¹ Calza, G., ‘La preeminenza dell’ “insula” nella edilizia romana’, Mon. Ant. 23 (1916), 575–6Google Scholar.

² Nash, E., A Pictorial Dictionary of Ancient Rome (London, 1961–1962) 11, 91, 232–3Google Scholar, cited hereafter as Nash. Throughout this paper, in numbering floors, I count the ground floor as the first.

³ Muñoz, A., Campidoglio (Rome, 1930), 45–52Google Scholar, plates 16 and 17; Nash 1, 506–7.

⁴ Ibid. 357–361. A. Prandi, Il complesso monumentale della basilica Celimontana dei SS. Giovanni e Paolo (Rome, 1935), 76.

⁵ The area near the Piazza Colonna under the large, modern commercial centre known as the Galleria Colonna was first published by Gatti, E., ‘Scoperte di antichità a Piazza Colonna’, NSc (1917), 9–20Google Scholar. The most recent work on the complex has been done by Gatti, G. in Saggi … in onore del … V. Fasolo (Rome, 1961), 49–66Google Scholar.

⁶ Calza, G., Architettura e Arti Decorative 3 (1923), 17Google Scholar; Nash 11, 123.

⁷ Lanciani, R., The Destruction of Ancient Rome (New York, 1899), 202Google Scholar.

⁸ Carettoni, G. et al. , La pianta marmorea di Roma (Rome, 1961) 1, 200Google Scholar, cited hereafter as Carettoni.

⁹ Ibid. 11, Plate 39, No. 165.

¹⁰ See below notes 14 and 15.

¹¹ Carettoni 11, Plate 38, No. 141; Plate 56, No. 599.

¹² Ibid. 1, 202.

¹³ De Beneficiis 4, 6, 2; 6, 15, 7; De Ira 3, 35, 4–5.

¹⁴ Tacitus, , Ann. 15, 43Google Scholar; Suetonius, , Nero 16, 1Google Scholar; Boëthius, Axel, The Golden House of Nero (Ann Arbor, 1960), 155–6Google Scholar, cited hereafter as Boëthius; Meiggs, R., Roman Ostia (Oxford, 1960), 237Google Scholar, cited hereafter as Meiggs. Nash 1, 506, assumes as a matter of course that insula signifies a concrete, brick-faced apartment-building.

¹⁵ Martial 1, 108, 3; 117, 6–7; 3, 30, 3; 4, 37; 5, 22; 6, 27, 1–2; 7, 20, 20 (if taken literally the ducentas scalas of this passage would imply that Martial knew of buildings six and seven stories high); 8, 14. Juvenal 3, 6 ff., 166, 190–202, 223–5, 235, 268–277; 11, 12–13.

¹⁶ [Aurelius Victor], Epitome de Caesaribus 13, 13.

¹⁷ Aulus Gellius 15, 1, 2–3.

¹⁸ Scriptores Historiae Augustae, Antoninus Pius 9.

¹⁹ Digest 9, 2, 27, 8.

²⁰ Herodian 7, 12, 5–6.

²¹ Symmachus, , Epistulae 6, 37Google Scholar.

²² For an account of these buildings in the Late Republic, see Cicero, , Ad Atticum 14, 9Google Scholar. Cicero here seems to regard the collapse of such buildings as an everyday occurrence.

²³ Carcopino, J., Daily Life in Ancient Rome (New Haven, 1940), 29–39Google Scholar, cited hereafter as Carcopino. Carcopino here gives only a generalized description of such buildings.

²⁴ Crema, L., L'architettura romana (Turin, 1959), 296Google Scholar, gives the figures for the Colosseum. The collapse of only a part of the canopies of the Circus Maximus involved 13,000 people in its ruin according to G. V. Gentili, BdA (1957), 25, 27 (n. 80). Lugli, G., Roma antica, il centro monumentale (Rome, 1946), 602Google Scholar, suggests a figure of 320,000 for the total capacity of the Circus, but points out that many of these spectators may have seen the races from elevated points outside the building itself.

²⁵ For a description of these baths, see Nash II, 429–477; for their operation, Carcopino 254–263.

²⁶ The various estimates made of the ancient population of Rome are summarized by Maier, F. G., Historia 2 (1953–1954), 321–2Google Scholar, cited hereafter as Maier.

²⁷ Cassius Dio 55, 10; RG 15 (discussed by Hardy, E. G., Mon. Ancyranum (Oxford, 1923), 76–81Google Scholar).

²⁸ Rostovtzeff, M., The Social and Economic History of the Roman Empire, 2nd ed. (Oxford, 1957), 81Google Scholar.

²⁹ See Appendix below.

³⁰ Maier 327. One of the most interesting attempts to estimate the population of Rome on the basis of the figures for the grain-dole has been by W. J. Oates. For a discussion of his views, see Appendix below.

³¹ The texts of both these important documents are contained in Nordh, A., Libellus de Regionibus Urbis Romae (Lund, 1949)Google Scholar, hereafter cited as Nordh.

