changed the silver from practical purity to 10 per cent alloy.

⁶² Silver was alloyed 15 per cent under Trajan, 25 per cent under Marcus, and 40 per cent under Septimius.- , Mattingly, Roman Coins, p. 125Google Scholar; , Frank in Economic Survey, V, 91–92. Caracalla “reformed” the coinage by displacing the silver denarius (legally 25 to the aureus) by a new coin, the antoninianus, supposedly worth two denarii but slightly less than twice the size of the denarius and so alloyed as to have only two thirds the silver of the Neronian denarius. Even this coin was soon debased and the gold aureus continued also to lose weight.-Google Scholar, Mattingly, Roman Coins, pp. 125–26Google Scholar; , Frank in Economic Survey, V, 93Google Scholar.

⁶³ For depreciation in Syria, see , Heichelheim in Economic Survey, IV, 219–23Google Scholar; in Asia, OBroughto n in Ibid., IV, 906-7. For depreciation and rise of prices, see , Rostovtzeff, Roman Empire, pp. 419–21Google Scholar; , Johnson in Economic Survey, II, 147–48 (Egypt)Google Scholar. For reversion from a money to a natural economy in the third century, see , Rostovtzeff, Roman Empire, p. 612, n. 57Google Scholar; Heichelheim, I, 784. Naturally, the “bad” money drove the “good” money out of circulation into hoarding or the melting pot. Ancient politicians had neither the ingenuity to devise nor the control of public opinion to enforce a “managed” currency. And excessive depreciation can destroy faith even in token money, as in Germany in the inflation of 1920.

⁶⁴ Heichelheim (I, 683-84) apparently regards the supply of metal as inadequate and connects the shortage not only with the exhaustion of reserves but with the flow out of the empire to pay for imports (see p. 717) and later with the tribute payments to barbarian tribes (see p. 689 and Frank, as cited above in n. 54). , Frank in Economic Survey (I, 349, and V, 20-21)Google Scholar regards the supply of metal as adequate under Caesar and Augustus because of influx of capital from the east (see above, n. 47). But coinage was severely restricted under Tiberius and the more liberal coinage of his successors may have been economically unsound.- Ibid., V, 32-36.

⁶⁵ Under hoarding might be included the wear and tear on coinage through use ( , Warmington, Commerce, p. 316Google Scholar, under item III b, and n. 66 below); and use for ornamentation ( , Thompson, Middle Ages, pp. 40–41)Google Scholar.

⁶⁶ Estimates of the flow of coinage out of the empire in connection with the eastern trade (see above, n. 38) are based on two statements by Pliny the Elder (Nat. Hist. vi. 23F.261.101) in connection with the sea route from Egypt to India: “India in no year absorbs less than 50,000,000 sesterces ($2,213,200 gold) from our empire and sends back goods which are sold with us at a hundredfold profit”; and (xii.18C411.84) in connection with the spice and pearl trade from Arabia Felix: “By the lowest reckoning, each year India and China and the [Arabian] peninsula take from our empire 100,000,000 sesterces ($4,426,400 gold). So much do our luxuries and our women cost us.” The above equivalents are figured on a Neronian gold aureus weighing 7.39 grams and worth 25 denarii or 100 sesterces ( , Frank in Economic Survey, V, 91Google Scholar) and a gold dollar weighing 1.6718 grams, which gives $4.4264 per aureus. The amounts would be somewhat less on a Flavian aureus weighing 7.3 grams. , Frank in Economic Survey (V, 32)Google Scholar thinks that the drain was considerable under the Julio-Claudians but that Vespasian (see p. 283) restricted the export of bullion. He feels, however (pp. 298-99), that a considerable flow continued. , Warmington (Commerce, pp. 39–40)Google Scholar says that finds of coins indicate a high rate of flow to India from Augustus to Vespasian but less thereafter. Yet he says later (pp. 120-24) that in Ceylon coin finds are of late imperial pieces. Warmington surveys the question generally in his Commerce, pp. 272-318, “The Adverse Balance,” and concludes that, though the outflow was always considerable, its ultimate effect was at most to hasten a financial collapse which would have come in any case. His sequence of economic causes for the collapse (p. 316) is: (1) lack of sufficient regular substitute for “spoils of war” as source of metal; (2) unproductive use of money and selfishness of the rich; (3) failure to maintain a good supply of silver coinage because: (a) mines were wastefully worked and slave labor fell off, (b) coinage in circulation wore out and was lost, and (c) coins were drained out of the empire.

