[Illustration: STREETS IN TIMGAD. From a photograph. ]

ANCIENT TOWN-PLANNING

By

F. HAVERFIELD

Oxfordat The Clarendon Press

1913

OXFORD UNIVERSITY PRESSLondon::Edinburgh::Glasgow::New YorkToronto::Melbourne::BombayHUMPHREY MILFORDPublisher to the University

PREFACE

The following pages are an enlargement of a paper read to theUniversity of London as the Creighton Lecture for 1910, and alsosubmitted in part to the London Conference on Town-planning in thesame year. 

The original lecture was written as a scholar's contribution to amodern movement. It looked on town-planning as one of those newmethods of social reform, which stand in somewhat sharp contrast withthe usual aims of political parties and parliaments. The latterconcern mainly the outward and public life of men as fellow-citizensin a state; they involve such problems as Home Rule, Disestablishment, Protection. The newer ideals centre round the daily life of humanbeings in their domestic environment. Men and women--or rather, womenand men--have begun to demand that the health and housing and food andcomfort of mankind, and much else that not long ago seemed to lieoutside the scope of legislation, should be treated with as closeattention and logic and intelligence as any of the older and moreconventional problems of politicians. They will not leave even thetubes of babies' feeding-bottles to an off-hand opportunism. 

Among these newer efforts town-planning is one of the better known. Most of us now admit that if some scores of dwellings have to be runup for working-men or city-clerks--or even for University teachersin North Oxford--they can and should be planned with regard to thehealth and convenience and occupations of their probable tenants. Town-planning has taken rank as an art; it is sometimes styled ascience and University professorships are named after it; in theLondon Conference of 1910 it got its _deductio in forum_ or atleast its first dance. But it is still young and its possibilitiesundefined. Its name is apt to be applied to all sorts ofbuilding-schemes, and little attempt is made to assign it any specificsense. It is only slowly making its way towards the recognized methodand the recognized principles which even an art requires. Here, itseemed, a student of ancient history might proffer parallels fromantiquity, and especially from the Hellenistic and Roman ages, whichsomewhat resemble the present day in their care for the well-being ofthe individual. 

In enlarging the lecture I have tried not only to preserve this pointof view, but also to treat the subject in a manner useful to classicalscholars and historians. The details of Greek and Roman town-planningare probably little known to many who study Greek and Roman life, andthough they have often been incidentally discussed, [1] they have neverbeen collected. The material, however, is plentiful, and itilluminates vividly the character and meaning of that city-life which, in its different forms, was a vital element in both the Greek and theRoman world. Even our little towns of Silchester and Caerwent in RomanBritain become more intelligible by its aid. The Roman student gainsperhaps more than the Hellenist from this inquiry, since the ancientRoman builder planned more regularly and the modern Romanarchaeologist has dug more widely. But admirable German excavations atPriene, Miletus, and elsewhere declare that much may be learnt aboutGreek towns and in Greek lands. 

 [1] For example, by Beloch in his volume on the cities of Campania, by Schulten in various essays, by Barthel in a recent inquiry into Roman Africa, and by others, to be cited below. Dr. J. Stübben in his _Städtebau_ (Darmstadt, ed. 2, 1907) and Mr. Raymond Unwin in his _Town planning in practice_ (London, 1909) have given interesting notices and illustrations of the subject for modern builders. 

The task of collecting and examining these details is not easy. Itneeds much local knowledge and many local books, all of which are hardto come by. Here, as in most branches of Roman history, we want aseries of special inquiries into the fortunes of individual Romantowns in Italy and the provinces, carried out by men who combine twothings which seldom go together, scientific and parochial knowledge. But a body of evidence already waits to be used, and though itsdiscussion may lead--as it has led me--into topographical minutiae, where completeness and certainty are too often unattainable and errorsare fatally easy, my results may nevertheless contain some newsuggestions and may help some future workers. 

I have avoided technical terms as far as I could, and that not merelyin the interests of the general reader. Such terms are too often bothugly and unnecessary. When a foreign scholar writes of a Roman town as'scamnirt' or 'strigirt', it is hard to avoid the feeling that thisis neither pleasant nor needful. Perhaps it is not even accurate, as Ishall point out below. I have accordingly tried to make my text asplain as possible and to confine technicalities to the footnotes. 

F. H. 

CONTENTS

 LIST OF PLANS AND ILLUSTRATIONS

 TABLE OF MEASURES

 1. PRELIMINARY REMARKS ON ANCIENT TOWN-PLANNING

 2. GREEK TOWN-PLANNING. THE ORIGINS: BABYLON

 3. GREEK TOWN-PLANNING. FIRST EFFORTS

 4. GREEK TOWN-PLANNING. THE MACEDONIAN AGE

 5. ITALY. THE ORIGINS

 6. ITALY. THE LATE REPUBLIC AND EMPIRE

 7. ITALIAN TOWNS

 8. ROMAN PROVINCIAL TOWNS. I

 9. ROMAN PROVINCIAL TOWNS. II

10. ROMAN BUILDING LAWS

11. THE SEQUEL

 APPENDIX. TOWN-PLANNING IN CHINA

 INDEX

LIST OF PLANS AND ILLUSTRATIONS

(For precise references to sources see the various footnotes. )

 STREETS IN TIMGAD. From a photograph 1. BABYLON. After Koldewey and others 2. PIRAEUS. After Milchõfer 3. SELINUS. After Cavallari and Hulot and Fougères 4. CYRENE. After Smith and Porcher, 1864 5. SOLUNTUM. After Cavallari, 1875 6. PRIENE, GENERAL OUTLINE. After Zippelius 7. PRIENE, DETAILS OF A PART OF THE EXCAVATED AREA. After the large plan by Wiegand and Schrader, 1904 8. PRIENE, PANORAMA OF THE TOWN. As restored by Zippelius 9. MILETUS. After Wiegand, 191110. GERASA. After Schumacher11. TERRAMARA OF CASTELLAZZO DI FONTANELLATO. After T. E. Peet12. MARZABOTTO. After Brizio and Levi13. POMPEII. After Mau, 191014. MODENA. From the plan of Zuccagni-Orlandini, 184415. TURIN. Reduced from a plan published by the Society for the diffusion of Useful Knowledge (_Maps_, London, 1844, vol. Ii) after Zuccagni-Orlandini, 184416. AOSTA. From Promis and others17. FLORENCE. (A) Modern Florence. (B) After L. Bardi (1795?) and Zuccagni-Orlandini18. LUCCA. From Sinibaldi, 184319. HERCULANEUM. After Ruggiero and Beloch20. NAPLES. From the Neapolitan Government map of 186521. INSCRIPTION OF ORANGE. From the _Comptes-rendus de l'Académie des Inscriptions et Belles-Lettres_, 190422. TIMGAD. After R. Cagnat and the large plan by A. Ballu (_Ruines de Timgad, Sept années de découvertes_ (Paris, 1911))23. DETAILS OF INSULAE IN TIMGAD. After R. Cagnat, _Timgad_, p. 33724. A PART OF CARTHAGE. Plan based on the _Carte archéologique des ruines de Carthage_, by Gauckler and Delattre25. A PART OF LAIBACH. From a plan by Dr. W. Schmid (_VI. Bericht der römisch-germanischen Kommission_, 1910-1911)26. LINCOLN, OUTLINE OF ROMAN WALLS27. LINCOLN, BASES OF THE COLONNADE UNDER BAILGATE. From a photograph28. LINCOLN, SEWER UNDER BAILGATE. From a photograph29. AUTUN. After H. De Fontenay (_Autun et ses Monuments_, Autun, 1889)30. TRIER. Plan reduced from plan (1:10, 000) by the late Dr. Hans Gräven, _Die Denkmalpflege_, 14 Dec. 190431. SILCHESTER, GENERAL PLAN. Reduced from the large plan by W. H. St. John Hope (1:1800), _Archaeologia_ lxi, plate 8532. SILCHESTER, DETAILS OF FOUR INSULAE, THE FORUM AND CHRISTIAN CHURCH. From _Archaeologia_33. CAERWENT, GENERAL PLAN. Reduced from plan by F. King (1:900), _Archaeologia_ lxii, plate 6434. BOSTRA. From a plan in Baedeker's _Guide to Palestine_35. SAUVETERRE-DE-GUYENNE, A BASTIDE OF A. D. 1281. From plan by Dr. A. E. Brinckmann36. RUINS OF KHARA-KHOTO, A CHINESE TOWN OF ABOUT A. D. 1100. _Geographical Journal_, Sept. 1910

For the loan of blocks I am indebted to the Académie des Inscriptionset Belles-Lettres (fig. 21), to the German Imperial ArchaeologicalInstitute (fig. 9), to the Royal Geographical Society (fig. 36), andto the Royal Institute of British Architects and the Editors of the_Transactions of the Town-Planning Conference_, 1911 (figs. 7, 8, 17, 30, 32, 35). Fig. 11 is from Mr. T. E. Peet's _Stone and Bronze Ages inItaly_. The other 26 blocks have been prepared for this volume. 

TABLE OF MEASURES

The following figures may be found convenient by readers who wish totake special account of the dimensions cited in the following pages, and may also help them to correct any errors which I have unwittinglyadmitted. 

1 Roman foot = 0. 296 metres = 0. 97 English feet. For practicalpurposes 100 Roman feet = 97 English feet. 

1 Iugerum = 120 x 240 Roman feet = 116. 4 x 233. 8 English feet. Forpractical purposes a _Iugerum_ may be taken to be rather over2/3 of an acre and rather over 1/4 of a hectare, and more exactly2523. 3 sq. Metres. 

1 Metre = 1. 09 English yards, a trifle less than 40 ins. 402. 5 metresequal a quarter of a mile. 

1 Hectare (10000 sq. Metres) = 2. 47 acres (11955 sq. Yds. ). 

1 Acre = nearly 69-1/2 x 69-1/2 yds. (208. 7 ft. Square) = 4840 sq. Yds. 

CHAPTER I

PRELIMINARY REMARKS

Town-planning--the art of laying out towns with due care for thehealth and comfort of inhabitants, for industrial and commercialefficiency, and for reasonable beauty of buildings--is an art ofintermittent activity. It belongs to special ages and circumstances. For its full unfolding two conditions are needed. The age must be onein which, whether through growth, or through movements of population, towns are being freely founded or freely enlarged, and almost as amatter of course attention is drawn to methods of arranging and layingout such towns. And secondly, the builders of these towns must havewit enough to care for the well-being of common men and the duearrangement of ordinary dwellings. That has not always happened. Inmany lands and centuries--in ages where civilization has been tingedby an under-current of barbarism--one or both of these conditions havebeen absent. In Asia during much of its history, in early Greece, inEurope during the first half of the Middle Ages, towns have consistedof one or two dominant buildings, temple or church or castle, of oneor two processional avenues for worshippers at sacred festivals, and alittle adjacent chaos of tortuous lanes and squalid houses. Architectshave devised beautiful buildings in such towns. But they have nottouched the chaos or treated the whole inhabited area as one unit. Town-planning has been here unknown. [2]

 [2] Compare Brinckmann's remarks on mediaeval towns: 'Der Nachdruck liegt auf den einzelnen Gebäuden, der Kathedrale, dem Palazzo publico, den festen Palästen des Adels, nicht auf ibrer einheitlichen Verbindung. Ebenso erscheint die ganze Stadt nur eine Ansammlung einzelner Bauten. Strassen und Plätze sind unbebaute Reste. '

In other periods towns have been founded in large numbers andfull-grown or nearly full-grown, to furnish homes for multitudes ofcommon men, and their founders have built them on some plan or system. One such period is, of course, our own. Within the last half-centurytowns have arisen all over Europe and America. They are many innumber. They are large in area. Most of them have been born almostfull-grown; some have been established complete; others have developedabruptly out of small villages; elsewhere, additions huge enough toform separate cities have sprung up beside towns already great. Throughout this development we can trace a tendency to plan, beginningwith the unconscious mechanical arrangements of industrial cities orsuburbs and ending in the conscious efforts of to-day. 

If we consider their size and their number together, these newEuropean and American towns surpass anything that the world has yetseen. But, save in respect of size, the process of founding orenlarging towns is no new thing. In the old world, alike in the Greeklands round the eastern Mediterranean and in the wide empire of Rome, urban life increased rapidly at certain periods through theestablishment of towns almost full-grown. The earliest towns of Greeceand Italy were, through sheer necessity, small. They could not growbeyond the steep hill-tops which kept them safe, or house moreinhabitants than their scanty fields could feed. [3] But the world wasthen large; new lands lay open to those who had no room at home, andbodies of willing exiles, keeping still their custom of civil life, planted new towns throughout the Mediterranean lands. The process wasextended by state aid. Republics or monarchs founded colonies toextend their power or to house their veterans, and the results wereequally towns springing up full-grown in southern Europe and, westernAsia and even northern Africa. So too in remoter regions. Obscureevidence from China suggests that there also in early times towns wereplanted and military colonies were sent to outlying regions onsomewhat the same methods as were used by the Greeks and Romans. 

 [3] For the connexion between such towns and their local food-supply, note the story of Alexander the Great and the architect Dinocrates told by Vitruvius (II. I). Dinocrates had planned a new town; Alexander asked if there were lands round it to supply it with corn, and on hearing there were none, at once ruled out the proposed site. 

Even under less kindly conditions, the art has not been whollydormant. Special circumstances or special men have called it intobrief activity. The 'bastides' and the 'villes neuves' ofthirteenth-century France were founded at a particular period andunder special circumstances, and, brief as the period was and governedby military urgencies, they were laid out on a more or less definiteplan (p. 143). The streets designed by Wood at Bath about 1735, byCraig at Edinburgh about 1770, by Grainger at Newcastle about 1835, show what individual genius could do at favourable moments. But suchinstances, however interesting in themselves, are obviously lessimportant than the larger manifestations of town-planning in Greeceand Rome. 

In almost all cases, the frequent establishment of towns has beenaccompanied by the adoption of a definite principle of town-planning, and throughout the principle has been essentially the same. It hasbeen based on the straight line and the right angle. These, indeed, are the marks which sunder even the simplest civilization frombarbarism. The savage, inconsistent in his moral life, is equallyinconsistent, equally unable to 'keep straight', in his house-buildingand his road-making. Compare, for example, a British and a Roman road. The Roman road ran proverbially direct; even its few curves were notseldom formed by straight lines joined together. The British road wasquite different. It curled as fancy dictated, wandered along the footor the scarp of a range of hills, followed the ridge of winding downs, and only by chance stumbled briefly into straightness. Wheneverancient remains show a long straight line or several correctly drawnright angles, we may be sure that they date from a civilized age. 

In general, ancient town-planning used not merely the straight lineand the right angle but the two together. It tried very fewexperiments involving other angles. Once or twice, as at Rhodes (pp. 31, 81), we hear of streets radiating fan-fashion from a commoncentre, like the gangways of an ancient theatre or the thoroughfaresof modern Karlsruhe, or that Palma Nuova, founded by Venice in 1593 todefend its north-eastern boundaries, which was shaped almost like astarfish. But, as a rule, the streets ran parallel or at right anglesto each other and the blocks of houses which they enclosed were eithersquare or oblong. 

Much variety is noticeable, however, in details. Sometimes the outlineof the ancient town was square or almost square, the house-blocks wereof the same shape, and the plan of the town was indistinguishable froma chess-board. Or, instead of squares, oblong house-blocks formed apattern not strictly that of a chess-board but geometrical andrectangular. Often the outline of the town was irregular and merelyconvenient, but the streets still kept, so far as they could, to arectangular plan. Sometimes, lastly, the rectangular planning waslimited to a few broad thoroughfares, while the smaller side-streets, were utterly irregular. Other variations may be seen in the prominencegranted or refused to public and especially to sacred buildings. Insome towns full provision was made for these; ample streets withstately vistas led up to them, and open spaces were left from whichthey could be seen with advantage. In others there were neither vistasnor open spaces nor even splendid buildings. 

A measure of historical continuity can be traced in the occurrence ofthese variations. The towns of the earlier Greeks were stately enoughin their public buildings and principal thoroughfares, but theyrevealed a half-barbaric spirit in their mean side-streets andunlovely dwellings. In the middle of the fifth century men rose abovethis ideal. They began to recognize private houses and to attempt anadequate grouping of their cities as units capable of a single plan. But they did not carry this conception very far. The decorative stilldominated the useful. Broad straight streets were still few and werelaid out mainly as avenues for processions and as ample spaces forgreat facades. [4] Private houses were still of small account. Thenotion that the City was the State, helpful and progressive as it was, did something also to paralyse in certain ways the development ofcities. 

 [4] Pindar mentions 'the paved road cut straight to be smitten by horse-hoofs in processions of men that besought Apollo's care' at Cyrene (_Pyth. _ v. 90). An inscription from the Piraeus, of 320 B. C. , orders the Agoranomi (p. 37) to take care 'of the broad roads by which the processions move to the temple of Zeus the Saviour'. 

A change came with the new philosophy and the new politics of theMacedonian era. The older Greek City-states had been large, wealthy, and independent; magnificent buildings and sumptuous festivals were asnatural to them as to the greater autonomous municipalities in allages. But in the Macedonian period the individual cities sank to beparts of a larger whole, items in a dominant state, subjects ofmilitary monarchies. The use of public buildings, the splendour ofpublic festivals in individual cities, declined. Instead, the claimsof the individual citizen, neglected too much by the City-states butnoted by the newer philosophy, found consideration even intown-planning. A more definite, more symmetrical, often more rigidly'chess-board' pattern was introduced for the towns which now began tobe founded in many countries round and east of the Aegean. Ornamentaledifices and broad streets were still indeed included, but in thehouse-blocks round them due space and place were left for thedwellings of common men. For a while the Greeks turned their minds tothose details of daily life which in their greater age they hadsomewhat ignored. 

Lastly, the town-planning of the Macedonian era combined, as Ibelieve, with other and Italian elements and formed the town system ofthe later Roman Republic and the Roman Empire. As in art andarchitecture, so also in city-planning, the civilization of Greece andof Italy merged almost inextricably into a result which, with all itsGreek affinities, is in the end Roman. The student now meets arigidity of street-plan and a conception of public buildings which areneither Greek nor Oriental. The Roman town was usually a rectanglebroken up into four more or less equal and rectangular parts by twomain streets which crossed at right angles at or near its centre. Tothese two streets all the other streets ran parallel or at rightangles, and there resulted a definite 'chess-board' pattern ofrectangular house-blocks (_insulae_), square or oblong in shape, moreor less uniform in size. The streets themselves were moderate inwidth; even the main thoroughfares were little wider than the rest, and the public buildings within the walls were now merged in thegeneral mass of houses. The chief structure, the Forum, was anenclosed court, decorated indeed by statues and girt with colonnades, but devoid of facades which could dominate a town. The town councilsof the Roman world were no more free than those of Greece or modernEngland from the municipal vice of over-building. But they had not thesame openings for error. On the other hand, there was in most of thema good municipal supply of water, and sewers were laid beneath theirstreets. 