³² See above, pp. 80–1 and Boëthius 137.

³³ In Rend. Line. 26 (1917), 60–87Google Scholar, Calza estimates the population of ancient Rome at 1,800,000 persons (p. 78). In his ‘La popolazione di Roma antica’, Bull. Comm. 69 (1941), 142–155Google Scholar, hereafter cited as Calza, PR, he has modified his figures to 1,215,648 (P. 155).

³⁴ von Gerkan, A., Röm. Mitt. 55 (1940), 159Google Scholar.

³⁵ Nordh 86.

³⁶ See above p. 80.

³⁷ Packer, J., The Insulae of Imperial Ostia, unpublished dissertation (Berkeley and Rome, 1964), 28–9Google Scholar, hereafter cited as Packer. E. Cuq, Mémoires de l'Académie des Inscriptions et Belles Lettres (1916), 335, and Maier 327 reach similarly negative conclusions.

³⁸ Vaglieri's work appeared in NSc during the first two decades of the present century. Calza's most important articles (not previously cited above) include : ‘La casa romana’, Capitolium 5 (1929), 521–531Google Scholar; ‘Contributi alla storia della edilizia imperiale romana’, Palladio 5 (1941), 1–33Google Scholar, hereafter cited as Calza, Contributi; ‘Gli scavi recenti nell' abitato di Ostia’, Mon. Ant. 26 (1920), 322–443Google Scholar, hereafter cited as Calza, SR.

³⁹ Boëthius 156.

⁴⁰ Meiggs 13–14.

⁴¹ Calza, PR 142–155.

⁴² Girri, G., La taberna nel quadro urbanistico e sociale di Ostia (Rome, 1956), 37–43Google Scholar, hereafter cited as Girri.

⁴³ Meiggs 532–4

⁴⁴ Girri 42.

⁴⁵ Meiggs 533.

⁴⁶ See Chart I, pp. 90–3 below, for a list of shop and factory buildings.

⁴⁷ Girri 36.

⁴⁸ A plan of I, iv, 1, appears in Calza, G., Becatti, G. et al. ,, Scavi di Ostia (Rome, 1953)Google Scholar I (hereafter cited as SO I), Map, and the structure has been treated by Calza, SR 334 ff. and by Girri 11.

⁴⁹ R. Paribeni, NSc (1916), 419–423. Paribeni's restoration of the building appears on p. 422.

⁵⁰ On I, iv, I, see above, n. 48; on the Horrea Epagathiana, Calza, Contributi 19 ff.

⁵¹ The Caseggiato dei Triclini is treated ibid. 3–6; also Harsh, P., MAAR 12 (1936), 22–4Google Scholar, hereafter cited as Harsh; Kleberg, T., Hôtels restaurants et cabarets dans Vantiquite romaine (Uppsala, 1957), 45Google Scholar. The Caseggiato del Serapide is best described by Calza, Contributi 8 ff.

⁵² Factories are connected with their mezzanines in the Caseggiato delle Fornaci (II, vi, 7) and the Insula Trapezoidale (III, iv, 1). II, vi, 7, has been published in NSc (1912), 388–9; (1913), 125–7. The plan of III, iv, 1, appears in SO 1, Map, and the same volume dates the structure to A.D. 128–30 (p. 235).

⁵³ Meiggs 533; Becatti, G. and Calza, G., Ostia (Rome, 1961), 17Google Scholar.

⁵⁴ Meiggs 533, n. 2, criticizes Girri's assumption that all shops were inhabited, but there seems to be no reason to doubt this suggestion, particularly since the preserved travertine thresholds of almost all shops were equipped with a slot for a small ‘night-door’ through which the inhabitants could enter and leave after the shutters which closed the establishment for the night had been put up. For an illustration of shop-doors at Ostia, see Paschetto, L., Ostia, colonia romana (Rome, 1912)Google Scholar, fig. 80, cited hereafter as Paschetto. Similar shop-doors appear also at Pompeii and are discussed for that site by Mau, A., Pompeii, 2nd ed. (New York, 1902), 289–90Google Scholar.

⁵⁵ See Chart I, pp. 90–3 below, for a list of the structures. In addition to this list the Chart also shows the estimated number of floors and apartments in each house, the average number of apartments per floor, the total number of apartments in each building, and the total population for each structure. We have no real way of knowing how many apartments existed on the vanished upper floors of most buildings, and therefore, in order to obtain such figures for each structure, I first counted the number of living units on the ground floor, whose plan is usually intact. I then assumed that the number of rooms on each upper floor was exactly the same (but see n. 56 below) and divided the total number of these rooms by two to obtain the number of upper-floor units, two being the average number of rooms in surviving upper-floor flats. Thus, I managed to reach an estimate of the number of apartments per upper floor. I used this procedure except in cases where the arrangement of the upper floors was reasonably clear from the plan of the ground-floor rooms and stairways. Finally, in order to obtain the total number of apartments for the entire structure, I multiplied the number of flats on each floor by the total number of floors in the structure. Deciding upon a satisfactory estimate of the number of persons in each apartment was more difficult, for, as Duncan-Jones, R. P. has pointed out in JRS LIII (1963), 84Google Scholar, such figures tend to be ‘unverifiable’. As a basis for my calculations, consequently, I have accepted and rounded off the figure of 35 persons per ancient household given by Russel, J. P., Transactions of the American Philosophical Society 48 (1958), 65Google Scholar. I have, therefore, assumed—and my assumption is no more than a rough estimate—that each Ostian family was composed of four persons. In order to obtain my population figures for each building, I thus multiplied by four the number of apartments in every structure.