⁶⁷ Syrian glass products were exported to China in considerable quantities and sold at great profit (see above, n. 38).

⁶⁸ Salvioli, G., Le Capitalisme dans de monde antique (Paris: Girard and Briere, 1906)Google Scholar, gives in his two last chapters (ix, “Le gran d Sbranlement economique,” and x, “L'Economie antique”) the most forceful presentation of the effects of the shortage of coinage and the waste of capital in luxury and other unproductive manners. He emphasizes the fundamental difference between classical capitalism, based on commerce and usury, and modern capitalism, based on industrial production, with its divorce between capital and labor and its monopolization of the means of production. This view is rejected by , Rostovtzeff, Roman Empire, pp. 302-3, 585Google Scholar, n. 91, 482-85, and by Frank, in Economic Survey, V, 299Google Scholar.

⁶⁹ The best discussion of slavery in the ancient world is , Westermann's article, “Sklaverei,” in RE, Suppl. 6, pp. 894–1068Google Scholar, in which slavery during the early empire is treated in cols. 994-1063. See also Barrow, R. H., Slavery in the Roman Empire (London: Methuen, 1928), andGoogle ScholarDuff, A. M., Freedmen in the Early Roman Empire (Oxford: Clarendon Press, 1928)Google Scholar. Buckland, W. W., The Roman Law of Slavery (Cambridge: The University Press, 1908)Google Scholar, treats the legal, not the social and economic, aspects of slavery. Barrow (p. 97) regards slave and free labor as not incompatible on farms and thinks that slavery failed because i t was expensive and inefficient. He does not think (p. 129) that slavery prevented technological advance in trade and industry, since slave and free labor competed on equal terms. He draws an interesting comparison (pp. 230-36) between slavery in antiquity and in the United States and concludes that slavery is profitable only in a large-scale operation with one crop (or grazing) which requires organization but not individual skill or initiative. See above, n. 9, for Heitland's similar view in his Agricola; also , Frank in Economic Survey, V, 175-82, 235-36, 297Google Scholar.

⁷⁰ “The ever increasing decline of slavery under the empire is to be regarded as the consequence, not as the cause, of the economic and political changes during this period.”- , Westermann, “Sklaverei,” RE, Suppl. 6, p. 1062, col. 1014Google Scholar. He gives as reasons for the decline: (1) the high cost of slave labor; (2) the risk of loss of property in slave by death; (3) the reduction of the free peasant, who might own a slave or two, to the colonate status, where he was too poor to do so; (4) the shift of industry to the newer western provinces where industrial slavery was not common; and (5) the ease of transition from slavery to freedom (see also , Rostovtzeff, Roman Empire, p. 313Google Scholar; Heichelheim, I, 756-57).

⁷¹ Duff thinks that agricultural slaves did not reproduce where the urban ones did, but urban slaves in particular were benefited by manumission and humanitarian legislation so that they too did not maintain their numbers.- Freedmen, pp. 3-5.

⁷² Beloch, J., Die Bevolkerung der Griechisch-romischen Welt (Leipzig: Duncker and Humbolt, 1886)Google Scholar, is still fundamental for population in the classical world. See also the various volumes of the Economic Survey. Beloch (pp. 496-505) thinks that an absolute decline of population began in Greece during the second century B.C. and in Italy after the time of Claudius. He connects this with the spread of slavery.

⁷³ For depopulation of Greece, see , Beloch, Bevolkerung, pp. 496–500Google Scholar; , Rostovtzeff, Hellenistic World, II, 623–25Google Scholar.