The reason for all this is plain. These Roman towns, even more thanthe Greek cities of the Macedonian world, were parts of a greaterwhole. They were items in the Roman Empire; their citizens werecitizens of Rome. They had neither the wealth nor the wish to buildvast temples or public halls or palaces, such as the Greeksconstructed. Their greatest edifices, the theatre and theamphitheatre, witness to the prosperity and population not so much ofsingle towns as of whole neighbourhoods which flocked in to periodicperformances. [5] But these towns had unity. Their various parts were, in some sense, harmonized, none being neglected and none grievouslyover-indulged, and the whole was treated as one organism. Despitelimitations which are obvious, the Roman world made a more real soberand consistent attempt to plan towns than any previous age hadwitnessed. 

 [5] Compare the crowd of Nucerians who made a riot in the amphitheatre at Pompeii in A. D. 59 (Tac. _Ann_. Xiv. 17). The common idea that the population of a town can be calculated by the number of seats in its theatre or amphitheatre is quite amiss. 

CHAPTER II

GREEK TOWN-PLANNING. THE ORIGINS, BABYLON

The beginnings of ideas and institutions are seldom well known or wellrecorded. They are necessarily insignificant and they win scant noticefrom contemporaries. Town-planning has fared like the rest. Earlyforms of it appear in Greece during the fourth and fifth centuriesB. C. ; the origin of these forms is obscure. The oldest settlement ofman in town fashion which has yet been explored in any land nearGreece is that of Kahun, in Egypt, dating from about 2500 B. C. HereProfessor Flinders Petrie unearthed many four-roomed cottages packedclose in parallel oblong blocks and a few larger rectangular houses:they are (it seems) the dwellings of the workmen and managers busywith the neighbouring Illahun pyramid. [6] But the settlement is verysmall, covering less than 20 acres; it is not in itself a real townand its plan has not the scheme or symmetry of a town-plan. For thatwe must turn to western Asia, to Babylonia and Assyria. 

 [6] W. F. Petrie, _Illahun, Kahun, and Gurob_ (London, 1891), ch. Ii, plate xiv. The plan is reproduced in Breasted's _History of Egypt_, p. 87, R. Unwin's _Town planning_, fig. 11 (with wrong scale), &c. 

Here we find clearer evidence. The great cities of the Mesopotamianplains show faint traces of town-planning datable to the eighth andfollowing centuries, of which the Greeks seem to have heard and whichthey may have copied. Our knowledge of these cities is, of course, still very fragmentary, and though it has been much widened by thelatest German excavations, it does not yet carry us to definiteconclusions. The evidence is twofold, in part literary, drawn fromGreek writers and above all Herodotus, and in part archaeological, yielded by Assyrian and Babylonian ruins. 

The description of Babylon given by Herodotus is, of course, famous. [7] Even in his own day, it was well enough known to beparodied by contemporary comedians in the Athenian theatre. Probablyit rests in part on first-hand knowledge. Herodotus gives us tounderstand that he visited Babylon in the course of his manywanderings and we have no cause to distrust him; we may even date hisvisit to somewhere about 450 B. C. He was not indeed the only Greek ofhis day, nor the first, to get so far afield. But his accountnevertheless neither is nor professes to be purely that of aneyewitness. Like other writers in various ages, [8] he drew no sharpdivision between details which he saw and details which he learnt fromothers. For the sake (it may be) of vividness, he sets them all on oneplane, and they must be judged, not as first-hand evidence but ontheir own merits. 

 [7] Hdt. I. 178 foil. The accounts of Ctesias and other ancient writers seem to throw no light on the town-planning and streets of Babylon, however useful they may otherwise be. 

 [8] The Elizabethan description of Britain by William Harrison is an example from a modern time. 

Babylon, says Herodotus, was planted in an open plain and formed anexact square of great size, 120 stades (that is, nearly 14 miles) eachway; the whole circuit was 480 stades, about 55 miles. It was girtwith immense brick walls, 340 ft. High and nearly 90 ft. Thick, and abroad deep moat full of water, and was entered through 100 gates;presumably we are intended to think of these gates as arrangedsymmetrically, 25 in each side. From corner to corner the city was cutdiagonally by the Euphrates, which thus halved it into two roughlyequal triangles, and the river banks were fortified by brickdefences--less formidable than the main outer walls--which ran alongthem from end to end of the city. There was, too, an inner wall on thelandward side. The streets were also remarkable:

 'The city itself (he says) is full of houses, three or four storeys high, and has been laid out with its streets straight, notably those which run at right angles, that is, those which lead to the river. Each road runs to a small gate in the brick river-wall: there are as many gates as lanes. '[9]

 [9] Hdt. I. 180 [Greek: To de astu auto, eon plêres ohikieôn triôrhofôn te kai tetrôrofôn, katatetmêtai tas hodous itheas, tas te aggas kai tas epikarsias, tas epi ton potamon echousas]. Apparently [Greek: epikarsias] means, as Stein says, those at right angles to the general course of the river, but this nearly = at right angles to the other roads. The course of the river appears to have been straighter then than it is now. 

In each part of the city (that is, on either bank of the Euphrates)were specially large buildings, in one part the royal palaces, in theother the temple of Zeus Belos, bronze-gated, square in outline, 400yards in breadth and length. 

So far, in brief, Herodotus. Clearly his words suggest town-planning. The streets that ran straight and the others that ran at right anglesare significant enough, even though we may doubt exactly what is meantby these other streets and what they met or cut at right angles. Buthis account cannot be accepted as it stands. Whatever he saw andwhatever his accuracy of observation and memory, not all of his storycan be true. His Babylon covers nearly 200 square miles; its walls areover 50 miles long and 30 yds. Thick and all but 120 yds. High; itsgates are a mile and a half apart. The area of London to-day is nomore than 130 square miles, and the topmost point of St. Paul's isbarely 130 yds. High. Nanking is the largest city-site in China andits walls are the work of an Empire greater than Babylon; but theymeasure less than 24 miles in circuit, and they are or were littlemore than 30 ft. Thick and 70 ft. High. [10] Moreover, Herodotus'saccount of the walls has to be set beside a statement which he makeselsewhere, that they had been razed by Darius sixty or seventy yearsbefore his visit. [11] The destruction can hardly have been complete. But in any case Herodotus can only have seen fragments, easilymisinterpreted, easily explained by local _ciceroni_ as relics ofsomething quite unlike the facts. 

 [10] L. Gaillard, _Variétés sinologiques_, xvi (plan) and xxiii. Pp. 8, 235 (Chang-hai, 1898, 1903). Others give the figures a little differently, but not so as to affect the argument. 

 [11] Hdt. Iii. 159. The theory that there were originally two parallel outer walls, that Darius razed one and Herodotus saw the other (Baumstark in Pauly-Wissowa, _Real-Encycl. _ ii. 2696), is meaningless. There could be no use in razing one and leaving the other, which was almost as strong (Hdt. I. 181). It is, however, not quite certain that Herodotus (i. 181) meant that there were two outer parallel walls. 

Turn now to the actual remains of Babylon, as known from surveys andexcavations. We find a large district extending to both banks of theEuphrates, which is covered rather irregularly by the mounds of manyruined buildings. Two sites in it are especially notable. At itssouthern end is Birs Nimrud and some adjacent mounds, ancientlyBorsippa; here stood a huge temple of the god Nebo. Near its northend, ten or eleven miles north of Borsippa, round Babil and Kasr, is alarger wilderness of ruin, three miles long and nearly as broad inextreme dimensions; here town-walls and palaces of Babylonian kingsand temples of Babylonian gods and streets and dwelling-houses ofordinary men have been detected and in part uncovered. Other signs ofinhabitation can be traced elsewhere in this district, as yetunexplored. 

Not unnaturally, some scholars have thought that this whole regionrepresents the ancient Babylon and that the vast walls of Herodotusenclosed it all. [12] This view, however, cannot be accepted. Quiteapart from the considerations urged above, the region in question isnot square but rather triangular, and traces of wall and ditchsurrounding it are altogether wanting, though city-walls have survivedelsewhere in this neighbourhood and though nothing can wholly deletean ancient ditch. We have, in short, no good reason to believe thatBabylon, in any form or sense whatever, covered at any time this largearea. 

 [12] So Baumstark, art. Babylon in Pauly-Wissowa, ii. 2696. 

On the other hand, the special ruins of Babil and Kasr and adjacentmounds seem to preserve both the name and the actual remains ofBabylon (fig. 1). Here, on the left bank of the Euphrates, are vastcity-walls, once five or six miles long. [13] They may be describedroughly as enclosing half of a square bisected diagonally by theriver, much as Herodotus writes; there is good reason to think thatthey had some smaller counterpart on the right bank, as yet scantilyexplored. Within these walls were the palaces of the Babylonian kings, Nabopolassar and Nebuchadnezzar (625-561 B. C. ), the temples of thenational god Marduk or Merodach and other Babylonian deities, a broadstraight road, Aiburschabu, running north and south from palaces totemples, a stately portal spanning this road at the Istar Gate, manyprivate houses in the Merkes quarter, and an inner town-wall perhapsof earlier date. Street and gate were built or rebuilt byNebuchadnezzar. He, as he declares in various inscriptions, 'paved thecauseway with limestone flags for the procession of the Great LordMarduk. ' He made the Istar Gate 'with glazed brick and placed on itsthreshold colossal bronze bulls and ferocious serpent dragons'. Alongthe street thus built the statue of Marduk was borne in solemn marchon the Babylonian New Year's Day, when the king paid yearly worship tothe god of his country. [14]

 [13] F. H. Weissbach, _Stadtbild von Babylon_ (_Der alte Orient_, fasc. 5); R. Koldewey, _Tempel von Babylon und Borsippa_, plates i, ii; S. Langdon, _Expositor_, 1909, pp. 82, 142; Hommel, _Geogr. Des alten Orients_, pp. 290, 331; E. Meyer, _Sitzungsber. Preuss. Akad_. 1912, p. 1102. I am indebted to Dr. Langdon for references to some of the treatises cited here and below. I cannot share the unfavourable view which is taken by Messrs. How and Wells, the latest good editors of Herodotus, of the views of these writers. 

 [14] Koldewey, _Pflastersteine von Aiburschabu_ (Leipzig, 1901). Some of the streets of Babylon are much older than 600 B. C. , but this point needs to be worked out further. 

[Illustration: FIG I. BABYLON]

Such are the remains of the city of Babylon, so far as they are knownat present. They do not fit ill with the words of Herodotus. We candetect in them the semblance not indeed of one square but of twounequal half-squares, divided by the river; we can trace at least onegreat street parallel to the river and others which run at rightangles to it towards the river. If the brick defences along thewater-side have vanished, that may be due to their less substantialcharacter and to the many changes of the river itself. To the studentof Babylonian topography, the account of Herodotus is of very littleworth. But it is as good as most modern travellers could compile, ifthey were let loose in a vast area of buildings, without plans, without instruments, and without any notion that a scientificdescription was expected of them. 

The remains show also--and this is more to our purpose--the idea ofthe sacred processional avenue which recurs in fifth-centuryGreece--and is indeed beloved of architects in the most modern times. Here is a germ of town-planning. But whether this laying out ofstreets extended beyond the main highways, is less clear. The Merkesexcavations occasionally show streets meeting at right angles and atleast one roughly rectangular _insula_, of 150 x 333 ft. But theadjoining house-blocks agree neither in size nor shape, and no hintseems to have yet come to light of a true chess-board pattern. [15]

 [15] _Mitteilungen der deutschen Orient-Gesellschaft_ 42, Dec. 1909, pp. 7, 19; 44, Dec. 1910, p. 26. 

A little further evidence can be drawn from other Mesopotamian sites. The city of Asshur had a long, broad avenue like the sacred road ofBabylon, but the one _insula_ of its private houses which has yet beenexcavated, planned and published, shows no sign of rectangularplanning. [16] There is also literary evidence that Sanherib (765-681B. C. ) laid out a 'Kingsway' 100 ft. Wide to promote easy movementthrough his city of Nineveh, and Delitzsch has even credited theSargonid dynasty generally (722-625 B. C. ) with a care for thedwellings of common men as well as of gods and of kings. [17]

 [16] _Mitt, deutsch. Orient-Gesell. _ 28, Sept. 1905; 31, May 1906. 

 [17] F. Delitzsch, _Asurbanipal und die assyr. Kultur seiner Zeit_ (_Der alte Orient_, Leipzig, 1909), p. 25. 

In conclusion, the mounds of Babil and Kasr and others near them seemto represent the Babylon alike of fact and of Herodotus. It was asmaller city than the Greek historian avers; its length and breadthwere nearer four than fourteen miles. But it had at least onestraight, ample, and far-stretching highway which gave space for theceremonies and the processions, if not for the business or thedomestic comforts, of life. In a sense at least, it was laid out withits streets straight. Nor was it the only city of such a kind in theMesopotamian region. Asshur and Nineveh, both of them somewhat earlierin date than Babylon, possessed similar features. These towns, or atleast Babylon, seem to have been known to Greek travellers, andprobably suggested to them the adornment of their Hellenic homes withsimilar streets. The germ of Greek town-planning came from the east. 

CHAPTER III

GREEK TOWN-PLANNING: FIRST EFFORTS

Greek town-planning began in the great age of Greece, the fifthcentury B. C. But that age had scant sympathy for such a movement, andits beginnings were crude and narrow. Before the middle of the centurythe use of the processional highway had established itself in Greece. Rather later, a real system of town-planning, based on streets thatcrossed at right angles, became known and practised. Later still, inthe early fourth century, the growing care for town-life produced townby-laws and special magistrates to execute them. In some form orother, town-planning had now taken root in the Greek world. 

The two chief cities of Greece failed, indeed, to welcome the newmovement. Both Athens, the city which by itself means Greece to mostof us, and Sparta, the rival of Athens, remained wholly untouched byit. Alike in the days of Themistocles and Pericles and in all itslater history, Athens was an almost Oriental mixture of splendidpublic buildings with mean and ill-grouped houses. An often-quotedsaying of Demosthenes puts the matter in its most favourable light:

'The great men of old built splendid edifices for the use of theState, and set up noble works of art which later ages can never match. But in private life they were severe and simple, and the dwelling ofan Aristides or a Miltiades was no more sumptuous than that of anyordinary Athenian citizen' (Third Olynthiac Oration, 25). 

This is that 'desire for beauty and economy' which Pericles (orThucydides) praised in the Funeral Oration. It has a less lovely side. Not a few passages in Greek literature speak, more or less clearly, ofthe streets of Athens as narrow and tortuous, unpaved, unlighted, andmore like a chaos of mud and sewage than even the usual Greek road. Sparta was worse. There neither public nor private buildings wereadmirable, and the historian Thucydides turned aside to note themeanness of the town. 

Nevertheless, the art of town-planning in Greece probably began inAthens. The architect to whom ancient writers ascribe the first step, Hippodamus of Miletus, --born about or before 480 B. C. , --seems to haveworked in Athens and in connexion with Athenian cities, under theauspices of Pericles. The exact nature of his theories has not beenrecorded by any of the Greek writers who name him. Aristotle, however, states that he introduced the principle of straight wide streets, andthat he, first of all architects, made provision for the propergrouping of dwelling-houses and also paid special heed to thecombination of the different parts of a town in a harmonious whole, centred round the market-place. But there seems to be no evidence forthe statement sometimes made, that he had any particular liking foreither a circular or a semicircular, fan-shaped town-plan. 

_Piraeus_ (fig. 2). 

Three cities are named as laid out by Hippodamus. Aristotle tells usthat he planned the Piraeus, the port of Athens, with broad straightstreets. He does not add the precise relation of these streets to oneanother. If, however, the results of recent German inquiries andconjectures are correct, and if they show us his work and not--as isunfortunately very possible--the work of some later man, his designincluded streets running parallel or at right angles to one anotherand rectangular blocks of houses; the longer and presumably the moreimportant streets ran parallel to the shore, while shorter streets ranat right angles to them down to the quays. Here is a rectangularscheme of streets, though the outline of the whole town is necessarilynot rectangular (fig. 2). 

[Illustration: FIG. 2. PLAN OF PIRAEUS]

_Thurii_. 

Another town ascribed to Hippodamus is the colony which the Atheniansand others planted in 443 B. C. At Thurii in southern Italy, of whichHerodotus himself is said to have been one of the original colonists. Its site has never been excavated, and indeed one might doubt whetherexcavation would show the street plan of 443 B. C. Or that of a laterand possibly even of a Roman age, when the town was recolonized on theRoman system. But the historian Diodorus, writing in the first centuryB. C. And no doubt embodying much older matter, records a pertinentdetail. The town, he says, was divided lengthways by four streets andcrossways by three. Plainly, therefore, it had a definite andrectangular street-planning, though the brevity of the historian doesnot enable us to decide how many house-blocks it had and how far thelesser streets were symmetrical with these seven principalthoroughfares. In most of the cases which we shall meet in thefollowing sections of this treatise, the number of streetsrunning-straight or at right angles is very much greater than thenumber assigned to Thurii. I may refer for example to the plans ofPriene, Miletus, and Timgad. 

_Rhodes_. 

A third city assigned to Hippodamus is Rhodes. This, according toStrabo, was laid out by 'the architect of the Piraeus'; according toothers, it was built round its harbour like the seats of an ancienttheatre round the orchestra, that is, fan-fashion like Karlsruhe. However, this case is doubtful. Rhodes was laid out in 408 B. C. , thirty-five years after the planting of Thurii and seventy years afterthe approximate date of the birth of Hippodamus. It is conceivable butnot altogether probable that Hippodamus was still planning towns inhis extreme old age, nor is it, on political grounds, very likely thathe would be planning in Rhodes. As, however, we do not know the realdate of his birth, and as Strabo does not specifically mention hisname, certainty is unattainable. [18]

 [18] On Hippodamus see K. F. Hermann, _de Hippodamo Milesio_ (Marburg, 1841) and Erdmann, _Philologus_ xlii. 193-227, and _Programm Protestant. Gymnasium zu Strassburg_, 1883. As will be seen, I do not accept all Erdmann's conclusions. For the Piraeus see Aristotle, _Politics_, II. 8 = p. 1267 and IV. 11 = p. 1330. For Thurii see Diodorus XII. 10. For Rhodes see Strabo 654 = XIV. Ii. 9: E. Meyer, _Gesch. Des Alt. _ iv. Pp. 60, 199 rejects the tale. For plans of the Piraeus see Wachsmuth, _Stadt Athen im Alterthum_, ii. 134, and Curtius and Kaupert, _Karten von Attika_ (1881), plan II_a_ by Milchhöfer. Foucart has adduced epigraphic reasons for dating the work of Hippodamus here to 480-470 B. C. (_Journal des Savants_, 1907, pp. 178-82); they are not conclusive, but, if he be right, the difficulty of assigning the Piraeus and Rhodes to the same architect becomes even greater. The town-plan of Piraeus given by Gustav Hirschfeld (_Berichte der sãchs. Ges. Der Wissenschaften_, 1878, xxx. I) is not convincing, nor do I feel very sure even about Milchhöfer's results. 