⁵⁶ This is of course a dangerous assumption. The upper floors of the apartment house embedded in the Aurelian Wall near the Porta Tiburtina (see above p. 80 and n. 2) certainly did not duplicate those below, and in some cases, it appears that the higher one went in a Roman building, the worse conditions became. On this subject, see Carcopino 44. The plans of Ostian multiple dwellings, so far as we can tell from the surviving remains, do, however, appear to have repeated themselves from floor to floor. There are several types of buildings at Ostia which have apartments on the first floor. The exact population statistics for these latter structures are given in Chart 11, pp. 93–5 below.

⁵⁷ Girri 32.

⁵⁸ Meiggs 533.

⁵⁹ These Ostian structures are treated in SO 1, 219 ff., 236. They appear also in NSc (1909), 164–5, 201, 411–412 (map facing p. 412). Paschetto 433–4 records the earliest work done on the buildings. See also Packer 251.

⁶⁰ The architectural superiority of Ostia as a city to Rome may be deduced from the fact that in Ostia almost all the buildings are regularly planned and constructed of brick-faced concrete—as does not appear to have been the case in Rome. On the dates of Ostian buildings, see SO 1, 233–8. As the chronological lists on these pages show, almost all Ostia was built between A.D. 90 and 180.

⁶¹ Although the highest population-concentrations were, generally speaking, within a comparatively short distance of the town's Forum, there were notable exceptions : 240 persons in 11, V, 1; 320 in IV, ii, 2, 3; and no less than 928 in the remarkable Garden House Complex (III, ix, 1–26).

⁶² On I, v, 1, 2, and I, vi, 1, see SO 1, 216, 235; Girri 11–12. For the buildings attached to the Baths of Neptune, see SO 1, 236, 238; Girri 20; Meiggs 409–411; Paschetto 265.

⁶³ See above, p. 85.

⁶⁴ The largest public latrines found in Ostia are those of the Forum Baths (SO I, Map, Section 13) and in the Caseggiato dei Triclini (1, xii, 1). On the latter see Harsh 22–4 and Calza, Contributi 3–6. Photographs of this last latrine appear in Meiggs, Plate 9, and in Packer 1280, Plate 10.

⁶⁵ Boëthius 156–7. Boëthius' point is supported not only by the archaeological evidence from Rome and Ostia but also by a passage from the jurist Paulus quoted in the Digest (1, 15, 3, 2) : ‘Effracturae fiunt plerumque in insulis in horreisque, ubi homines pretiosissimam partem fortunarum suarum reponunt…’ See also MacMullen, R., Enemies of the Roman Order (Cambridge, Mass., 1966), 166–67CrossRefGoogle Scholar.

⁶⁶ Oates, W. J., ‘The Population of Rome’, CP 29 (April, 1934), 101–116Google Scholar (quoted by Carcopino 18 and n. 38, p. 291), hereafter cited as Oates.

⁶⁷ Ibid. 104–5.

⁶⁸ Ibid. 107.

⁶⁹ Cicero, , II Verr. 3, 70, 163Google Scholar, quoted by Oates 112.

⁷⁰ Ibid. 114.

⁷¹ Ibid. 115.

⁷² Ibid. 102, n. 4.

⁷³ Ibid. 104, n. 9.

⁷⁴ Friedländer, L., Roman Life and Manners Under the Empire (London, 1913) IV, 22Google Scholar, hereafter cited as Friedländer.

⁷⁵ Oates 104, n. 10.

⁷⁶ Friedländer 25.

⁷⁷ Oates 105.

⁷⁸ CIL XIII, 5, 708.

⁷⁹ See above, pp. 87–8.

⁸⁰ Oates 104.

⁸¹ J. Beloch, Die Bevölkerung der griechisch-römischen Welt (1884), 392–413, argued that this area extended for a distance of some 40 kilometres around Rome and included not only Ostia but also Velitrae, Praeneste, Tibur, etc. (cited by Friedländer 27). The notion is attacked by Friedländer, loc. cit.

⁸² Severus 23, 2, ‘Moriens … reliquit … olei vero tantum ut per quinquennium non solum urbis usibus, sed totius Italiae, quae oleo eget, sufficeret’.