⁷⁴ For depopulation of Italy, see , Beloch, Bevolkerung, pp. 418–19Google Scholar; , Rostovtzeff, Roman Empire, p. 311 (under Trajan)Google Scholar. State funds for orphans (alimenta) are attested only for Italy and were aimed at helping agriculture as well, since the capital was put out on loan to farmers. In the end, this probably had a disastrous effect, since the loan was on a permanent basis and the interest actually constituted a permanent added tax on the farms in question. By the time of Septimius, the state funds were in financial difficulty and they vanished during the third century. Private alimenta are attested in the provinces as well as n i Italy. See the article, “Alimenta,” in Diz. Epigrafico, I, 402–11Google Scholar, which is fuller than that in RE, 1(2), 1484-89; also , Frank in Economic Survey, V, 65-67, 70. 75, 86, n. 50, 88, 101, 106-7, 173–74Google Scholar; , Rostovtzeff, Roman Empire, pp. 311-13, 314 (balustrade of Trajan from Forum), 356, 544, n. 4, 587, n. 6Google Scholar.

⁷⁵ For settlement of barbarians, see , Nilsson, Imperial Rome, pp. 351–52, and above, n. 19Google Scholar.

⁷⁶ Stein, E., Geschichte der Spdtromischen Retches (Vienna: Seidel and Sohn, 1928—), pp. 1, 3Google Scholar, states that the general population of the empire fell from about seventy million at the birth of Christ to about sixty million at the end of the third century. Oertel, in C. A. H. (XII, 267-68Google Scholar) accepts this view. , Johnson in Economic Survey (II, 246)Google Scholar accepts a decline in Egypt in the third century; Collingwood in Ibid., (III, 77) adduces evidence for infanticide in Britain; Scramuzza in Ibid., (III, 368) thinks that the population of Sicily was static rather than declining (see below, n. 78).

⁷⁷ For plague under Marcus, see , Rostovtzeff, Roman Empire, p. 325Google Scholar; , Frank in Economic Survey, V, 76Google Scholar, with references.

⁷⁸ Rostovtzeff disputes the view of Seeck and others (see above, n. 76) that depopulation was general throughout the empire and a cause of the collapse of ancient civilization.- Roman Empire, pp. 328-29, 591, nn. 30-31, for references. It is equally difficult to blame the collapse on the mixture of races, as is done by Nilsson, Imperial Rome, in his chapter on “The Population Problem,” pp. 317-67 (see below, n. 104).

⁷⁹ Syme, R., The Roman Revolution (Oxford: Clarendon Press, 1939), chaps, xxxi-xxxxiii, pp. 476–524Google Scholar, gives a vigorous picture of the opposition and doom of the republican noble families and their replacement by “new men” under Augustus and his immediate successors. Syme's picture of Augustus as the coming politician is perhaps overdrawn.- Hammond, M., The Augustan Principate (Cambridge: Harvard University Press, 1933), pp. 117–20Google Scholar; Jones, H. S. in C. A. H., X, 176–81Google Scholar. Augustus' policy was to maintain the supremacy of the Romano-Italic elements against the provincial ( , Rostovtzeff, Roman Empire, pp. 47–49)Google Scholar and it is questionable how far this policy was seriously abandoned until the Flavians, despite Claudius' admission of certain Gallic chiefs to the Senate.- Scramuzza, V., The Emperor Claudius (Cambridge: Harvard University Press, 1940), pp. 90–110Google Scholar. For Vespasian's admission of provincials, see Last in C. A. H., XI, 418-20. During the Flavian and Antonine period, not only were provincials increasingly admitted to the senatorial class and high administrative positions (see below, n. 81) but also citizenship was widely extended and its privileges came to count for little while its obligations became more burdensome. Finally, Caracalla, by an edict in 212 A.D. (the Constitutio Antoniniana), extended citizenship to practically all who were then inhabitants of the empire. At least one ancient author, the contemporary Dio Cassius (lxxvii [lxxviii]. 9.5) gives as a reason Caracalla's desire to collect from everybody the taxes previously paid by Roman citizens only, presumably since the exemptions which citizens had originally enjoyed from other taxes (such as that on land) had been forgotten. Dio's explanation has not, however, been unanimously accepted. See Sherwin-White, A. N., The Roman Citizenship (Oxford: Clarendon Press, 1939), Part II, pp. 167–230Google Scholar, passim. For the edict, Ibid., pp. 220-27; , Miller in C. A. H., XII, 45–47Google Scholar and bibliography on p. 734. The fullest study of the edict is Capocci, V., “La Constitutio An-toniniana,” in Studi di papirologia e di diritlo pubblico Romano, Memorie delta R. Acca-demia Nazionale dei Lincei, classe di scienze morali, storiche e filologiche, anno CCCXXII, Serie VI, Vol. I, Fasc. I (Rome: Bardi, 1925)Google Scholar.