If we cannot tell exactly how Hippodamus planned cities or exactlywhich he planned, still less do we know how far town-planning on hisor on any theory came into general use in his lifetime or indeedbefore the middle of the fourth century. Few Greek cities have beensystematically uncovered, even in part. Fewer still have revealedstreet-planning which can be dated previous to that time. It does notfollow, when we find streets in the ruins of an ancient city, thatthey must belong to its earliest period. That is not true of towns inany age, modern or mediaeval, Roman or Greek. Some Greek cities werefounded in early times, were rebuilt in the Macedonian period, andagain rebuilt in the Roman period. Without minute excavation it may beimpossible to assign the town-plan of such a place to its proper placeamong these three periods. 

We have, however, at Selinus in Sicily and Cyrene on the north coastof Africa, two cases which may belong to the age of Hippodamus. Theyare worth describing, since they illustrate both the difficulty ofreaching quite certain conclusions and also the system which probablydid obtain in the later fifth and the early fourth century. 

_Selinus_ (fig. 3). 

At Selinus the Italian archaeologists discovered some years ago, inthe so-called Acropolis, a town of irregular, rudely pear-shapedoutline with a distinct though not yet fully excavated town-plan. Twomain thoroughfares ran straight from end to end and crossed at rightangles (fig. 3), the longer of these thoroughfares being just aquarter of a mile long and 30 ft. Wide. From these two main streetsother narrower streets (12-18 ft. Wide) ran off at right angles; theresult, though not chess-board pattern, is a rectangular town-plan. Unfortunately, it cannot be dated. Selinus was founded in 648 B. C. , was destroyed in 409, then reoccupied and rebuilt, and finallydestroyed for ever in 249. Its town-planning, therefore, might be asearly as the seventh century B. C. Or (and this is the most probableconclusion) it may date from the days of Selinuntine prosperity justbefore 409, when the city was growing and the great Temple of Zeus orApollo was rising on its eastern hill. Or again, though less probably, it may have been introduced after 400. We may conclude that we havehere a clear case of town-planning and we may best refer it to thelater part of the fifth century. [19]

[Illustration: FIG. 3. PLAN OF SELINUS]

 [19] Koldewey and Puchstein, _Die griech. Tempel in Unteritalien und Sicilien_, p. 90, plan 29, from Cavallari; Hulot and Fougères, _Sélinonte_, Paris, 1910, pp. 121, 168, 196. The latter writers assign the rebuilding to Hermocrates, 408-407 B. C. But our accounts of Hermocrates do not suggest that he rebuilt anything at Selinus of any sort, except defences. 

_Cyrene_ (fig. 4). 

[Illustration: FIG. 4. PLAN OF CYRENE]

At Cyrene the researches of two English archaeologists about 1860disclosed a town-plan based, like that of Selinus, on two main streetswhich crossed at right angles (fig. 4). Here, however, the otherstreets do not seem to have been planned uniformly at right angles tothe two main thoroughfares, and the rectangular scheme is thereforeless complete and definite than at Selinus. Cyrene, unfortunately, resembles Selinus in another respect, that we have no proper knowledgeof the date when its main streets were laid out. It was foundedsomewhere in the seventh century B. C. And Pindar, in an ode writtenabout 466 B. C. , mentions a great processional highway there. Whetherthis was one of the two roads above mentioned is not clear. But it isnot probable, since Pindar's road seems hardly to have been inside thecity at all. [20]

 [20] Smith and Porcher, _Discoveries at Cyrene_ (1864), plate 40; hence Studnickza, _Kyrene_ (1890, p. 167, fig. 35), and Malten, _Kyrene_ (Berlin, 1911). For Pindar's reference see Pyth. V. 90 and p. 16 above. 

In these two cases and in one or two others which might be noted fromthe same or later times, the town-scheme includes rectangular elementswithout any strict resemblance to the chess-board pattern. Thedominant feature is the long straight street, of great width andsplendour, which served less as the main artery of a town than as afrontage for great buildings and a route for solemn processions. Here, almost as in Babylon, we have the spectacular element which architectslove, but which is, in itself, insufficient for the proper dispositionof a town. Long and ample streets, such as those in question, mighteasily be combined, as indeed they are combined in some modern townsof southern Europe and Asia, with squalid and ill-groupeddwelling-houses. Hippodamus himself aimed at something much better, asAristotle tells us. But it was not till after 350 B. C. Or someapproximate date, that dwelling-houses were actually arranged andgrouped on a definite system. [21]

 [21] Soluntum, near Palermo, on the north coast of Sicily, was found by Cavallari in 1875 to exhibit a rectangular street-plan; one main street ran north and south along level ground and several lesser streets lay at right angles to it mounting a hillside by means of steps (as at Priene, p. 42). See the _Bullettino delta Commissione di Antichità e Belle Arti in Sicilia_, viii. Palermo, August 1875. Cavallari himself assigned this plan to the date when Soluntum was founded--which is unfortunately uncertain--but only on the general ground that 'in una città, una volta tracciate le strade e disposte le arterie dicommunicazione, non è facile cambiarne la disposizione generale'. I attach less weight than he does to this reason. Soluntum was in the main and by origin a Phoenician town, with a Greek colouring; in 307 B. C. It was refounded for the discharged soldiers of Agathocles; later still, in Roman times, it had the rank of 'municipium'; most of its ruins are generally considered to be of Roman date and small objects found in it are also mostly Roman, and its street-plan may also be Roman. As the 'Bullettino' is somewhat rare, I add a reduced plan (fig. 5). 

[Illustration: FIG. 5. SOLUNTUM]

It was probably, however, in the first half of the fourth century thatthe Greek cities began to pass by-laws relating to the police, thescavenging and the general public order of their markets and streets, and to establish Agoranomi to control the markets and Astynomi tocontrol the streets. These officials first appear in inscriptionsafter 350, but are mentioned in literature somewhat earlier. Anaccount of the Athenian constitution, ascribed formerly to Xenophonand written (as is now generally agreed) about 430-424 B. C. , mentionsbriefly the prosecution of those who built on to the public land, thatis (apparently), who encroached upon the streets. But it is silent asto specific officers, Astynomi or other. Plato, however, in his'Laws', which must date a little earlier than his death in 347, alludes on several occasions to such officers. They were to look afterthe private houses 'in order that they may all be built according tolaws', and to police and clean the roads and water-channels, bothinside and outside of the city. A prohibition of balconies leaningover the public streets, and of verandas projecting into them, is alsomentioned in two or three writers of the fourth century and is said togo back to a much earlier date, though its antiquity was probablyexaggerated. [22]

 [22] Plato, _Laws_ 763 c, 779 c, &c. ; Aristotle, _Ath. Pol. _ 50; Arist. , _Oec. _ ii. 5, p. 134; Xenophon, _Ath. Pol. _ iii. 4; Schol. To Aeschines, iii. 24. The fact that the word 'Astynomos' occurs in Aeschylus does not justify the writer of an article in Pauly-Wissowa (_Real-Encycl. _ ii. 1870) in stating that magistrates of this title were already at work in the earlier part of the fifth century; the poet uses the noun in a general sense from which it was afterwards specialized. Some of the regulations recur at Rome (p. 137). 

The municipal by-laws which these passages suggest clearly came intouse before, though perhaps not long before, the middle of the fourthcentury. They do not directly concern town-planning; they involvebuilding regulations only as one among many subjects, and thoseregulations are such as might be, and in many cases have been, adoptedwhere town-planning was unknown. But they are natural forerunners ofan interest in town-planning. As in modern England, so infourth-century Greece, their appearance suggests the growth of a carefor well-ordered town life and for municipal well-being which leadsdirectly to a more elaborate and methodical oversight of the town asan organized combination of houses and groups of houses. 

As we part from this early Greek town-planning, we must admit thataltogether we know little of it. There was such a thing: among itsmain features was a care for stately avenues: its chief architect wasHippodamus. Thus much is clear. But save in so far as Milchhöfer'splans reproduce the Piraeus of B. C. 450 or 400, we cannot discerneither the shape or the size of the house-blocks, or the groupingadopted for any of the ordinary buildings, or the scheme of theordinary roads. We may even wonder whether such things were of muchaccount in the town-planning of that period. 

CHAPTER IV

GREEK TOWN-PLANNING: THE MACEDONIAN AGE, 330-130 B. C. 

The Macedonian age brought with it, if not a new, at least a moresystematic, method of town-planning. That was the age when Alexanderand his Macedonian army conquered the East and his successors forseveral generations ruled over western Asia, when Macedonians andGreeks alike flocked into the newly-opened world and Graeco-Macedoniancities were planted in bewildering numbers throughout its length andbreadth. Most of these cities sprang up full-grown; not seldom theirfirst citizens were the discharged Macedonian soldiery of the armiesof Alexander and his successors. The map of Turkey in Asia is full ofthem. They are easily recognized by their names, which were oftentaken from those of Alexander and his generals and successors, theirwives, daughters, and relatives. Thus, one of Alexander's youngestgenerals, afterwards Seleucus I, sometimes styled Nicator, foundedseveral towns called Seleucia, at least three called Apamea, andothers named Laodicea and Antiochia, thereby recording himself, hisIranian wife Apama, his mother Laodice and his father Antiochus, andhis successors seem to have added other towns bearing the same name. Indeed, two-thirds of the town-names which are prominent in the laterhistory of Asia Minor and Syria, date from the age of Alexander andhis Macedonians. 

Many discoveries show that these towns were laid out with a regular'chess-board' street-plan. That method of town-planning now madedefinite entry into the European world. No architect or statesman isrecorded to have invented or systematically encouraged it. Alexanderhimself and his architect, one Dinocrates of Rhodes or perhaps ofMacedonia, seem to have employed it at Alexandria in Egypt, and thismay have set the fashion. Seven years after Alexander's death itrecurs at Nicaea in Bithynia, which was refounded by one ofAlexander's successors in 323 B. C. And was laid out on this fashion. But no ancient writer credits either the founder or the architect ofAlexandria or the founder of Nicaea with any particular theory on thesubject. If the chess-board fashion becomes now, with seemingsuddenness, the common--although not the universal--rule, that isprobably the outcome of the developments sketched in the last chapter. Approximations to chess-board planning had been here and thereemployed in the century before Alexander. When his conquests and theircomplicated sequel led, amongst other results, to the foundation ofmany new towns, it was natural that the most definite form of planningshould be chosen for general use. 

We might, however, wonder whether its adoption was helped by themilitary character of the generals who founded, and the dischargedsoldiers who formed the first inhabitants of so many among thesetowns. Military men are seldom averse to rigidity. It is worth noting, in this connexion, that when chess-board planning came into common usein the Roman Empire, many--perhaps most--of the towns to which it wasapplied were 'coloniae' manned by time-expired soldiers. So, too, inthe Middle Ages and even in comparatively modern times, the towns laidout with rectangular street-plans in northern Italy, in Provence, inthe Rhine Valley, are for the most part due in some way or other tomilitary needs. [23] In our own days rectangular planning is a dominantfeature of the largest and newest industrial towns. They are adaptinga military device to the purposes of an industrial age. 

 [23] Since the invention of artillery, the rectangular street-plan has been regarded by soldiers as useful in defending the streets of a town. Aristotle, however, expressly observes in the _Politics_ that, in street warfare, tortuous lanes were far better than straight avenues for the defence, and he recommends that the rectangular pattern should be adopted only 'in parts and in places', though he does not explain how this would work out (_Politics, _ iv. 11, p. 1330). 

[Illustration: FIG. 6. GENERAL OUTLINE OF PRIENE. A, B, C. Gates. D, E, F, H, M, P. Temples (see fig. 7). G. Agora, Market. I. Council House, K. Prytaneion. L, Q. Gymnasium. N. Theatre, O. Water-reservoir, R. Race-course. ]

[Illustration: FIG. 7. PART OF PRIENE AS EXCAVATED 1895-8. (From the large plan by Wiegand and Schrader. )]

[Illustration: FIG. 8. PRIENE, PANORAMA OF THE TOWN. (As restored by Zippelius. )]

_Priene_ (figs. 6-8). 

The best instance of the new system is not perhaps the most famous. Priene was a little town on the east coast of the Aegean. The highridge of Mycale towered above it; Miletus faced it across an estuary;Samos stood out seawards to the west. In its first dim days it hadbeen perched on a crag that juts out from the overhanging mountain;there its life began, we hardly know when, in the dawn of Greekhistory. But it had been worn down in the fifth century between theupper and the nether millstone of the rival powers of Samos andMiletus. Early in the Macedonian age it was refounded. The oldAcropolis was given up. Instead, a broad sloping terrace, or moreexactly a series of terraces, nearer the foot of the hill, was laidout with public buildings--Agora, Theatre, Stoa, Gymnasium, Temples, and so forth--and with private houses. The whole covered an area ofabout 750 yds. In length and 500 yds. In width. Priene was, therefore, about half the size of Pompeii (p. 63). It had, as its excavatorscalculate, about 400 individual dwelling-houses and a populationpossibly to be reckoned at 4, 000. 

In the centre was the Agora or market-place, with a temple and otherlarge buildings facing on to it; round them were other publicbuildings and some eighty blocks of private houses, each blockmeasuring on an average 40 x 50 yds. And containing four or fivehouses. The broader streets, rarely more than 23 ft. Wide, ran levelalong the terraces and parallel to one another. Other narrowerstreets, generally about 10 ft. Wide, ran at right angles up theslopes, with steps like those of the older Scarborough or ofAssisi. [24] The whole area has not yet been explored and we do notknow whether the houses were smaller or larger, richer or poorer, inone quarter than in another, but the regularity of the street-plancertainly extended over the whole site. 

 [24] Compare Soluntum, p. 36, n. 2. 

Despite this reasoned and systematic arrangement, no striking artisticeffects appear to have been attempted. No streets give vistas ofstately buildings. No squares, save that of the Agora--120 by 230 ft. Within an encircling colonnade--provide open spaces where largerbuildings might be grouped and properly seen. Open spaces, indeed, such as we meet, in mediaeval and Renaissance Italy or in modernEnglish towns of eighteenth century construction, were very rare inPriene. Gardens, too, must have been almost entirely absent. In thearea as yet uncovered, scarcely a single dwelling-house possessed anygarden ground or yard. [25]

 [25] Wiegand and Schrader, _Priene, Ergebnisse der Ausgrabung in den Jahren 1895-8_ (Berlin, 1904). Professor P. Gardner gave a good account to the Town-Planning Conference (_Proceedings_, pp. 112-122). I am indebted to him for two of my illustrations. 

_Miletus_ (fig. 9). 

The skill of German archaeologists has revealed what town-planningmeant in a small town rebuilt in the Alexandrine period. No other evenapproximately complete example has been as yet uncovered on any othersite. But spade-work at the neighbouring and more famous city ofMiletus has uncovered similar street-planning there. In one quarter, the only one yet fully excavated, the streets crossed at right anglesand enclosed regular blocks of dwelling-houses measuring 32 x 60 yds. (according to the excavators) but sub-divided into blocks of about 32yds. Square (fig. 9). These blocks differ somewhat in shape from thoseof Priene, which are more nearly square; whether they differ in dateis more doubtful. They are certainly not earlier than the Macedonianera, and one German archaeologist places the building or rebuilding ofthis quarter of Miletus after that of Priene and in a 'lateHellenistic' and apparently Roman period. There is unquestionably muchRoman work in Miletus; there seems, however, no sufficient reason forascribing the house-blocks shown on fig. 7 to any date but some partof the Macedonian period. Though differently shaped, they do notdiffer very greatly in actual area from those of Priene. They aresomewhat smaller, but only by about 60 sq. Yds. In each average-sizedplot. [26]

 [26] Wiegand, _Abhandlungen der Berliner Akademie_, 1911, Anhang; _Archäol. Anzeiger_, 1911, 420 foll. 

[Illustration: FIG. 9. MILETUS, AS EXCAVATED BY WIEGAND. (_Archãologischer Anzeiger_, 1911, p. 421. )]

_Alexandria_. 

A yet more famous town, founded by Alexander himself, is definitelyrecorded by ancient writers to have been laid out in the samequasi-chess-board fashion, with one long highway, the Canopic Street, running through it from end to end for something like four miles. [27]Unfortunately the details of the plan are not known with anycertainty. Excavations were conducted at the instigation of NapoleonIII in 1866 by an Arab archaeologist, Mahmud Bey el Fallaki, and, according to him, showed a regular and rectangular scheme in whichseven streets ran east and west while thirteen ran north and south atright angles to them. The house-blocks divided by these streets werethought to vary somewhat in size but to measure in general about 300 x330 metres. [28] More recent research, however, has not confirmedMahmud's plans. The excavations of Mr. Hogarth and M. Botti suggestthat many of his lines are wrong and that even his Canopic Street isincorrectly laid down. Mr. Hogarth, indeed, concludes that 'it ishopeless now to sift his work; those who would treat the site ofAlexandria scientifically must ignore him and start _de novo_'. Morerecent excavation, carried out by Dr. Noack in 1898-9, seemed to showthat the ancient streets which can now be traced beneath Alexandriabelong to a Roman age, though they may of course follow older lines, and that, if some items in Mahmud's plans are possibly right, theerrors and omissions are serious. We may accept as certain thestatement that Alexandria was laid out with a rectangular town-plan;we cannot safely assume that Mahmud has given a faithful picture ofit. [29]

 [27] Strabo, xvii. 793. 

 [28] Mahmud Bey, _Mémoire sur l'ancienne Alexandrie_ (Copenhagen, 1872); Néroutsos Bey, _L'ancienne Alexandrie_ (Paris, 1888). 

 [29] D. G. Hogarth, _Archaeological Report of the Egypt Exploration Fund_, 1894-5, p. 28, and _Hellenic Journal_, xix. 326; F. Noack, _Athen. Mitteil. _ xxv. (1900), pp. 232, 237. Dr. Noack thought that his results confirmed Mahmud; to me, as to some others, they seem rather to yield the conclusions indicated in the text. 

_Nicaea_. 

Priene, Miletus, and Alexandria supply more or less well-knowninstances of Macedonian town-planning. They can be reinforced by acrowd of less famous examples, attested by literature or by actualremains. One of the most characteristic is known to us fromliterature, Nicaea in Bithynia, founded by one of the Macedonians in316 B. C. And renamed by another some years later in honour of his wifeNicaea. Strabo, writing about A. D. 15, describes it and hisdescription no doubt refers to arrangements older than the Romans. Itformed, he says, a perfect square in which each side measured fourstades, a little over 800 yds. In each side--apparently in the middleof each side--there was one gate, and the streets within the wallswere laid out at right angles to one another. A man who stood at acertain spot in the middle of the Gymnasium could see straight to allthe four gates. [30] Here is the chess-board pattern in definite form, though the central portion of the city may have been laid out underthe influence of spectacular effect rather than of geometry. 

 [30] Strabo, 565, 566. 

_Sicyon, Thebes, &c. _

Another Macedonian town-plan may be found at Sicyon, a little west ofCorinth. This old Greek city was rebuilt by Demetrius Poliorcetesabout 300 B. C. , and is described by a Greek writer of the firstcentury B. C. As possessing a regular plan and roads crossing at rightangles. The actual remains of the site, explored in part by Englishand French archaeologists early in the nineteenth century, show somestreets which run with mathematical straightness from north-east tosouth-west and others which run from north-west to south-east. [31]These streets might, indeed, date from the period when Sicyon was thechief town of the Roman province of Achaia, the period (that is)between the overthrow of Corinth in 146 B. C. And its restoration justa century later. But that was not an epoch when such rebuilding islikely to have been carried through. Friendly as the Republicangovernment of Rome showed itself in other ways to Hellas, there is noreason to think that it spent money on town-planning in Helleniccities. It is far more probable that the town-plan of Sicyon datesfrom the Macedonians. 

 [31] Diodorus Sic. Xx. 102; _Expédition scientifique de Morée, archit. Et sculpture_, iii (1838), plate LXXXI. 

To the same Macedonian epoch we may perhaps ascribe the building orrather the rebuilding of Boeotian Thebes, which one who passes for acontemporary writer under the name of Dicaearchus, describes as'recently divided up into straight streets'. [32] To the same periodStrabo definitely assigns the newer town of Smyrna, lying in the plainclose to the harbour. It was due, he says, to the labours of theMacedonians, Antigonus, and Lysimachus. [33] We may perhaps assign tothe same period the town-planning of Mitylene in Lesbos, whichVitruvius mentions as so splendid and so unhealthy, were it not thathis explanation of its unhealthiness suggests rather a fan-shapedoutline than a square. It was, he says, intolerable, whatever windmight blow. With a south wind, the wind of damp and rain, every onewas ill. With a north-west wind, every one coughed. With a north wind, no one could stand out of doors for the chilliness of its blasts. [34]Streets that lay open to the north and the north-west and the south, equally and alike, could only be found in a town-plan fashioned like afan. But perhaps Vitruvius only selected three of the plagues ofLesbos. 

 [32] Dicaearchus, p. 143. 

 [33] Strabo, 646. 

 [34] Vitruvius, i. 6. 

In other cases the same planning was probably adopted, although theevidence as yet known shows only a rectangular plan of main streets, such as we have met in Pre-Macedonian Greece. In Macedonia itself, Thessalonika, laid out perhaps about 315 B. C. , had at least one mainstreet running southwards to the sea and two more running east andwest at right angles to that. [35] In Asia two Syrian towns, whichoccupy sites closed to Hellenic culture before Alexander, may serve asexamples. Apamea on the Orontes was built by the Macedonians, roseforthwith to importance, and retained its vigorous prosperity throughthe Roman Empire; in A. D. 6 it was 'numbered' by Sulpicius Quirinius, then the governor of Syria, and the census showed as many as 117, 000citizens settled in the city and its adjacent 'territory'. Its ruinsseem to be mainly earlier than the Romans, and its streets may welldate from its Macedonian founders. In outline it is an irregularoblong, nearly an English mile in length and varying in width fromhalf to two-thirds of a mile. A broad and straight street, linedthroughout with colonnades, runs from end to end of its length andpasses at least five great buildings, which seem to be the temples andpalaces of the Seleucid kings. Two other streets cross this mainstreet at right angles. Whether the smaller thoroughfares took thesame lines can be determined only by excavation. It would be a gentleguess to think so. [36]

 [35] Tafrali, _Topographie de Thess. _ pp. 121 foll. And plan. 

 [36] E. Sachau, _Reise in Syrien_ (1883), p. 76; Mommsen, _Ephemeris epigr_. Iv, p. 514, and _Mon. Ancyr. _ (ed. 2), p. 540. 

Further south, on the edge of the Haurân, stood the town of Gerasa. This too, like Apamea, was built by the Macedonians and flourished notonly in their days but during the following Roman age. Its generaloutline was ovoid, its greatest diameter three quarters of a mile, itsarea some 235 acres--nearly the same with Roman Cologne and RomanCirencester. Its streets resembled those of Apamea. A colonnadedhighway ran straight through from north to south; two other streetscrossed at right angles, and its chief public buildings, the Temple ofthe Sun and three other temples, two theatres and two public baths, stood near these three streets (fig. 10). Again the evidence provesrectangular town-planning in broad outline; excavation alone can tellthe rest. [37]

 [37] _Zeitschrift des deutschen Palãstina-Vereins_, xxv (1902), plate 6; Bãdeker, _Palestine and Syria_ (1906), p. 140. For the neighbouring Bostra, see p. 136. 

[Illustration: FIG. 10. GERASA]

In the towns just described a distinctive feature is the 'chess-board'pattern of streets and rectangular house-blocks. That, of course, isthe feature which most concerns us here. It may not have looked sopredominant to their builders and inhabitants. The towns which theMacedonians founded were not seldom rich and large; several were thecapitals of powerful and despotic rulers. In such towns we expectgreat public buildings, temples, palaces. It is not surprising ifsometimes those who reared them cared solely for the spectaculargrouping of magnificent structures and forgot the private houses andthe general plan of the town. 

_Pergamum_. 

One such instance from the Macedonian age, perhaps the mostinstructive which we could ever hope to get, [38] is Pergamum, in thenorth-west of Asia Minor. This has been thoroughly explored by Germanscience; its remains are superb; its chief buildings date from an agewhen town-planning had grown familiar to the Greek world. About 300B. C. It was a hill-town where a Macedonian chief could bestow awar-chest. It grew both populous and splendid in the third and secondcenturies B. C. Under the Attalid kings; later builders, Augustus orTrajan or other, added little either to its general design or to itsarchitectural glory. The dominant idea was that of a semi-circle ofgreat edifices, crowning the crest and inner slopes of a highcrescent-shaped ridge. Near the northern and highest end of this ridgestood the palace of the Attalid princes, afterwards buried beneath atemple in honour of Trajan. Next, to the south, was the Library--withstores of papyri worth more perhaps to the world than all thearchitecture of Pergamon. The middle of the crescent held the shrineof Athena, goddess of Pergamon, and beside it the Altar of Zeus theSaviour, gigantic in size, splendid with sculpture, itself the equalof an Acropolis. Lastly, the southern or lower end of the ridge bore atemple of Dionysus and an Agora for Assemblies. 

 [38] Ephesus, refounded by Lysimachus about 281 B. C. , might perhaps be another. But the repeated excavations there, though they have taught us much about the temples and other large edifices of the great city, seem to have left the streets comparatively unexplored. 

These buildings ringed the hill-top in stately semi-circle; belowthem, a theatre was hewn out of the slopes and a terrace 250 yds. Longwas held up by buttresses against precipitous cliffs. Lower yet, beneath the Agora, the town of common men covered the lower hill-sidein such order or disorder as its steepness allowed. Here was noconventional town-planning. Only a yet lower and later city, built inRoman days on more or less level spaces beside the stream Selinus, seems perhaps to have been laid out in chess-board fashion. [39] TheAttalid kings, the founders of Pergamon, cared only for splendidbuildings splendidly adorned. If their abrupt hill-side forbade thestraight and broad processional avenues of some other Greek cities, they crowned their summits instead with a crescent of temples andpalaces which had not its like on the shores of the Aegean. 

 [39] P. Schatzmann, _Athen. Mitteil_. Xxxv. (1910) 385; _Archãol. Anzeiger_ (1910), p. 541. This lowest city is covered by a swarm of modern houses and hovels, and has not been very fully explored. 

Yet even Pergamon had its building-laws and by-laws for the protectionof common life. A Pergamene inscription contains part of a 'Royal Law'which apparently dates from one of the Attalid rulers. It isimperfect. But we can recognize some of the items for which itprovided. Houses which fell or threatened to fall on to the publicstreet, or which otherwise became ruinous, could be dealt with by theAstynomi; if their owners failed to repair them, these magistrateswere to make good the defects themselves and to recover the cost, anda fine over and above it, from the owners; if the Astynomi neglectedtheir duty, the higher magistrates, the Strategi, were to take up thematter. Streets were to be cleaned and scavenged by the same Astynomi. Brick-fields were expressly forbidden within the city. The widths ofroads outside the town were fixed and owners of adjacent land wereheld liable for their repair, and there was possibly some similarrule, not preserved on the inscription, for roads inside the walls; atPriene, it seems, these latter were in the care of the municipality. There were provisions, too, for the repair of common walls whichdivided houses belonging to two owners, and also for the prevention ofdamp where two houses stood side by side on a slope and the wall ofthe lower house stood against the soil beneath the upper house. [40]

 [40] Kolbe, _Athen. Mitteil_. Xxvii. 47 and xxix. 75; Hitzig, _Zeitschrift der Savigny-Stiftung, roman. Abteilung_ xxvi. 433. 

These rules are very like those which were coming into use before 330B. C. (p. 37). Only, they are more elaborate, and it is significantthat the inscriptions begin in Macedonian and later days to give moreand fuller details as to the character of these laws and as to theexistence in many cities of officials to execute them. It is notsurprising to find that Roman legislation of the time of Caesar andthe early Empire applies these or very similar rules to the localgovernment of the Roman municipalities of the Empire (p. 137). 

So common in the Macedonian world was the town-planning which has beendescribed above, that the literature of the period, even in its casualphrases and incidental similes, speaks of towns as being normallyplanned in this fashion. Two examples from two very different authorswill suffice as illustration. Polybius, writing somewhere about B. C. 150, described in well-known chapters the scheme of the Roman camp, and he concludes much as follows: 'This being so, the whole outline ofthe camp may be summed up as right-angled and four-sided andequal-sided, while the details of its street-planning and its generalarrangement are precisely parallel to those of a city' (VI. 31, 10). He was comparing the Greek town, as he knew it in his own country, with the encampment of the Roman army; he found in the town the aptestand simplest parallel which he could put before his readers. A muchlater writer, living in a very different environment and concernedwith a very different subject, fell nevertheless under the influenceof the same ideas. Despite his 'sombre scorn' for things Greek andRoman, St. John, when he wished to figure the Holy City Jerusalem, centre of the New Heaven and New Earth, pictured it as a city lyingfoursquare, the length as large as the breadth, and entered by twelvegates, 'on the east three gates, on the north three gates, on thesouth three gates, and on the west three gates. '[41]

 [41] Revelation xxi. 13, 16. Some of the details are, no doubt, drawn from the later chapters of Ezekiel, but the difference between the two writers is plain. 

The instances and items cited in the preceding paragraphs lie withinthe limits of the Greek world and of the Roman Empire. We mightperhaps wish to pursue our speculations and ask whether this vigoroussystem influenced foreign lands, and whether the Macedonian armycarried the town-plan of their age, in more or less perfect form, asfar as their conquests reached. Alexander settled many soldiers inlands which were to form his eastern and north-eastern frontiers, asif against the central-asiatic nomads. Merv and Herat, Khokand andKandahar, [42] have been thought--and, it seems, thought with somereason--to date from the Macedonian age and in their first period tohave borne the name Alexandria. But no Aurel Stein has as yetuncovered their ruins, and speculation about them is mere speculation. 

 [42] See p. 145 below. 

CHAPTER V

ITALIAN TOWN-PLANNING. THE ORIGINS

If Greek and Macedonian town-planning are fairly well known, the RomanEmpire offers a yet larger mass of certain facts, both in Italy and inthe provinces. The beginnings, naturally, are veiled in obscurity. Wecan trace the system in full work at the outset of the Empire; wecannot trace the steps by which it grew. Evidences of something thatresembles town-planning on a rectangular scheme can be noted in two orthree corners of early Italian history--first in the prehistoricBronze Age, then in a very much later Etruscan town, and thirdly onone or two sites of middle Italy connected with the third or fourthcentury B. C. These evidences are scanty and in part uncertain, andtheir bearing on our problem is not always clear, but they claim aplace in an account of Italian town-planning. To them must be added, fourthly, the important evidence which points to the use of a systemclosely akin to town-planning in early Rome itself. 

_The Terremare_ (fig. 11). 

(i) We begin in the Bronze Age, somewhere between 1400 and 800 B. C. , amidst the so-called Terremare. More than a hundred of these strangesettlements have been examined by Pigorini, Chierici, and othercompetent Italians. Most of them occur in a well-defined districtbetween the Po and the Apennines, with Piacenza at its west end andBologna at its east end. Some have also been noted on the north bankof the Po near Mantua, both east and west of the Mincio, and two orthree elsewhere in Italy. Archaeologically, they all belong to theBronze Age; they seem, further, to be the work of a race distinct fromany previous dwellers in North Italy, which had probably just movedsouth from the Danubian plains. At some time or other this race haddwelt in lake-villages. They were now settled on dry ground and faraway from lakes--one of their hamlets is high in the Apennines, nearly1, 900 ft. Above the sea. But they still kept in the Terremare thelacustrine fashion of their former homes. 

The nature of these strange villages can best be explained by anaccount of the best-known and the largest example of them (fig. 11). At Castellazzo di Fontanellato, a little west of Parma, are thevestiges of a settlement which, with its defences, covered an area ofabout forty-three acres. In outline it was four-sided; its east andwest sides were parallel to one another, and the whole resembled arectangle which had been pulled a trifle askew. Round it ran a solidearthen rampart, 50 ft. Broad at the base and strengthened withwoodwork (plan, B). In front of the rampart was a wet ditch (A), 100ft. Wide, fed with fresh water from a neighbouring brook by an inletat the south-western corner (C) and emptied by an outfall on the east(D). One wooden bridge gave access to this artificial island at itssouthern end (E). The area within the rampart, a little less thanthirty acres in extent, was divided into four parts by two mainstreets, which would have intersected at right angles had the placebeen strictly rectangular; other narrower streets ran parallel tothese main thoroughfares. On the east side (F) was a small'citadel'--_arx_ or _templum_--with ditch, rampart and bridge of itsown (G, H); in this were a trench and some pits (K) which seemed bytheir contents to be connected with ritual and religion. Outside thewhole (L, M) were two cemeteries, platforms of urns set curiously likethe village itself, and also a little burning _ghat_. [43] Thepopulation of the village is necessarily doubtful. A German writer, Nissen, has reckoned it at four or five thousand, men, women andchildren together, crowded into small huts. But this estimate may betoo high. In any case, many of the Terremare are much smaller. 

 [43] The literature of the Terremare is very large. The results obtained up to 1894 were summarized by F. Von Duhn in the _Neue Heidelberger Jahrbücher_, iv. 144; the best recent accounts are by T. E. Peet, _Stone and Bronze Ages in Italy_ (Oxford, 1909), chaps. 14 and 17, from which fig. 11 is taken, and R. Munro, _Palaeolithic Man and Terramara Settlements_ (Edin. , 1912), pp. 291-487 and plates xxxiii foll. A good brief sketch is given by Mr. H. S. Jones, _Companion to Roman History_, pp. 4-6. One point in the arrangement seems not quite clear. It is generally stated that the trapezoidal outline was adopted in order to allow the water to enter the ditch from a running stream and to part easily into two channels (fig. 11). That is quite intelligible. But, if so, one would expect the outlet to be at the opposite end, and not (as it actually is) in the middle of one side, where it would 'short-circuit' the current. (Mr. H. S. Jones seems to have confused inlet and outlet. )

[Illustration: FIG. 11. TERRAMARA OF CASTELLAZZO DI FONTANELLATO]

These Terremare bear a strong likeness to the later Italiantown-planning, and they are usually taken to be the oldestdiscoverable traces of that system. This means that the Italiantown-planning was derived from other sources besides Greece or theEast, since the Terremare are far older than Hippodamus or evenNebuchadnezzar and Sennacherib (pp. 23, 29). It must be added that ourpresent knowledge does not allow us to follow the actual developmentof the Terremare into historic times, and to link them closely withthe later civilization of Central Italy. When some modern scholarscall the men of the Terremare by the name 'Italici', they express ahope rather than a proven fact. It may be safer, for the moment, toavoid that name and to refrain from theories as to the exact relationbetween prehistoric and historic. But we shall see below that theexistence of a relation between the two is highly probable. 

_Marzabotto_ (fig. 12). 

[Illlustration: FIG. 12. MARZABOTTO. (AB, FG, CD, main streets. The shading represents excavated houses. )]

(ii) A greater puzzle, dating probably from the fifth century B. C. , meets us in the ruins of a nameless little Etruscan town which stoodoutside of Etruria proper, on the north slopes of the Apennines. Itssite is fifteen miles south of Bologna, close to the modernMarzabotto, on the left bank of the little river Reno. Only a tinypart has been uncovered. But the excavators have not hesitated tocomplete their results conjecturally into a rectangular town-plan, with streets crossing at right angles and oblong blocks of housesmeasuring from 158 to 176 yds. In length and 37 or 44 or 71 yds. Inwidth (fig. 12). The whole must have been laid out at once, and thesmaller remains seem to show that this was done by Etruscans. In thefourth century the place was sacked by the Gauls, and though there waslater occupation, [44] its extent is doubtful. [45]

Further excavation is, however, needed to confirm this generallyaccepted interpretation of the place. Nothing has been noted elsewherein Etruria or its confines to connect the Etruscans with anyrectangular form of town-plan. At Veii, for example, most of theEtruscan city has lain desolate and unoccupied ever since the Romansdestroyed it, but the site shows no vestige of streets crossing atright angles or of oblong blocks of houses. At Vetulonia the excavatedfragment of an Etruscan city shows only curving and irregularstreets. [46] Nor is there real reason to believe that the 'Etruscanteaching' learnt by Rome included an art of town-planning (p. 71) orthat, as a recent French writer has conjectured, the Etruscans broughtany such art with them from the East and communicated it to the West. We must conclude that at Marzabotto we have a piece of evidence whichwe cannot set into its proper historical framework. We might perhapscall it an early blend of Greek and Italian methods and compare itwith Naples (p. 100). It is odd that four out of seven house-blocksshould measure just under 120 Roman ft. In width and thus approximateto a figure which we meet often elsewhere in the Roman world (p. 79). But it would be well to learn more of the plan by further excavation. 

 [44] _Archaeological Journal_, 1903, p. 237. 

 [45] Brizio, _Monumenti Antichi_, i. 252, superseding Gozzadini's _Antica Necropoli a Marzabotto_ (Bologna, 1865-70); Grenier, _Bologne villanovienne_ &c. (Paris, 1912) p. 98. Compare _Authority and Archaeology_, pp. 305, 306. 

 [46] _Notizie degli Scavi_ 1895, p. 272; Durm, _Baukunst der Etr_. P. 39. 

_Pompeii_ (fig. 13). 