⁸⁰ For the social policy of Augustus, see Last in C. A. H., X, 425-64; for the “eugenic” purpose of the Leges Iulia et Papia Poppaea, see Field, J. A., “The Purpose of the Lex Iulia et Papia Poppaea,” The Classical Journal, XL (1945), 398–416Google Scholar.

⁸¹ For studies of the composition of the senate, see Willems, J., “Le Senat romain en l'an 65,” Le Musée Belge, IV (1900), 236–77Google Scholar; V (1901), 82-126; VI (1902), 100-51; Stech, B., “Senatores Romani qui fuerint inde a Vespasiano usque ad Traiani exitum,” Klio, Beiheft X (1910), 70–117Google Scholar; Lambrechts, P., La Composition du Sénat romain de I'accession au thrône d'Hadrien à la mort de Commode (Antwerp: De Sikkel, 1936), pp. 117–92Google Scholar, and idem, La Composition du Sénat romain de Septime Sévére à Dioclétien, Dissertationes Pannonicae (Budapest: Inst. de Num. et d'Arch. de l'Univ. P. Pazmany, 1937), Série I, Fasc. 8, pp. 193-284. See , Duff in C. A. H. (XI, 746–48)Google Scholar for a brief statement of the position of the senate in the second century.

⁸² For the senate of the third century, see , Ensslin in C. A. H., XII, 375–76Google Scholar. For a picture of the great Roman and provincial senators of the later empire, see Dill, Samuel, Roman Society in the Last Century of the Western Empire (2d ed.; London: Macmillan and Company, 1899)Google Scholar, Bk. II, pp. 115-223.

⁸³ For the development of “litourgies” into a system of state compulsion, see , Rostovtzeff, Roman Empire, pp. 333–43. “Litourgy” is the Greek word for uncompensated public service (other than military) expected from the individualGoogle Scholar.

⁸⁴ It is noted by one of the editorial readers for this JOURNAL that government control as expressed in “mercantilism” between about 1550 and 1800 did not prevent the great expansion of the European economic system.

⁸⁵ , Frank in Economic Survey (V, 21)Google Scholar holds that Augustus promoted commerce only to increase port dues and (pp. 294-95) denies in general any conscious commercial policy on the part of the Roman government. Charlesworth thinks on the contrary that there was conscious encouragement of commerce.- Trade-routes, pp. 228-34. Rostovtzeff praises highly Hadrian's appreciation of the economic problems of his day and the measures that he took to combat them, particularly with reference to agriculture, taxation, and municipal expenditures.- Roman Empire, pp. 315-25. Heichelheim (I, 678-82) regards Augustus as too conservative and reactionary in his economic program, as against Caesar (see above, n. 13). Conscious economic policy was familiar to the Greek world, particularly in the monopolistic Ptolemaic state and in the commercial states of Carthage and Rhodes.- Bullock, C. J., Politics, Finance, and Consequences (Harvard Economic Studies, LXV, 1939)Google Scholar. But , Frank argued in his Roman Imperialism (New York: The Macmillan Company, 1914)Google Scholar that the Roman aristocracy was not guided by economic motives until the emergence of the equestrian class in the first century B.C., whose policies are chiefly represented by Pompey. The Augustan reaction favored the senatorial landowning class, with its Junker prejudices, and tended to shepherd the republican financiers, the equestrians, into government service and to reduce their opportunities for gain in handling the collection of taxes (see below,

⁸⁶ See above, n. 16.

⁸⁷ Subsidies to shipowners under Claudius.-Suetonius Cl. 18.2; Scramuzza, , Claudius, pp. 167-69Google Scholar. In general, see , Rostovtzeff, Roman Empire, pp. 149 and 532, n. 22 (with further references), 336, 359, 361, 379-80, 3971 408-9, 474. 592, n. 37; Heichelheim, I, 711-12Google Scholar; , Thompson, Middle Ages, pp. 26–28Google Scholar.Waltzing, J. P., Etude historique sur les corporations projessionelles chez les Romains (4 vols.; Louvain: Peeters, 1895-1900) is still the fundamental study of the collegiaGoogle Scholar.

⁸⁸ For the contract system of tax collection under the Republic, see Last in C. A. H., IX, 65-66 (Gracchan period) and Stevenson in Ibid., IX, 469-71 (Ciceronian period); for the Augustan reforms, see Stevenson in Ibid., X, 190-93. Cf. , Rostovtzeff, Roman Empire, pp. 159–60Google Scholar; Stevenson, G. H., Roman Provincial Administration till the Age of the An-tonines (Oxford: B. Blackwell, 1939), chap, v, pp. 133–55Google Scholar.

⁸⁹ For public contracting of the collection of taxes and the management of public properties, such as estates and mines, see M. Rostovtzeff (Rostowzew) “Geschichte der Staats-pacht in der rémischen Kaiserzeit bis Diocletian,” Philologus, Supplement band IX (1901-1904)Google Scholar, Heft 3, pp. 329-512. In pp. 367-74, he discusses briefly the publicani under the later republic.

⁹⁰ For the collection of taxes by municipalities and the responsibility of the members (curiales) of the municipal councils (curiae, see , Rostovtzeff, Roman Empire, pp. 159-60, 317-18, 339-43, 593–95Google Scholar, nn. 40-46; Abbott, F. F. and Johnson, A. C., Municipal Administration in the Roman Empire (Princeton: Princeton University Press, 1926), chap, ix, pp. 17–37Google Scholar.

⁹¹ For the decay of municipalities, see , Abbott and , Johnson, Municipal Administration, chap, xiv, pp. 117–37Google Scholar.

⁹² For plight of curiales, see , Rostovtzeff, Roman Empire, pp. 469–77Google Scholar; , Thompson, Middle Ages, pp. 45–46Google Scholar; , Dill, Roman Society, pp. 245–81Google Scholar.

⁹³ For Scaeva, see , Dessau, Ins. Lat. Sel. no. 915Google Scholar.

⁹⁴ For Vespasian and the eastern frontier, see C. A. H., XI, 140-41.

⁹⁵ For Trajan and the financial difficulties of cities, see Rostovtzeff, Roman Empire, pp. 313-15; for Pliny, see C. A. H., XI, 218-20; for Pontus and Bithynia, see Ibid., pp. 575-80.

⁹⁶ For Maximus in Greece (province of Achaea), see Pliny the Younger Ep. viii.24.2.

⁹⁷ Articles on “Curatores,” Sec. 10 in RE, IV (8), 1806-11 and “Curator republicae” in Diz. Epigrafico, II, 2, 1345-86.

⁹⁸ For the “state socialism” of the later empire, see Oertel in C. A. H., XII, 270-81. The most interesting instance of the attempt of the government to regulate -.conomic activity was Diocletian's edict in 301 which fixed maximum prices (anticipating the O.P.A.?). The text is reconstructed by E. R. Graser in an appendix to Economic Survey, V, 305-421. Frank (in Ibid., pp. 299-300) thinks that the effort was justified, but this is not the general view and certainly the attempt was short-lived; see Heichelheim, I, 788-91, and II, 1202, n. 9, who finds parallels in the near eastern monarchies; , Ensslin in C. A. H., XII, 405 and bibliography on p. 762Google Scholar; and , Bliimner's article, “Edictum Diocletiani,” in RE, V (10), 1941-1957Google Scholar.