(iii) A third piece of evidence can be found on a site whichhistorians and novelists alike connect mainly with the Roman Empire, but which dates back to the days of the early or middle Republic. Pompeii began in or before the sixth century B. C. As an Oscan city. For a while, we hardly know when, it was ruled by Etruscans. Later, about 420 B. C. , it was occupied by Samnites. Finally, it became Roman;it was refounded in 80 B. C. As a 'colonia' and repeopled by soldiersdischarged from the armies of Sulla. In A. D. 79 it reached its end inthe disaster to which it owes its fame. Its life, therefore, was longand full of destruction, re-building, enlargement. Its architecturalhistory is naturally hard to follow. Many of its buildings, however, can be dated more or less roughly by the style of their ornament orthe character of their material, and the lines of its streets suggestsome conjectures as to its growth which deserve to be stated eventhough they may conflict with the received opinions about Pompeii. Itwill be understood, of course, that these conjectures, like allspeculations on Pompeii, are limited by the fact that barely half ofits area has been as yet uncovered, and that very little search hasbeen made beneath the floors and pavements of its latest period. [47]

 [47] For recent plans of Pompeii the reader may consult the second edition (1908) of August Mau's _Pompeii_, or the fifth edition (1910) of his _Führer durch Pompeii_, re-edited by W. Barthel. A plan on a large scale is given in the last part of _CIL_. Iv (1909); there are also occasional plans in the _Notizie degli Scavi_. See also C. Weichardt, _Pompeji vor der Zerstorung_ (Leipzig, 1897). 

[Illustration: FIG. 13. POMPEII. (T = Temple. The area of the supposed original settlement is outlinedin black. )]

As we know it at present, Pompeii is an irregular oval area of about160 acres, planted on a small natural hill and girt with a stone wallnearly two miles in circumference (fig. 13). On the west there wasoriginally access to the sea, and on this side the walls havedisappeared or have not been yet uncovered. Near this end of the townis the Forum, with the principal temples and public buildings roundit. At the east end of the town, nearly 1200 yds. From the westernextremity, is the amphitheatre, and the town-walls appear to have beendrawn so as to include it. Two main streets, now called the Strada diNola and the Strada dell' Abbondanza, cross the town from SW. To NE. The main streets from NW. To SE. Are less distinct, but the StradaStabiana certainly ran from wall to wall. While there is someappearance of symmetry in the streets generally, it does not go veryfar; there is hardly a right angle, or any close approach to a rightangle, at any street corner. 

It is generally held, as Mau has argued, that the whole town was laidout at once, perhaps during the Etruscan period, on one plan ofstreets crossing at right angles. Two principal streets, those nowstyled the Strada di Mercurio and the Strada di Nola, are consideredto be the main streets of this earliest town-plan, and to give it itsgeneral direction. A third main street, the Strada Stabiana, whichcuts obliquely across from the Vesuvian to the Stabian Gate and marsthe supposed symmetry of this town-plan, is ascribed to the influenceof a small natural depression along which it runs, while a small areaeast of the Forum, which also breaks loose from the general scheme, isthought to have been laid out abnormally in order to remedy the effectof this obliquity. [48]

This theory is open to objections. In the first place the streets(even apart from those just east of the Forum) do not really form onesymmetrical plan. Region VI fits very ill with Regions I and III. Bothindicate systematic planning. But Region VI is laid out in oblongblocks 110 ft. Wide and either 310 ft. Or 480 ft. Long, while RegionsI and III are made up of approximately square blocks about 200 ft. Each way. Moreover, the orientation of the blocks is different. Thosein Region VI follow the lines of the Strada di Mercurio; those ofRegions I and II, and perhaps also of Region V, are dominated by theStrada Stabiana. Yet there is no obvious reason why this differenceshould not have been avoided; it results, indeed, in awkward cornersand inconvenient spaces. Nor, again, can we accept as in any degreeadequate the cause assigned by Mau for the odd orientation of thestreets next to the east side of the Forum. 

 [48] Mau, _Führer_ (1910), p. 5, 'um die Schiefwinkeligkeit zu vermindern. ' Truly, a very inadequate reason. 

These streets which lie round and east of the Forum suggest adifferent development. Pompeii may have begun with a little Oscan townplanted in what became its south-western corner, near the Water-Gateand the Forum, within the area of Regions II and IV. Here is a littlenetwork of streets, about 300 by 400 yds. Across (25 acres), whichharmonizes ill with the streets in the rest of the town, which liesclose to the river-haven on the Sarno, which includes the Forum andBasilica--probably the oldest public sites, though not the oldestsurviving structures, in Pompeii--and which is large enough to haveformed the greater part or even the whole of a prehistoric city. Theearliest building as yet excavated at Pompeii, the Doric Temple, withits precinct now known as the Forum Triangulare, stood on the edge ofthis area looking out from its high cliff over the plain of the Sarno. Originally this Temple may have stood just within the first town-wall, or perhaps just without it, sheltered by the precipice which itcrowns. This area has all the appearance of an 'Altstadt'. No doubt ithas been much altered by later changes. In particular, Forum andBasilica have grown far beyond their first proportions, and thebuildings which surround them have been added, altered, enlarged outof all resemblance to the original plan. Nevertheless, this theoryseems to account better than any other for this curious little cornerof streets that are hardly regular even in their relations to oneanother and are wholly irreconcilable to the rest of the town. 

Round this primitive city grew up the greater Pompeii. The growth musthave been rather by two or three distinct accretions than a gradualand continuous development. At present we cannot trace these stages. To do that we must wait till the excavations can be carried deeperdown, and till the other half of the city has been uncovered, or atleast till the lines of its streets and the shapes of its house-blockshave been determined, like those of Priene (p. 42), by specialinquiry. All that is as yet certain is that Regions I, III, V, and VIwere laid out, and their houses were (in part at least) in existencebefore--perhaps long before--80 B. C. , when the Sullan colony wasplanted, [49] and we see also that Region VI is planned differentlyfrom I and III. 

 [49] Region VI contains an ancient column of the sixth century B. C. (Mau, _Führer_, p. 113), but this may not be _in situ_. 

Another fact claims notice. The town-planning of Pompeii is in themain trapezoidal, not rectangular. Neither its oblongs, nor itssquares, nor its street-crossings exhibit true right angles, thoughmany of the rooms and peristyles in the private houses are regularenough. In this feature Pompeii resembles the trapezoidal outlines ofthe Terremare (fig. 11). It resembles also much Roman military work, both of Republican and of Imperial date, which disregards the strictright angle and accepts squares and oblongs which are, so to say, askew. The motive of the Terremare is supposed to have been, as I havesaid above, that of providing an easy flow for the water in theencircling moat. The motive of various military camps may perhaps befound rather in a wish to secure the same area as that of an orthodoxrectangle, even though the ground forbade the strict execution of theorthodox figure. Whatever the reason, the trapezoidal house-blocks ofPompeii exhibit a feature which is not alien to the earliertown-planning of Italy, though it is strange to the cities of Greece. 

_Norba_. 

Not only do we need to know more of Pompeii itself. We need evidencealso from other Italian towns of similar age. Here our ignorance isdeep. Only one site which can help has been even tentatively explored. Norba, which once crowned a spur of the Monti Lepini above the Pontinemarshes, was founded as a Roman town, according to the orthodoxchronology, in 492 B. C. [50] But the received chronology of the earlierRepublic, minute as it looks, probably deserves no more credence thanthe equally minute but mainly fictitious dates assigned by the SaxonChronicle to the beginnings of English History. Actual remains foundat Norba suggest rather that it was founded (not necessarily by Rome)about, or a little before, 300 B. C. ; it is therefore later than theTerremare and Marzabotto, and later also than the Oscan age ofPompeii. On the other hand, it came to an end in the Sullan period (82B. C. ). Its excavation has little more than begun, but it alreadyindicates a scheme of streets somewhat resembling that of Pompeii, [51]and it is a useful adjunct to our better knowledge of the more famoustown. The two together furnish examples of the town-planning of middleItaly of about 400-300 B. C. , in days that are only half historic, andthus help to fill the gap between the Terremare and the fullydeveloped system of the Roman Imperial period. 

 [50] Livy ii. 34, contradicted, however, by xxvii. 10 and by Dionysius Halic. Vii. 13 _ad fin_. 

 [51] _Notizie degli Scavi_, 191, p. 558, 1903, p. 261; Frothingham, _Roman Cities_, plate ix. I am indebted to Dr. T. Ashby, Director of the British School at Rome, for information as to the site. Excavations made in 1823 at the Roman Falerii (founded 241 B. C. ) show streets crossing at right angles, but the piece unearthed was small and the date uncertain (Canina, _Etruria Maritima_ i, plate ix). 

It may be permitted in this context to add a plan of a north Italiancity, in which some of the modern streets recall one quarter ofPompeii (fig. 14). Modena, the Roman Mutina, was founded as a'colonia' with 2, 000 male settlers in 183 B. C. , and despite variousmisfortunes became one of the chief towns in the Lombard plain. Onepart of this town shows a row of long narrow blocks measuring about 20x 160 metres (fig. 14, plan A), with a second row of shorter blocks ofthe same width and about half the length (plan B). These blocks havebeen much marred and curtailed by the inevitable changes of town life, but their symmetry cannot be accidental, and if they date back, as isquite possible, to Roman days, they may be put beside the Sixth Regionof Pompeii which contains two rows of similar blocks. [52]

 [52] Fig. 14 is taken from Zuccagni-Orlandini (1844). Kornemann suggests that Mutina was refounded about 40-20 B. C. , but there seems to be no evidence of this break in its continuity. 

[Illustration: FIG. 14. MODENA. See p. 69. ]

(iv) There remains, fourthly, evidence relating to early Rome itself, and to customs and observances which obtained there. These customsbelong to the three fields of religion, agrarian land-settlement andwar. All three exhibit the same principle, the division of a definitespace by two straight lines crossing at right angles at its centre, and (if need be) the further division of such space by other linesparallel to the two main lines. The Roman augur who asked the will ofHeaven marked off a square piece of sky or earth--his _templum_--intofour quarters; in them he sought for his signs. The Roman general whoencamped his troops, laid out their tents on a rectangular patterngoverned by the same idea. The commissioners who assignedfarming-plots on the public domains to emigrant citizens of Rome, planned these plots on the same rectangular scheme--as the map ofrural Italy is witness to this day. 

These Roman customs are very ancient. Later Romans deemed them asancient as Rome itself, and, though such patriotic traditions belongrather to politics than to history, we find the actual customs wellestablished when our knowledge first becomes full, about 200 B. C. [53]The Roman camp, for example, had reached its complex form long beforethe middle of the second century, when Polybius described it in words. Here, one can hardly doubt, are things older even than Rome. Scholarshave talked, indeed, of a Greek origin or of an Etruscan origin, andthe technical term for the Roman surveying instrument, _groma_, hasbeen explained as the Greek word 'gnomon', borrowed through anEtruscan medium. But the name of a single instrument would not carrywith it the origin of a whole art, even if this etymology were morecertain than it actually is. Save for the riddle of Marzabotto (p. 61), we have no reason to connect the Etruscans with town-planning orwith the Roman system of surveying. When the Roman antiquary Varroalleged that 'the Romans founded towns with Etruscan ritual', he setthe fashion for many later assertions by Roman and modern writers. [54]But he did not prove his allegation, and it is not so clear as isgenerally assumed, that he meant 'Etruscan ritual' to includearchitectural town-planning as well as religious ceremonial. 

 [53] The prologue to the Poenulus of Plautus (verse 49) which mentions 'limites' and a 'finitor', may well be as old as Plautus himself. But the 'centuriation' still visible in north Italy around colonies planted about 180 B. C. Is no full proof of rectangular surveying at that date. These towns were re-founded at a much later date, and their lands, and even their streets, _may_ have been laid out anew. 

 [54] Varro _ling. Lat_. 5. 143 _oppida condebant Etrusco ritu, id est, iunctis bobus_, cf. Frontinus _de limit_. (grom. I. P. 27). 

These are Italian customs, far older than the beginnings of Greekinfluence on Rome, older than the systematic town-planning of theGreek lands, and older also than the Etruscans. They should be treatedas an ancestral heritage of the Italian tribes kindred with Rome, andshould be connected with the plan of Pompeii and with the far olderTerremare. Many generations in the family tree have no doubt beenlost. The genealogy can only be taken as conjectural. But it is areasonable conjecture. 

In their original character these customs were probably secular ratherthan religious. They took their rise as methods proved by primitivepractice to be good methods for laying out land for farming or forencamping armies. But in early communities all customs that touchedthe State were quasi-religious; to ensure their due performance, theywere carried out by religious officials. At Rome, therefore, moreespecially in early times, the augurs were concerned with thedelimitation alike of farm-plots and of soldiers' tents. Theytestified that the settlement, whether rural or military, was dulymade according to the ancestral customs sanctioned by the gods. After-ages secularized once more, and as they secularized, they alsointroduced science. It was, perhaps, Greek influence which brought ina stricter use of the rectangle and a greater care for regularplanning. 

It may be asked how all this applies to the planning of towns. Wepossess certainly no such clear evidence with respect to towns as withrespect to divisions agrarian or military. But the town-plans which weshall meet in the following chapters show very much the same outlinesas those of the camp or of the farm plots. They are based on the sameessential element of two straight lines crossing at right angles inthe centre of a (usually) square or oblong plot. This is an elementwhich does not occur, at least in quite the same form, at Priene or inother Greek towns of which we know the plans, and it may well becalled Italian. We need not hesitate to put town and camp side byside, and to accept the statement that the Roman camp was a city inarms. Nor need we hesitate to conjecture further that in the planningof the town, as in that of the camp, Greek influence may have added amore rigid use of rectangular 'insulae'. When that occurred, will bediscussed in Chapter VI. 

Whether the nomenclature of the augur, the soldier and theland-commissioner was adopted in the towns, is a more difficult, butfortunately a less important question. Modern writers speak of the_cardo_ and the _decumanus_ of Roman towns, and even apply to themmore highly technical terms such as _striga_ and _scamnum_. For theuse of _cardo_ in relation to towns there is some evidence (p. 107). But it is very slight, and for the use of the other terms there isnext to no evidence at all. [55] The silence alike of literature and ofinscriptions shows that they were, at the best, theoreticalexpressions, confined to the surveyor's office. [56]

 [55] Whether the _possessores ex vico Lucretio scamno primo_ of Cologne (Corpus XIII. 8254) had their property inside the 'colonia' of that place or in the country outside, may be doubted (Schulten, _Bonner Jahrb. _ ciii. 28). 

 [56] The phrase Roma Quadrata ought, perhaps, to be mentioned in this chapter. It does not seem, however, to be demonstrably older than the Ciceronian age. The line _et qui sextus erat Romae regnare quadratae_, once attributed to Ennius (ed. Vablen, 1854, 158), is clearly of much later date. As a piece of historical evidence, the phrase merely sums up some archaeologist's theory (very likely a correct theory, but still a theory) that the earliest Rome on the Palatine had a more or less rectangular outline. 

CHAPTER VI

ITALIAN TOWN-PLANNING: THE LATE REPUBLIC AND EARLY EMPIRE

During the later Republic and the earlier Empire many Italian townswere founded or re-founded. To this result several causes contributed. Like the Greeks before them, the Romans of the Republic sent out fromtime to time compact bodies of emigrants whenever the home populationhad grown too large for its narrow space. These bodies were each largeenough to form a small town, and thus each migration meant--or mightmean--the foundation of a new town full-grown from its birth. TheGreeks generally established new and politically independent towns. The Romans followed another method. Their colonists remained subjectto Rome and constituted new centres of Roman rule, smallquasi-fortresses of Roman dominion in outlying lands. Often themilitary need for such a stronghold had more to do with the foundationof a 'colonia' than the presence of too many mouths in the city. Cicero, speaking of a 'colonia' planted at Narbo (now Narbonne) insouthern Gaul about 118 B. C. , and planted perhaps with some regard toan actual overflow of population in contemporary Rome, calls itnevertheless 'a colonia of Roman citizens, a watch-tower of the Romanpeople, a bulwark against the wild tribes of Gaul'. Those words statevery clearly the main object of many such foundations under Republicand Empire alike. 

Another reason for the establishment of 'coloniae' may be found in thehistory of the dying Republic and nascent Empire. During the civilwars of Sulla, of Caesar and of Octavian, huge armies were broughtinto the field by the rival military chiefs. As each conflict ended, huge masses of soldiery had to be discharged almost at once. For thesake of future peace it was imperative that these men should bequickly settled in some form of civic life in which they would abide. The form chosen was the familiar form of the 'colonia'. Thetime-expired soldiers were treated--not altogether unreasonably--assurplus population, and they were planted out in large bodies, sometimes in existing towns which needed population or at least aloyal population, sometimes in new towns established full-grown forthe purpose. This method of dealing with discharged soldiers wascontinued during the early Empire, though it was then employedsomewhat intermittently and the 'coloniae' were oftener planted in theprovinces than in Italy itself; indeed the establishment of Italian'coloniae', as distinct from grants of colonial rank by way of honour, almost ceased after A. D. 68. 

It is not easy to determine the number of such new foundations oftowns in Italy. Some seventy or eighty are recorded from the early andmiddle periods of the Republic--previous to about 120 B. C. ; Sullaadded a dozen or so; Octavian (Augustus) in his earlier yearsestablished or helped to establish about thirty. [57] But these figurescan hardly represent the whole facts. The one certainty is that, through the causes just detailed, a very large number of the Italiantowns were either founded full-grown or re-founded under newconditions during the later Roman Republic and the earlier Empire. Fewtowns in Italy developed as Rome herself developed, expanding fromsmall beginnings in a slow continuous growth which was governed byconvenience and opportunism and untouched by any new birth orsystematic reconstruction. 

 [57] See Mommsen, _Gesamm. Schriften_ v. 203; Nissen, _Ital. Landeskunde_ ii. 27; Kornemann in Pauly-Wissowa, _Encycl. _ iv. 520 foll. 

Coincident with these processes of urban expansion, we find, in manytowns which can be connected with the later Republic or the Empire, examples of a definite type of town-planning. This type has obviousanalogies with earlier Italy and with the town-planning of the Greekworld, but is also in certain respects distinct from either. The townareas with which we have now to deal are small squares or oblongs;they are divided by two main streets into four parts and by other andparallel streets into square or oblong house-blocks ('insulae'), andthe rectangular scheme is carried through with some geometricalprecision. The 'insulae', whatever their shape--square or oblong--arefairly uniform throughout. Only, those which line the north side ofthe E. And W. Street are often larger than the rest (pp. 88, 125). [58]The two main streets appear to follow some method of orientationconnected with augural science. As a rule, one of them runs north andsouth, the other east and west, and now and again the latter streetseems to point to the spot where the sun rises above the horizon onthe dawn of some day important in the history of the town. [59]

 [58] Modern plans seem sometimes to imply that the 'insulae' which abutted on the walls were also abnormally large. That is because the corresponding modern blocks often include, with the original 'insula', the space between it and the wall, and also the wall itself which has been disused and built over. 

 [59] See on this point some remarks by W. Barthel, _Bonner Jahrbücher_, cxx. 101-108. 

The public buildings of these towns are in general somewhat small andarranged with little attempt at processional or architecturalsplendour; they seldom dominate or even cross the scheme of streets. Open spaces are rare; the Forum, which corresponds to the Greek Agora, contains, like that, a paved open court, but this court is almost asmuch enclosed as the cloister of a mediaeval church or the quadrangleof a mediaeval college. Theatre and amphitheatre[60] might, no doubt, reach huge dimensions, but externally they were more often massivethan ornamental and the amphitheatre often stood outside the citywalls. Here and there a triumphal arch spanned a road where itapproached a town, and provided the only architectural vista to beseen in most of these Roman towns. 

 [60] In western Europe the provincial Roman amphitheatre averaged 45 x 70 yds. For its arena. 

Dimensions, of course, varied. There was no normal size for an infanttown. Some, when first established, covered little more than 30 acres, the area of mediaeval Warwick. Others were four or five times asspacious; they were twice or nearly twice as large as mediaevalOxford, no mean city in thirteenth-century England. Most of them, doubtless, grew beyond their first limits; a few spread as far as asquare mile, twice the extent of mediaeval London. Similarly the'insulae' varied from town to town. In one, Timgad, they were only 70to 80 ft. Square. Often they measured 75 to 80 yds. Square, rathermore than an acre, as at Florence, Turin, Pavia, Piacenza. [61]Occasionally they were larger, but they seldom exceeded three acres, and their average fell below the prevalent practice of modernchess-board planning. 

 [61] For Florence and Turin see below; for Piacenza, the plans on the scale of 1:1000 and 1:5000 in L. Buroni's _Acque potabili di Piacenza_ (1895). 

In most towns, though not in all, the dimensions of the 'insulae' showa common element. In length or in breadth or in both, they usuallyapproximate to 120 ft. Or some multiple of that. The figure issignificant. The unit of Roman land-surveying, the 'iugerum', was arectangular space of 120 by 240 Roman feet--in English feet a tinytrifle less--and it seems to follow that 'insulae' were often laid outwith definite reference to the 'iugerum'. The divisions may not havealways been mathematically correct; our available plans are seldomgood enough to let us judge of that, [62] and we do not know whether weought to count the surface of the streets with the measurement of the'insulae'. But the general practice seems clear, and it extended evento Britain (p. 129), and though blocks forming exactly a 'iugerum' ora half 'iugerum' are rare, the Italian land-measure certainly affectedthe civilization of the provincial towns. 

 [62] Silchester and Timgad are the only two sites which have been planned well enough to provide accurate measurements. The large modern town-plans (e. G. Of Turin, p. 86) are useful, but inadequate to our purpose; for one thing, they often exaggerate the width of the streets. One really needs actual measurements made on the spot. 

In this system perhaps the most peculiar feature is the intermixtureof square and oblong 'insulae'. It is not merely the variation whichcan be traced in Priene (fig. 5), where some blocks are rather moresquare or oblong than others, but where all approach the same norm. The Roman towns which we are now considering show two varieties ofhouse-blocks. Sometimes the blocks are square; sometimes, perhaps moreoften, they are oblong approximating to a square, like the blocks ofPriene. But in a few cases, as at Naples among the more ancient, andat Carthage among the later foundations, they are oblong and theoblongs are very long and narrow. 

It is hard to detect any principle underlying the use of these variousforms. No doubt differences of historical origin are ultimately thecauses of the mixture. But our present knowledge does not reveal theseorigins. The evidence is, indeed, contradictory at every point. If theGraeco-Macedonian fashion be quoted as precedent for square orsquarish 'insulae', the Terremare show the same. If the theoreticalscheme of the earlier Roman camp seemed based on the long narrowoblong, the actual remains of legionary encampments of the secondcentury B. C. At Numantia include many squares. If one part of Pompeiiexhibits oblongs, another part is made up of squares. If Piacenza, first founded in north Italy about 183 B. C. , and founded again ahundred and fifty years later, is laid out in squares, its coevalneighbour Modena prefers the oblong. If the old Greek city of Naplesembodies an extreme type of oblong, so does the later AugustanCarthage (pp. 100, 113). In the historic period, it would seem, nosharp line was drawn, or felt to exist, between the various types of'insulae'. In the main, the square or squarish-oblong was preferred. Local accidents, such as the convenience of the site at Carthage, ledto occasional adoption of the narrower oblong. 

The Roman land-surveyors, it is true, distinguished the square and theoblong in a very definite way. The square, they alleged, was proper tothe Italian land or to such provincial soil as enjoyed the privilegeof being taxed--or freed from taxation--on the Italian scale. Theoblong they connected with the ordinary tax-paying soil of theprovinces. This distinction, however, was not carried out even in theagrarian surveys with which these writers were especiallyconcerned, [63] and it applies still less to the towns. No doubt it isa fiction of the office. It would be only human nature if thesurveyors, finding both forms in use, should invent a theory toaccount for them. 

 [63] Schulten, _Bonner Jahrbücher_, ciii. 23, and references given there. 

The system sketched in the preceding paragraphs seems, as has beensaid (p. 73), to have sprung from a fusion of Greek or Graeco-Macedonianwith Italian customs. Roman town-planning, like Roman art, was recastunder Hellenistic influence and thus gained mathematical precision andsymmetry. When this happened is doubtful. Foreign scholars oftenascribe it to Augustus and find a special connexion between the firstemperor and the chess-board town-plan. But the architect Vitruvius, who dedicated his book to Augustus and who gives some brief notice totown-planning, urges strongly that towns should not be laid out on thechess-board pattern, but rather on an eight-sided or (as we might callit) star-shaped plan. [64] He would hardly have denounced a schemewhich had been specially taken up by his patron, nor indeed does hiscriticism of the chess-board system sound as if he were denouncing anovelty in Italian building. 

 [64] i. 5 (21), 6 (28, 29). 

On the other hand there seems no great difficulty in the idea thatthe regularization of the old Italian town-plan by Greek influencetook place spontaneously in the late Republic. We cannot, indeed, date the change. It must remain doubtful whether it came by degreesor all at once, [65] and whether the right-angled plans of towns likeAquileia[66] or Piacenza belonged to their first foundation, i. E. Toabout 180 B. C. , or to later rearrangements. But it seems reasonableto believe that a Graeco-Italian rectangular fashion of town-planningdid supersede an earlier, irregular, Italian style, and had becomesupreme before the end of the Republic. 

 [65] Perhaps about 180 B. C. , Mommsen, _Roman Hist. _ iii. 206. 

 [66] Aquileia was set up in 181 B. C. To guard the north-east gate of Italy, and was reinforced in 169. Its remains, so far as excavated, show a rectangular plan of oblong 'insulae'--some of 1-1/2 acres (74 by 94 yards), some larger--while, till its downfall, about A. D. 450, we hear no word of refoundation or wholesale rebuilding. But if its original area be the space of 70 acres which is usually assigned, that is not rectangular but a square somewhat askew, which fits very badly with the rectangular street-plan, and one would incline to ascribe the latter to a later date. See Maionica, _Fundkarte von Aquileia_. 

CHAPTER VII

INSTANCES OF ITALIAN TOWN-PLANS

The preceding chapters have dealt with the origins and generalcharacter of the Italian town-plan. We pass now to the remains whichit has left in its own home, in Italy. These are many. In one cityindeed, the greatest of all, no town-planning can be detected. LikeAthens and Sparta, Rome shows that conservatism which marks so manycapital cities. No part of it, so far as we know, was laid out on arectangular or indeed on any plan. [67] It grew as it could. Itsbuilders, above all its imperial builders, cared much for spectaculareffects and architectural pomp. Even in late Republican times thegloomy mass of the Tabularium and the temples of the Capitol must havetowered above the Forum in no mere accidental stateliness, andimperial Rome contained many buildings in many quarters to show thatit was the capital of an Empire. But for town-planning we must goelsewhere. 

 [67] The traces of prehistoric planning detected by some writers in Rome are very dubious. 

The sources of our knowledge are twofold. In a few casesarchaeological excavation has laid bare the paving of Roman streets orthe foundation of Roman house-blocks. More often mediaeval and modernstreets seem to follow ancient lines and the ancient town-plan, or apart of it, survives in use to-day. Such survivals are especiallycommon in the north of Italy. It is not, indeed, possible to gather afull list of them. He who would do that needs a longer series of goodtown-maps and good local histories than exist at present; he needs, too, a wider knowledge of mediaeval Italian history and a closerpersonal acquaintance with modern Italian towns, than a classicalscholar can attempt. But much can be learnt even from our limitedmaterial. [68]

 [68] See the seventeenth century Atlases of Blaeu, Janssons, and others, the modern maps prepared by Grassellini and others about 1840-50 (some on the scale 1:4, 000), and in particular the _Atlante geografico_ of Attilio Zuccagni-Orlandini (Firenze, 1844), and the recent town-maps of various Italian cities (mostly about 1:10, 500). Different maps of the same town sometimes differ much in their detail. The Italian Government maps of the largest scale (1:25, 000) are small for our present purpose and have been issued mainly for northern Italy. 

The evidence of the streets needs, however, to be checked in everycase. It would be rash to assume a Roman origin for an Italian townsimply because its streets are old and their plan rectangular. Thereare many rectangular towns of mediaeval or modern origin. Such isTerra Nova, near the ancient Gela in Sicily, built by Frederick StuporMundi early in the thirteenth century. Such, too, Livorno, built bythe Medici in the sixteenth century. Such, too, the many littlemilitary colonies of the Italian Republics, dotted over parts ofnorthern and middle Italy. Often it is easy to prove that, despitetheir chess-board plans, these towns do not stand on Roman sites. Often the inquiry leads into regions remote from the study of ancienthistory. 

Fortunately, enough examples can be identified as Roman to serve ourpurpose. Some of these occur in the Lombardy plain where, both underthe Republic and at the outset of the Empire, many 'coloniae' wereplanted full-grown and where town-life on the Roman model wasotherwise developed. Not all these towns survive to-day; not all ofthe survivors retain clear traces of their Roman town-plan; in ninecases, at least, the streets seem unmistakably to follow Roman lines. Four of the nine date from early days; in the late third and the earlysecond centuries (218-183 B. C. ), Piacenza, Bologna, Parma, and Modena, were built as new towns with the rank of 'colonia'. The first three ofthese were later refounded, about 40-20 B. C. --whether their streetswere then laid out afresh is an open question--and Turin and Bresciawere added. In addition, Verona, Pavia, and Como won municipal statusin or before this later date, though when or how they came to be laidout symmetrically is not certain. [69] And there are other less certainexamples. 

 [69] Milan (Mediolanium), once the chief Roman town of north Italy, is usually stated to preserve to-day no trace of Roman street-planning. But the line of the Via Manzoni, Via Margherita, and Via Nerino (cutting the Ambrosian Library) seems really to represent one of its main streets, and the line of the Fulcorino and Corso di Porta Romana the other, while one or two traces of 'insulae' can be detected near the Ambrosian Library. The town was destroyed in A. D. 539 and again in 1162, and more survivals cannot be expected. 

Other instances, but not so many, may be quoted from south of theApennines. At Florence, for example, and at Lucca 'coloniae' wereplanted full-grown and the street-plans still record the fact. AtNaples, at Herculaneum, perhaps at Sorrento, [70] proofs survive ofsimilar planning. But the towns of central Italy were in great partmore ancient than the era of precise town-planning, and many of themwere perched in true Italian fashion on lofty crags--_praeruptisoppida saxis_--which gave no room for square or oblong house-blocks. In the period of the dying Republic and nascent Empire fewer'coloniae' were planted here than in the north, while in much ofsouthern Italy towns have in all ages been comparatively rare. 

 [70] Beloch, _Campanien_, p. 252. 

In the towns just noted we can trace many, though not all, of theoriginal house-blocks. Usually the blocks are square or nearly so, asat Turin, Verona, Pavia, Piacenza, Florence, Lucca. Less often theyare long and even narrow rectangles, as at Modena, and Sorrento, andabove all Naples, and as usual it is not easy to understand the reasonfor the difference (p. 80). 

_Turin_ (fig. 15). 

Of all the examples of Roman town-planning known to us in Italy, Turinis by far the most famous. [71] Here the streets have survived almostintact, and excavations have confirmed the truth of the survival byrevealing both the ancient road-metalling and the ancient town-wallsand gates. Turin, Augusta Taurinorum, began about 28 B. C. As a'colonia' planted by Augustus. Its walls enclosed an oblong of about745 x 695 metres (127 acres). [72] The sides are represented (1) on thenorth by the Via Giulio, in the western part of which the southernedge of the street actually coincides with the line of the Romantown-wall, while further east the Porta Palatina enshrines an ancientgate; (2) on the west by the Via della Consolata, and the ViaSiccardi, the east side of which latter street seems to stand upon theRoman town-wall; and (3) on the south by the Via della Cernaia and ViaTeresa, the north side of which stands over the Roman southerntown-wall. (4) The east wall agrees with no existing street but may berepresented by a line drawn through the Carignano Theatre and thewestern front of the Palazzo Madama, which contains the actual towersof the Roman east gate. [73] The north-west corner, uncovered in 1884, is a sharp right angle. This feature recurs at Aosta and at Laibach(pp. 90, 116), both founded, like Turin, in the Augustan age, andseems to belong to that period; later, it gave place to the roundedangle visible at Timgad (p. 109) and in many Roman forts of the middleEmpire. 

 [71] Carlo Promis, _Storia dell' antico Torino_ (Torino, 1869); Alfredo d'Andrade, _Relazione dell' ufficio regionale per la conservazione dei monumenti del Piemonte_, 1883-91 (Torino, 1899); Schultze, _Bonner Jahrbücher_, cxviii. 339; Barthel, _ibid_. Cxx. 105; Pianta di Torino (1-10, 000), by G. B. Paravia. 

 [72] I take these figures from the plan of Paravia, which is said to be the most correct plan of Turin at present available. Promis gives smaller dimensions, 720 x 670 m. , and he measured from what is now known to be a point too far to the east (the Via Accademia delle Scienze) instead of from the west front of the Palazzo Madama; he has, however, been usually followed. Other maps give other dimensions, Orlandini (1844), 758 x 780 m. ; Vallardi (1869), 680 x 740 m. ; Maggi (1876), 730 x 800 m. ; Ashby (Art. 'Turin' in _Encycl. Britannica_) gives 2, 526 x 2, 330 ft. Which must be too large. I reproduce here (fig. 15) the plan of Orlandini, since it shows well the extent of street-survivals in Turin before the great modern rebuildings or expansions. 

 [73] d'Andrade, _Relazione_, pp. 8-20; _Notizie degli Scavi_, 1885, pp. 173, 271, and 1902, p. 277. 

Of the interior buildings of the town little is known. The Forumperhaps stood near the present Palazzo di Città, and the Theatre wastraced in 1899 in the north-east corner of the town, occupyingapparently, a complete insula;[74] of the private houses nothingdefinite seems to be recorded. 

 [74] _Notizie_, 1903, p. 3. 

But the street-plan has survived intact, except in two outlyingcorners. The town was divided up into square or nearly square blocks, of which there were nine counting from east to west and eight fromnorth to south. Most of these 'insulae' measured about 80 yds. Square. [75] A few were larger, 80 x 120 yds. ; these were ranged alongthe north side of the street now called Via Garibaldi (formerly DoraGrossa), which represents the Roman main street between the east andwest gates--in the language of the Roman land-surveyors, the_decumanus maximus_. This street cut the town into two equal halves. The other divisions of the town were no less symmetrical. But, asthere were nine 'insulae' from east to west, the main north and southstreet could not bisect the town. Indeed, the south gate seems to havehad five house-blocks west of it and four east of it, while the PortaPalatina stands further west, with six blocks on the west side of it. The north and south gates, therefore, are not opposite. [76] Whetherthis was the original plan is not clear, nor is the age of thesurviving walls and gates quite certain; the bonding courses in someof the masonry of the walls does not seem Augustan. But the streetplan may unhesitatingly be assigned to the first establishment of thetown, about 28 B. C. Since, it has been extended far beyond the Romanwalls. Nearly all modern Turin has been laid out, bit by bit, inimitation and continuation of the original Roman lines. 

 [75] An insula is mentioned in _Notizie_, 1901, p. 391, which measured 74 x 80 metres. It is likely that there were small unevennesses in the ancient as there are in the modern house-blocks. The 'insulae' which abutted on the town-walls are represented to-day by unduly large blocks, oblong rather than square, but these latter contain not only the areas of the Roman 'insulae' in question, but also the space between them and the town-walls and the lines of the wall themselves (p. 77). 

 [76] This failure in symmetry recurs in one or two other Roman towns as probably at Timgad (p. 109) and at Cologne (E. And W. Gates), at Silchester and Caerwent, but it may sometimes be the result of alteration. Occasionally it appears in military sites (Ritterling, _Lager bei Hofheim_, p. 29 _note_). It is presumably a mere matter of convenience; no superstition attaches to it such as that which led the Chinese not to put their gates opposite each other (p. 148). 

[Illustration: FIG. 15. TURIN. FROM A PLAN OF 1844]

_Aosta_ (fig. 16). 

Another example of an Italian town-plan, from the same date anddistrict as Turin, is supplied by Augusta Praetoria, now Aosta, somefifty miles north of Turin in the Dora Baltea Valley, not far from thefoot of Mont Blanc. [77] Aosta was founded by Augustus in 25 B. C. On ahitherto empty spot, to provide homes for time-expired soldiers and toserve as a quasi-fortress in an important Alpine valley. Its firstinhabitants were 3, 000 men discharged from the Praetorian Guard, withtheir wives and children; its population may have numbered at theoutset some 15, 000 free persons, besides slaves. The town, as it isknown to us from excavation and observation, formed a rectangle 620yds. Long and 780 yds. Wide, and covered an area of about 100 acres(fig. 16). The walls formed sharp right angles at the corners, as atTurin. Within the walls were an amphitheatre, a theatre, public baths, a structure covering nearly 2 acres and interpreted as a granary or(perhaps more correctly) as a cistern, [78] and private houses as yetunexplored. Beneath the chief streets were sewers, by which indeedthese streets were mainly traced. 

 [77] C. Promis, _Antichità di Aosta_ (Torino, 1862), with plan, plate 3, dating from 1838; _Notizie degli Scavi_, 1899, p. 108, with a later plan, but lacking a scale; Nissen, _Ital. Landeskunde_, ii. 171. 

 [78] Durm _Baukunst der Römer_, p. 458. 