⁹⁹ That progress in the Roman economic system was.in extension, not depth, is the view of , Oertel in C. A. H., XII, 237–42Google Scholar. In pp. 252-56, he attempts to explain the failure in respect to technological progress: (1) by the unstable and limited system of capital; (2) by the low purchasing power of the masses (Rostovtzeff's reason for the lack of industrial development, Roman Empire, pp. 303-5); (3) by slave labor which made economy of labor unnecessary; (4) by lack of creative energy (see , Frank in Economic Survey, V, 298)Google Scholar, which meant that increased demand was met by decentralization and not by mass production; (5) by the emergence of the economically self-contained large estates at the expense of the towns; and (6) by the development of state socialism. See above, n. 68, for the view of Salvioli (and others) that the failure was due to inability to develop beyond a house economy and for the rejection of this view by Rostovtzeff and Frank.

¹⁰⁰ , Oertel in C. A. H., XII, 250–54Google Scholar.

¹⁰¹ Augustus definitely turned his back on geographical expansion, but his motives are unknown. He may have done so because of innate caution (Heichelheim, I, 679) or because he felt that the empire could not afford the men or money necessary (see above, n. 49) or because he realized that the empire had reached the natural limits of the Mediterranean basin; or because he realized that the maintenance of peace demanded a stable frontier.- , Rostovtzeff, Roman Empire, pp. 52–53Google Scholar.

¹⁰² , Oertel in C. A. H., XII, 237–42Google Scholar.

¹⁰³ On the crisis of the third century and the economic restoration, see , Oertel in C. A. H., XII, 259–70Google Scholar. In pp. 279-80, he argues that the fourth century witnessed a considerable degree of recovery but that this was not adequate and was purchased at the price of state socialism. See also , Rostovtzeff, Roman Empire, chaps, x-xii, pp. 381–478Google Scholar;, Frank in Economic Survey, V, 302–3Google Scholar; , Thompson, Middle Ages, p. 42Google Scholar.

¹⁰⁴ A popular but interesting summary of various explanations which have been given for the decline and fall of the Roman empire is White, E. L., Why Rome Fell (New York and London: Harper and Brothers, 1927), pp. 260–323Google Scholar. White agrees with Gibbon that the introduction of Christianity, with its otherworldliness, undermined the will of the Romans to resist the barbarians. Nüsson (Imperial Rome, p. 299) blames the fall on the barbarization of the army because of the growing disinclination of the more civilized populations for military service and (pp. 360-67; see also above, n. 78) on the loss of the superiority of the Greco-Roman stock in consequence of “mongrelization.” For wealth as a factor in the fall, see above, n. 17. For Frank's general conclusion, see his concluding essay in Economic Survey, V, 296-304; for Rostovtzeff's, his two concluding paragraphs, Roman Empire, pp. 486-87; for Heichelheim's, his concluding paragraph, I, 859.

¹⁰⁵ So , Thompson (Middle Ages, pp. 55)Google Scholar citing the survival of the Byzantine state, the even more corrupt, stagnant, and oppressive successor to the later empire. He concludes: “Why did the Roman Empire pass away ? To this question Clio vouchsafes no clear answer.” It is symptomatic of the fascination which the fall of the Roman empire exercised over the medieval and modern mind that thinkers from Augustine and Orosius to the present have nevertheless continued to plague Clio for an answer.

¹⁰⁶ This pressure became acute under Marcus (see above, n. 55) just at the time when the internal difficulties had become apparent under Trajan and Hadrian. , Rostovtzeff, Roman Empire, pp. 306–25Google Scholar. It seems like a turn of the tide from the high-water mark of Trajan's advances into Dacia and Parthia.

¹⁰⁷ , Rostovtzeff, Roman Empire, pp. 452–53Google Scholar; , Frank in Economic Survey, V, 303–4Google Scholar.