[Illustration: FIG. 16. AOSTA]

The whole was divided by a regular network of streets into rectangularblocks. According to the latest plan of the site, there were sixteenblocks, nearly identical in shape and averaging 145 x 180 yds. (5-1/2acres). That, however, is an incredible area for single house-blocks, and it is to be noted that Promis shows two further roads (A, A infig. 16). If these are survivals of other such roads, Aosta may havecontained thirty-two oblong 'insulae', each nearly 220 x 540 ft. , oreven sixty-four smaller and squarer 'insulae', measuring half thatsize. [79] Four gates gave entrance; those in the two longer sideswhich face north-west and south-east, are curiously far from thecentre and indeed close to the south-western end of the town. It is, of course, impossible to determine, without spade-work, which of therecognizable buildings of Aosta date from the foundation of the placein 25 B. C. But the general internal scheme and the symmetrical andpractically 'chess-board' pattern of streets must date from the firstfoundation. [80]

 [79] Promis, p. 140; his plan has no proper scale. There seems no decisive evidence and the modern streets of Aosta do not help us. 

 [80] The town of Concordia in north-east Italy, where Augustus planted a 'colonia', doubtless of discharged soldiers, is said to have possessed a ground-plan of oblong blocks very like that of Augusta Praetoria. But this plan rests mainly on the authority of a workman who apparently did not know how to read or write (he is described as 'analfabeta') and I therefore omit it here. See _Notizie degli Scavi_, 1880, p. 412, and Plate XII (the text gives no dimensions and the plan lacks a scale), and compare 1882, p. 426, and 1894, p. 399. 

_Florence_ (fig. 17). 

A yet more interesting instance of a Roman town-plan preserved in manystreets may be found in Florence. [81] In Roman times Florence was a'colonia'. When this 'colonia' was planted is very doubtful. Perhapsthe age of Sulla (90-80 B. C. ) is the likeliest date; all that isactually certain is that the foundation was made before the end of thefirst century A. D. This 'colonia', like others, was laid out inchess-board fashion, and vestiges of its streets survive in the Centrowhich forms the heart of the present town. The Centro of Florence, aswe see it to-day, is very modern. It was, indeed, laid out ageneration ago by Italian architects who designed the broad streetscrossing at right angles which form its characteristic. But this'Haussmannization' revived, consciously or unconsciously, an oldarrangement. The plan of Florence in 1427 shows a group of twentyunmistakable 'insulae', each of them about 1-1/8 acre in area, thatis, very similar in size to the 'insulae' of Turin. This group isbounded by the modern streets Tornabuoni on the west, Porta Rossa onthe south, Calzaioli on the east, Teatina on the north; it covers arectangle of some 305 x 327 yds. , not quite 21 acres. 

 [81] On Roman and early mediaeval Florence see Villani, _Cronica_ (written about 1345, published 1845), i. 61, 89, 120; R. Davidsohn, _Geschichte von Florenz_ and _Forschungen_ (Berlin, 1886); L. A. Milani, _Notizie degli Scavi_, 1887, p. 129; plan of the Centro in 1427 by Comm. Guido Carocci, _Studi storici sul Centro di Firenze_ (Florence, 1889); _Monumenti antichi_, vi. 15. Nissen _(Ital. Landeskunde_, ii. 296) fixes its area at 400 x 600 m. , about 58 acres. 

[Illustration: FIG. 17A. FLORENCE, SINCE THE REBUILDING OF THECENTRAL PORTION (Centro shaded). ]

[Illustration: FIG. 17B. FLORENCE ABOUT 1795, FROM L. BARDI. The chief streets which seem to have preserved Roman lines are markedin black. ]

The original Roman town presumably extended beyond these narrowlimits. But it is not easy to fix its area, nor are unmistakable'insulae' to be detected outside them. On the west the Via Tornabuoniseems to have marked the Roman limit, as it does to-day. On the north, a probable line is given by the gateway, Por Episcopi, which oncespanned the passage--now an open space--on the east side of theArchbishop's Palace (plan 17 B). That gateway stood between the ViaTeatina and the next street to the north, the Via dei Cerretani, andthe Roman north wall and ditch apparently ran along the intervalsbetween these two modern streets--as indeed the lines of certainmediaeval lanes suggest. On the east the 'colonia' is supposed to havestretched to the Via del Proconsolo and the old Por S. Piero, probablythe original east gate. Here the traces of 'insulae' are illpreserved; the space in question would contain, and the mediaevalstreets would admit of, twelve blocks in addition to the twenty notedabove. 

The southern limit of Roman Florence towards the Arno is altogetherdoubtful. There are, or were, traces of Roman baths in the Via delleTerme, and it has been thought that the town stretched riverwards asfar as the old gate Por S. Maria and the Piazza S. Trinità. The gate, however, is ill-placed and the line of wall implied by this theory isirregular. The mediaeval streets point rather to a south wall near theVia Porta Rossa. The baths might perhaps be due to a later Romanextension, such as we shall meet at Timgad (p. 113). The Por S. Mariamay even be due to one of the reconstructions of Florence in theMiddle Ages. At the end we must admit that without further evidencethe limits of Roman Florence cannot be fixed for certain. But thelimits indicated above give the not unsuitable dimensions of 46 acres(380 x 590 yds. ), while the history of the twenty indubitable insulaeof the Centro remains full of interest. We see here, as clearly asanywhere in the Roman world, how the regular Roman plan has graduallybeen distorted by encroachments and how, even in its irregularity, ithas had power to drive modern builders towards its ancient fashion. 

Of the interior of the Roman town little is known. The streets nowcalled Strozzi and Speziali plainly preserve the Roman main streetfrom east to west, while the Via Calimara overlies that which ran fromnorth to south. Where these crossed was the mediaeval Mercato Vecchio, now enlarged into a patriotic Piazza Vittorio Emmanuele; here we mayput the Roman forum, and here too, by the former church of S. Maria inCampidoglio, was the temple of Capitoline Juppiter. There were alsotheatres, a shrine of Isis, and, outside the Roman limit, anamphitheatre still discernible in the curves of certain streets (plan17 B). However small Florentia was, it possessed the true elements ofthe Roman town. 

_Lucca_ (fig. 18). 

A good parallel to Florence may be found at Lucca, the ancient Luca, where again the streets preserve a rectangular pattern without showingclearly what was its full extent. Luca is said to have been founded asa 'colonia' in 177 B. C. , but the statement is of doubtful truth. Certainly it was a 'municipium' in Cicero's days, and a little later, in the period 40-20 B. C. , it received the rank of 'colonia' and manycolonists, taken (as an inscription says) from discharged soldiers ofLegions VII and XXVI. Whether the surviving traces of town-planningdate from this latter event or from some earlier age is not easy tosay. But of the street-plan there can be no doubt, though its originalsize is uncertain. A rectangular area about 700 yds. From east to westand 360 yds. From north to south is divided into fifteen square orsquarish 'insulae' arranged in three rows. Each insula is about 3acres, but those of the middle row are larger than the rest (150 x 150yds. ). The Via S. Croce which runs along the south side of this rowwas perhaps the main east and west thoroughfare of the town, the'decumanus maximus', so that the larger 'insulae' correspond to thosewhich appear in the same position at Turin and elsewhere (p. 88). 

[Illustration: FIG. 18. LUCCA. (The streets which preserve Roman lines are marked in black. )]

Whether there were other 'insulae' besides the fifteen is doubtful. Onthe east there were certainly none: the two narrow parallel streets atthe east end of the area just described are obviously due to a growthof houses along the line of the original east wall. The other limitsare more obscure. Probably the north and west walls stood a littleoutside of the Via Galli Tassi (once S. Pellegrino) and the Via S. Giorgio, but there may well have been a row of insulae, nowobliterated, south of the Via del Battistero. One or two interiorbuildings are known. The Forum appears to have stood where is now thePiazza S. Michele in Foro; close by was a temple; in the north-easternquarter, at the Piazza del Carmine, was probably the theatre; near itbut outside the walls was the amphitheatre, its outlines still visiblein the Piazza del Mercato (110 x 80 yds. In greatest dimensions). [82]

 [82] Plan by P. Sinibaldi, 1843, 1:4, 000. _Notizie degli Scavi_, 1906, p. 117, &c. Nissen (_Ital. Landeskunde_, ii. 288) gives the area as 800 x 1, 200 metres, which seems much too large. 

_Herculaneum_ (fig. 19). 

To these examples from north Italy may be added two from the south, Herculaneum and Naples. Herculaneum had much the same early history asits more important neighbour Pompeii. First an Oscan settlement, thenEtruscan, then Samnite, it passed later under Roman rule. After theSocial Wars (89 B. C. ) it appears as a 'municipium'; of its historyfrom that date till its destruction (A. D. 79) we know next to nothing. But excavations, commenced in the eighteenth century and now longsuspended, have thrown light on its ground-plan. [83] This was arectangular pattern of oblong house-blocks, measuring 54 x 89 yds. , orin some cases a little more, and divided by streets varying from 15 to30 ft. In width which ran at right angles or parallel to one another. Only a part of the town has been as yet unearthed. In that a broadcolonnaded main street ran from north-west to south-east; on thenorth-east side of this street stood a row of house-blocks with astructure taken to be a Basilica, and on the south-west of it were tenhouse-blocks, one of which includes some public baths. At the northend of this area are a theatre and temple, at the south end two largestructures which have been called temples but are more like largeprivate houses; on the east (according to the eighteenth-centurysearchers) are graves. 

 [83] M. Ruggiero, _Scavi di Ercolano_ (Naples, 1885), plates ii and xii; Beloch. _Campanien_, pp. 215 foll. ; Nissen, _Ital. Landeskunde_, ii. 759; Waldstein and Shoobridge, _Herculaneum_ (London, 1908), pp. 60 foil. ; E. R. Barker, _Buried Herculaneum_ (1908); Gall in Pauly-Wissowa, viii. (1912) 532-48. 

[Illustration: FIG. 19. HERCULANEUM]

How much of the town has been uncovered, how much still lies hiddenbeneath the lava which overflowed it in A. D. 79, is disputed. Of itstown-walls and gates no trace has yet been found. But nearly all itspublic buildings seem to be known; the graves on the east side, ifcorrectly mapped by their discoverers and if coeval with the streetsand houses, leave no room for further 'insulae' in that direction, while the great country-house called the 'Casa dei Papiri' plainlystood outside the town on the north-west. From these facts one modernwriter has calculated that Herculaneum was less than a quarter of amile long, less than 350 yds. Broad, and less than 26 acres inextent--in short, not a sixth part of Pompeii. These measures areprobably too small. The 'Basilica' on the north side of the mainstreet cannot have stood on the extreme edge of the town. There musthave been not three but four rows of house-blocks from south-west tonorth-east; the graves once noted in this quarter must be older thanour Herculaneum or otherwise unconnected with it. The whole town musthave been 40 or 45 rather than 25 acres in area. Even so it is alittle town. The unenthusiastic references to it in ancient literatureare, after all, truthful. Apart from the great villa outsideit--possibly an imperial residence--it hardly deserved, or to-daydeserves, to be excavated at the extraordinary cost which itsexcavation would involve. 

The date of its planning is as doubtful as the extent of its area. Onerecent writer, Nissen, has suggested that it was reconstructed afteran earthquake in A. D. 63 and was hardly completed before the eruptionof 79. The earthquake is well attested. But it cannot possibly havewrecked the town so utterly as to cause wholesale rebuilding on newlines, and an inscription points rather to the time of Augustus. OneMarcus Nonius Balbus (the text runs) built 'a basilica, gates and awall at his own cost', and this builder Balbus was probably acontemporary of Augustus. [84] Others have preferred to think that thetown-planning reveals Greek influences; they point to the Greek cityof Naples, 7 miles west of Herculaneum, and the Doric temple atPompeii, much the same distance east of it. However, neither thetown-planning of Naples, to be discussed in the next paragraphs, northat of Pompeii (p. 68), seems to be necessarily Greek, andHerculaneum itself contains nothing which cannot be explained asItalian. It is possible, though there is no record of the fact, thatit received a settlement of discharged soldiers somewhere about 30B. C. And was then laid out afresh. But here, as throughout thisinquiry, more light is needed if the inquirer is to pass fromguesswork to proven fact. 

 [84] _CIL_. X. 1425; compare Dessau, 896. It is, no doubt, possible that this Nonius Balbus is the M. Nonius . .. Who built something in honour of Titus in A. D. 72, but the identification is not likely. 

_Naples_ (fig. 20). 

One more example, from the neighbourhood of Herculaneum, may completethe list of Italian street-plans. Naples, the Greek and RomanNeapolis, was a Greek city, the most prosperous of the Greek towns inCampania. [85] After 90 B. C. It appears to have become a Roman'municipium'. But it retained much of its Greek civilization. A writerof the early first century after Christ, Strabo, states that abundanttraces of Greek life survived there, 'gymnasia, and athletic schools, and tribal divisions, and Greek names even for Roman things. ' Evenlater Tacitus calls it a 'Greek city', and Greek was still used forofficial inscriptions there in the third century. 

 [85] Beloch, _Campanien_ (Berlin, 1879), p. 26; Capasso, _Napoli Greco-Romana_ (Napoli, 1905). The Forum, Market, and some other buildings marked by Capasso seem to me (and even to him or his editors) very dubious (p. 63). Two theatres (p. 82) and a Temple of the Dioscuri are better established. For plans see _Piante topogr. Dei quartieri di Napoli_ 1861-5 (1:3, 888) and _Pianta della città di N. _ (Off. Della Guerra, 1865), from which latter fig. 20 is adapted. 

[Illustration: FIG. 20. NAPLES. ADAPTED FROM A PLAN OF 1865. (TH = Theatre, T = Temple. )]

This Neapolis town had, as certain existing streets declare, apeculiar form of town-planning. The area covered by these streets isan irregular space of 250 acres in the heart of the modern city, about850 yds. From north to south and 1, 000 yds. From east to west. [86] InRoman days three straight streets ran parallel from east to west and alarge number of smaller streets, twenty or so, ran at right angles tothem from north to south. The house-blocks enclosed by these streetswere all of similar size and shape, a thin oblong of 35 x 180 metres(39 x 198 yds. ). Some of the public buildings naturally trespassed onto more than one 'insula'; a theatre appears indeed to have stretchedover parts of three. In general, the oblongs seem to have been laidout with great regularity and the angles are right angles, though the'insulae' in the northern and southern rows of house-blocks cannothave been fully rectangular and symmetrical. 

 [86] The limits are the Castel Capuano on the east, the Strada dell' Orticello on the north, the church of S. Pietro a Majella on the west, and on the south the churches of S. Marcellino and S. Severino. 

This town-plan of Naples differs from any of those noted above. Itsblocks are narrower than those in any Italian town, unless in Modena, and while they resemble the 'insulae' of the sixth region of Pompeii(fig. 13), are far more regular than those. Almost the only closeparallel is that of Roman Carthage (fig. 24). As Naples was by originand character a Greek city, these narrow oblongs have been supposed torepresent a Greek arrangement. They do not, however, correspond toanything that is known in the Greek lands, either of the Macedonian orof any earlier period. The conclusion is difficult to avoid that thisGreek city of Naples adopted an Italian street-scheme, but laid it outwith more scientific regularity than the early Italians themselves. When this occurred and why, is wholly unknown. That the result is notan unpractical form of building is shown by the fact that similar longand narrow house-blocks are a characteristic feature of modernLiverpool, though they seldom occur in other English towns, unlessintermixed with square and other blocks. 

CHAPTER VIII

ROMAN PROVINCIAL TOWN-PLANS. I

The provinces, and above all the western provinces of the RomanEmpire, tell us even more than Italy about Roman town-planning. Butthey tell it in another way. They contain many towns which werefounded full-grown, or re-founded and at the same time rebuilt, andwhich were in either case laid out on the Roman plan. But the modernsuccessors of these towns have rarely kept the network of theirancient streets in recognizable detail. Though walls, gates, temples, baths, palaces, amphitheatres still stand stubbornly erect amidst aflood of modern dwellings, they are but the islands which mark asubmerged area. The paths and passages by which men once moved acrossthat area have vanished beneath the waves and cannot be recovered fromany survey of these visible fragments. There is hardly one modern townin all the European and African provinces of the Roman Empire whichstill uses any considerable part of its ancient street-plan. In ourown country there is no single case. In Gaul and Germany, two or threestreets in Cologne and one or two in Trier are the sole survivals. [87]In Illyricum there is no example unless possibly at Belgrade. In theSpanish peninsula the town of Braga in northern Portugal seems tostand alone. In Roman Africa--Tunis, Algiers and Morocco--no instancehas survived the Arab conquest. [88]

 [87] For Orange see p. 107. Nîmes may possibly retain one or two streets of the Roman Nemausus, but it is very doubtful; see Menard's map of 1752. See further in general p. 142. 

 [88] Though, curiously enough, the chess-board pattern of field divisions has survived in the neighbourhood of Carthage. 

If, however, survivals of ancient streets are as rare in the provincesas they are common in Italy, the provinces yield other evidenceunknown to Italy. In these lands, and above all in Africa, the sitesof many Roman towns have lain desolate and untouched since Roman days, waiting for the excavator to recover the unspoilt pattern of theirstreets. If the Roman Empire brought to certain provinces, as itunquestionably did to Africa, the happiest period in their historytill almost the present day, that only makes their remains the morenoteworthy and instructive. Here the new art of excavation has alreadyachieved many and varied successes. In the western Empire one town, Silchester in Britain, has been wholly uncovered within the circuit ofits walls. Others, like Caerwent in Britain or Timgad and Carthage inAfrica, have been methodically examined, though the inquiries have notyet touched or perhaps can never touch their whole areas. In othersagain, some of which lie in the east, occasional search or even chancediscoveries have shed welcome light. Our knowledge is more than enoughalready for the purposes of this chapter. 

We can already see that the town-plan described in the foregoing pageswas widely used in the provinces of the Empire. We find it in Africa, in Central and Western Europe, and indeed wherever Rorrran remainshave been carefully excavated; we find it even in remote Britainamidst conditions which make its use seem premature. Where excavationhas as yet yielded no proofs, other evidence fills the gap. Insouthern Gaul, as it happens, archaeological remains are unhelpful. But just there an inscription has come to light, the only one of itskind in the Roman world, which proves that one at least of the'coloniae' of Gallia Narbonensis was laid out in rectangular oblongplots. It is clear enough that this town-plan was one of the formsthrough which the Italian civilization diffused itself over thewestern provinces. 

The exact measure of its popularity is, however, hard to determine. Inthe east it found little entrance. There, the very similar Macedonianand Greek methods of town-planning were rooted firmly, long beforeRome conquered Greece or Asia Minor or Syria or Egypt. The fewtown-plans which have been noted in these lands, and which may beassigned more or less conjecturally to the Roman era, seem to beHellenic or Hellenistic rather than Italian. They show broad statelystreets, colonnades, vistas, which belong to the east and not toItaly. Even in the west, the rule of the chess-board was sometimesbroken. Aquincum, near Budapest, became a 'municipium' under Hadrian;its ruins, so far as hitherto planned, exhibit no true street-planning. But that may be due to its history, for it seems not to have beenfounded full-grown, but to have slowly developed as best it could, and to have won municipal status at the end. 

Roman Africa is here, as so often, our best source of knowledge. AtTimgad (p. 109), a town laid out in Roman fashion with a rigid'chess-board' of streets was subsequently enlarged on irregular andalmost chaotic lines. At Gigthi, in the south-east of Tunis, thestreets around the Forum, itself rectangular enough, do not runparallel or at right angles to it or to one another. [89] At Thibilis, on the border of Tunis and Algeria, the streets, so far as they haveyet been uncovered, diverge widely from the chess-board pattern. [90]One French archaeologist has even declared that most of the towns inRoman Africa lacked this pattern. [91] Our evidence is perhaps stilltoo slight to prove or disprove that conclusion. Few African townshave been sufficiently uncovered to show the street-plan. [92] Buttown-life was well developed in Roman Africa. It is hardly crediblethat the Africans learnt all the rest of Roman city civilization andcity government, and left out the planning. The individual cases ofsuch planning which will be quoted in the following pages tell theirown tale--that, while the strict rule was often broken, it was therule. 

 [89] _Archives nouvelles des Missions scientifiques_, xv. 1907, fasc. 4. 

 [90] Plan by Joly, _Arch. Anzeiger_, 1911, p. 270, fig. 17. The plan has been thought to imply 'insulae' twice as large as those of Timgad. To me it suggests nothing so regular. 

 [91] Toutain, _Cités romaines de la Tunisie_, p. 79 note: 'Ce qui toutefois est incontestable, c'est que cette disposition d'une régularité artificielle, autour de deux grandes voies exactement orientées et se coupant a angle droit, est très rare dans l'Afrique romaine. Les villes de ce pays n'out pas été toutes construites sur le mème plan: chacune d'elles a, pour ainsi dire, épousé la forme de son emplacement. '

 [92] There are many in which it could be traced with some ease, apparently. Thelepte, Cillium, Ammaedara, Sufetula, _Archives des Missions_, 1887, pp. 68, 121, 161-171, Simitthu, _Mémoires présentés par divers savants_, ser. I. X. 462, and Thuccabor, Tissot, _Géogr. D'Afrique_, ii. 292, seem to have visible streets, but no one has recorded them exactly. The plan of Utica, given by Tissot (_Atlas_, by Reinach, plate vi) on the authority of Daux, is open to doubt. 

[Illustration: FIG. 21. INSCRIPTION OF ORANGE. (From the _Comptes-rendus de l'Académie des Inscriptions_. )

Plot (_meris_) I (_lost_) . .. 

Plot II . .. Perpetual lessee (_manceps_) C. Naevius Rusticus: suretyfor him C. Vesidius Quadratus. Fronting the Kardo. 

 (5) Plot III, frontage of 34-1/2 feet and Plot IV, frontage of 35 feet;ground rent (?), 69-1/2 denarii (_in margin_). Yearly rent II . .. (?). Lessee and surety, as above. Fronting the Kardo. 

(10) Plot V, frontage 55-1/2 feet, and Plot VI, next to the Ludus(gladiators' school), frontage 75 feet . .. ]

_Orange_ (fig. 21). 

The case which deserves the first place stands by itself. It is theone piece of written evidence (as distinct from structural remains)which has survived from Roman town-planning. Curiously enough, it wasfound not in Italy but in a province, and a province which, for allits wealth of Roman buildings, has not yet revealed the smalleststructural proof of Roman town-planning. In April 1904 a scrap ofinscribed marble, little more than 18 in. Broad and high, was dug upat Orange, in southern France, right in the centre of the town. It isa waif from a lengthy document. But it chances to be intelligible. Itenumerates six plots of land--'merides' it calls them, from a Greekword meaning 'share' or 'division'--which seem to have formed oneparcel: each plot is numbered, and the length of its frontage on thepublic way (_in fronte_), the name of its lessee or _manceps_ and thatof his surety (_fideiussor_) are added. The frontages of four plotsmake up 200 ft. (those of the other two are lost), and it has beensuggested that the six together made up 240 ft. The depth--which isnot stated on the surviving fragment, but was doubtless uniform forall the plots--may then have been 120 ft. , and the whole parcel mayhave covered 120 x 240 ft. , that is, a Roman 'iugerum'. It was plainlya piece of town property. The largest 'meris', Plot v, measured only25 by 40 yds. And no one would care for such a field or farm. Besides, this plot at one end adjoined a 'ludus' or gladiatorial school, and itfronted AD K, _ad kardinem_, on to the street called in surveyinglanguage the 'cardo'. The whole land apparently belonged to one lesseewho held it from the municipality on something like a perpetuallease. [93]

 [93] For the inscription see Esperandieu, _Acad. Des Inscriptions, Comptes rendus_, 1904, p. 497; Cagnat, _Année Épigr. _, 1905, 12; and especially Schulten, _Hermes_, 1906, 1; a convenient English account is given by H. S. Jones, _Companion to Roman Hist. _, p. 22. It has been suggested by Schulten that the blocks were at first divided into plots of 35 ft. Frontage, and that the boundaries had become changed in the ordinary course of things before the survey was made. But this seems to carry conjecture rather far. 

Here, in short, is the record of an oblong 'insula' in the Roman townof Orange. It is doubtless part of a longer record, a register ofhouse-property in the whole town. Orange, Colonia Iulia SecundanorumArausio, was a 'colonia' founded about 45 B. C. With dischargedsoldiers of Caesar's Second Legion. Possibly the register was drawn upat this date; more probably it is rather later and may be connectedwith a _census_ of Gaul begun about 27 B. C. Certainly it was preservedwith much care, as if one of the 'muniments' of the citizens. The spotwhere it was dug up is in the heart of the ancient as well as of themodern town, close to the probable site of the Forum, and theinscription may have been fastened up in all its length on the wallsof some public building. If, as is likely, the town owned the soil ofthe town, the connexion of the inscription with the Forum becomes evenclearer. In any case, the town was plainly laid out in a rectangularstreet-plan. To-day its lanes are as tortuous as those of any otherProvencal town. [94] A strange chance reveals what it and many other ofthese towns must once have been. 

 [94] It has been said to show marks of streets laid out rectangularly, but neither the look of the town itself nor the plans of it seem to me to confirm this idea; compare Lentheric, _Le Rhone_, ii. 110. 

[Illustration: FIG. 22. AFTER CAGNAT AND BALLU (1911). (The six 'insulae' marked A are shown in detail in fig. 23. Unshaded'insulae' are as yet unexcavated. )]

_Timgad_ (figs. 22, 23). 

From this piece of half-literary evidence we pass to purelyarchaeological remains, and first to the province of Numidia in RomanAfrica and to the town of Timgad. The town of Thamugadi, now Timgad, lay on the northern skirts of Mount Aurès, halfway between Constantineand Biskra and about a hundred miles from the Mediterranean coast. Here the emperor Trajan founded in A. D. 100 a 'colonia' on ground thenwholly uninhabited, and peopled it with time-expired soldiers from theThird Legion which garrisoned the neighbouring fortress of Lambaesis. The town grew. Soon after the middle of the second century it was morethan half a mile in width from east to west, and its extent from northto south, though not definitely known, cannot have been much less. Thefirst settlement was smaller. So far as it has been uncovered byFrench archaeologists--sufficiently for our purpose, though notcompletely--the 'colonia' of Trajan appears to have been some 29 or 30acres in extent within the walls and almost square in outline (360 x390 yds. ). It was entered by four principal gates, three of which canstill be traced quite clearly, and which stood in the middle of theirrespective sides; the position of the south gate is doubtful. According to Dr. Barthel, the street which joins the east and westgates was laid out to point to the sunrise of September 18, thebirthday of Trajan. 

[Illustration: FIG. 23. SIX 'INSULAE' IN S. W. TIMGAD(after Prof. Cagnat). Nos. 91, 92, 99, one house each; 108, 109, 3 houses; 100, Baths. ]

The interior of the town was divided by streets into a chess-boardpattern of small square house-blocks; from north to south there weretwelve such blocks and from east to west eleven--not twelve, as isoften stated. The possible total of 132 'insulae' was, however, diminished by the space needed for public buildings, though it is noteasy to tell how great this space was in the original town. Ultimately, as the excavations show, eight 'insulae' were taken up bythe Forum, four by the Theatre, three by the various Baths, one by aMarket, one by a Public Library, and one by a Christian church. Butsome of these edifices were certainly not established till long afterA. D. 100 and the others, which must have existed from the first, weresoon extended and enlarged. A competent writer on the subject, Dr. Barthel, allows seven blocks for public purposes in the original town, but this seems too little. The blocks themselves measured on theaverage a square of 70 Roman feet (23 x 23 yards), and may havecontained one, two, three, or even four houses apiece, but they haveundergone so many changes that their original arrangements are not atall clear. The streets which divided these blocks were 15 to 16 ft. Wide; the two main streets, which ran to the principal gates, werefurther widened by colonnades and paved with superior flagging. Allthe streets had well-built sewers beneath them. 

Trajan's Timgad was plainly small. On any estimate of the number ofhouses, the original draft of veterans sent there in A. D. 100 canhardly have exceeded 400, and the first population, apart from slaves, must have been under 2, 000. This agrees with the figures of Aosta (p. 89). There, 100 acres took 3, 000 veterans and their families; here thearea is about one-third of 100 acres and the ground available fordwellings may perhaps have been one-sixth. In neither case was spacewasted. There was not probably at Aosta, there certainly was not atTimgad, any provision of open squares, of handsome facades, of templesseen down the vista of stately avenues; there were not even privategardens. The one large unroofed space in Timgad was the half-acre shutwithin the Forum cloister. This economy of room is no doubt due to thefact that the 'colonia' was not only a home for time-expired soldiers, but, as Prof. Cagnat has justly observed, a quasi-fortress watchingthe slopes of Mount Aurès south of it, just as Aosta watched itsAlpine valley. As Machiavelli thought it worth while to observe, theshorter the line of a town's defence, the fewer the men who can holdit. The town-planning of Timgad was designed on other than purelyarchitectural or municipal principles. For this reason, too, we shouldprobably seek in vain any marked distinction between richer and poorerquarters and larger or smaller houses. [95] The centurions and otherofficers may have formed the first municipal aristocracy of Timgad, asretired officers did in many Roman towns, but there can have been nodefinite element of poor among the common soldiers. 

 [95] Ballu detects a 'quartier industriel' in the outer town, but the evidence does not seem to warrant so grand a term. 

Such was Trajan's Timgad, as revealed by excavations now abouttwo-thirds complete. The town soon burst its narrow bounds. A Capitol, Baths, a large Meat-market, and much else sprang up outside the walls. Soon the walls themselves, like those of many mediaeval towns--forexample, the north and west town-walls of Oxford--were built over andhidden by later structures. The town grew from one of 360 to a breadthof over 800 yds. And as it expanded, it broke loose from thechess-board pattern. The builders of later Timgad did not resemblethose of later Turin. Even the _decumanus_, the main 'east and west'street, wandered away north-west in an uncertain curve, and all thathas been discovered of streets outside the walls of Trajan isirregular and complicated. A town-plan, it seems, was binding on thefirst builders of the 'colonia'. It lost its power within a very fewyears. [96]

 [96] Boeswillwald, Cagnat and Ballu, _Timgad_ (Paris, 1891-1905); see especially Appendix, pp. 339-349; Ballu, _Ruines de Timgad_ (Paris, 1897-1911); Barthel, _Bonner Jahrbücher_, cxx. 101. 

_Carthage_ (fig. 24). 

It remains to note another example of town-planning in a Romanmunicipality of the western Empire, which is as important as it isabnormal. Carthage, first founded--though only in an abortivefashion--as a Roman 'colonia' in 123 B. C. And re-established with thesame rank by Julius Caesar or Augustus, shows a rectangular town-planin a city which speedily became one among the three or four largestand wealthiest cities in the Empire. The regularity of its planningwas noted in ancient times by a topographical writer. [97] But theplan, though rectangular, is not normal. According to the Frencharchaeologists who have worked it out, it comprised a large number ofstreets--perhaps as many as forty--running parallel to the coast, asmaller number running at right angles to these down the hillsidetowards the shore, and many oblong 'insulae', measuring each about130 x 500 ft. , roughly two Roman _iugera_. The whole town stretchedfor some two miles parallel to the shore and for about a mile inland, and covered perhaps 1, 200 acres. Its street-plan can hardly be olderthan Caesar or Augustus, but the shape of its 'insulae' appears to bewithout parallel in that age. It comes closest to the oblong blocks ofPompeii and of Naples (pp. 63, 100), and its two theatres also recallthose towns. One reason for its plan may no doubt be found in thephysical character of the site. The ground slopes down from hillstowards the shore, and encourages the use of streets which run levelalong the slopes, parallel to the shore, and not more or less steeplytowards it. [98]

 [97] _Totius orbis descriptio_, 61 (Müller, _geogr. Graeci min. _ ii. 527); dispositione gloriosissima constat . .. In directione vicorum et platearum aequalibus lineis currens' (written probably about A. D. 350). 

 [98] _Carte archéologique et topogr. Des Ruines de Carthage_, by Gauckler and Delattre (1:5, 000); Schulten, _Archäol. Anzeiger_, 1905, p. 77; 1909, p. 190; 1911, p. 246; Audollent, _Carthage romaine_ (Paris, 1901), pp. 309, 846. The older accounts of Daux and Tissot seem less trustworthy. 

[Illustration: FIG. 24. A PART OF CARTHAGE. Plan based on the _Carte archéologique des ruines de Carthage_, byGauckler and Delattre. ]

_Laibach_ (fig. 25), _Numantia, Lincoln_ (fig. 26). 

Three or four more ordinary examples chosen at random from provincialmunicipalities may show the diffusion of town-planning in the westernRoman world. One example, from the borders of Italy, may be found justoutside the pleasant town of Laibach in southern Austria. HereAugustus in 34 B. C. Planted a 'Colonia Iulia Augusta Emona', andrecent work of Dr. W. Schmid has thrown much light on its character. The colony was in outline a rectangle of nearly 55 acres (480 x 560yds. ), and was divided up into forty-eight blocks by five streetswhich ran north and south and seven which crossed them at rightangles; of these forty-eight blocks some must, of course, have beentaken up by public buildings. They varied in size: the largest as yetplanned (II in fig. 25) measured 170 x 195 ft. , or 3/4 acre; twoothers measured 163 x 170 ft. ; while one block, which contained onelarge house not unlike the Silchester 'inn', was 112 x 168 ft. (Plan, II), and the block next it was a trifle smaller. None of thedimensions show any trace of the normal 120 or 240 ft. (p. 79). Thestreets were very broad (37-40 ft. ); one, which may be the 'cardomaximus', measured as much as 47 ft. Across. Beneath the main streetswere sewers, in the usual fashion. Round the whole town stood strongwalls, reinforced at regular intervals by square projecting towers;the four corners were not rounded but rectangular, after the fashionof Aosta and Turin (pp. 87, 90). [99]

 [99] _Correspondenzblatt des Gesamtvereins der deutschen Geschichts und Altertumsvereine_, April 1912; _Bericht vi der römisch-germanischen Kommission_ 1910-11, p. 96. Müllner's _Emona_ (Laibach, 1879), p. 19, plate 2, is wholly inadequate. 

[Illustration: FIG. 25. A PART OF LAIBACH. (From W. Schmid. )]

[Illustration: FIG. 26. LINCOLN, OUTLINE OF ROMAN WALLS. (See p. 118. )]

[Illustration: FIG. 27. LINCOLN, BASES OF COLONNADE UNDER BAILGATE. (p. 118). ]

For a second example turn to a remote corner of central Spain. Thetown of Numantia was famous in early days for its long struggle withthe armies of the Roman Republic. Under Roman rule it was whollyinsignificant. Over the débris of Numantine liberty a little Romantown grew up. But it is hardly mentioned save in one or tworoad-books. Yet it enjoyed some form of municipal status and itsstreets and houses show to the excavator traces of Romantown-planning. The streets ran parallel or at right angles to oneanother; the house-blocks measured some 50 yds. Square. [100]

 [100] Schulten, _Abhandlungen der k. Gesellschaft der Wissenschaften zu Göttingen, phil. -hist. Kl. _, viii. (1905), p. 61, plan 2; the evidence seems adequate though not wholly decisive. The Roman town Emporiae, now Ampurias, in the extreme north-east of Spain, seems to have had a rectangular street-plan, though its Greek predecessor was irregular, _Institut d'estudis catalans, anuari 1908_, p. 185. 

A third example may be drawn from our own country. Lincoln, the RomanLindum, was established as a 'colonia' about A. D. 75, and the lines ofits original area, its 'Altstadt'--for it was perhaps enlarged inRoman times, --can still be traced 'Above Hill' round the Castle andCathedral (fig. 26). It formed a rectangle just over 41 acres inextent (400 x 500 yds. ). Four gates, one of which still keeps itsRoman arch, gave access to the two main streets which divided the towninto four symmetrical quarters and crossed at right angles in thecentre. Along one of these streets, which agrees, if only roughly, with the modern Bailgate, ran a stately colonnade (fig. 27), thoughwhether this belonged to some special building or adorned the wholeextent of street is not quite certain. Beneath the same street ran, asat Timgad and Laibach and elsewhere, the town sewer (fig. 28). Of theother main street and of side streets nothing is known, but we canhardly doubt that they carried out the chess-board pattern. [101]

 [101] _Archaeologia_, liii. 236 and lvi. 371. The plan given by Mr. Fox in liii. 236 represents his own theory, which may be open to doubt. 

Probably the other four municipalities in Britain were plannedsimilarly, though the evidence is too slender to prove it. AtVerulamium (for example) near St. Albans, a local archaeologist longago claimed to detect a scheme of symmetrical house-blocks, resemblingsquares very slightly askew. Subsequent inquiry has shown that thisscheme was merely or mostly imagination. [102]

 [102] J. W. Grover, _Brit. Archaeol. Assoc. Journal_, xxvi. (1870), p. 45, plate 1. The theories of the late Mr. Bellows about the streets of Roman and modern Gloucester were equally astray, though in other ways. 

[Illustration: FIG. 28. LINCOLN. SEWER UNDER BAILGATE]

CHAPTER IX

ROMAN PROVINCIAL TOWN-PLANS. II

In the preceding chapters Roman town-planning has been treated inconnexion with towns of definite municipal rank, which bore the titles'colonia' or 'municipium'. The system is, of course, closely akin tosuch foundation or refoundation as the establishment of a 'colonia'implied in the early Empire, while the no less Roman character of the'municipium' made town-planning appropriate to this class of townalso. 

It was, however, not limited to these towns. It appears not seldom inprovincial towns of lower legal status, such as were not uncommon inBritain, in Gaul, and in some other districts. Four instances may bequoted from the two provinces just named. In the first, Autun, thetown-planning is explained by the establishment of the town full-grownunder Roman official influence. Unfortunately, however, little isknown of the buildings, and it is difficult to judge of the actualcharacter of the place. In the second case, Trier, we may conjecture asimilar official origin. At Silchester, official influence seems alsoto have been at work, and it is not impossible that the fourth case, Caerwent, may be explained by the same cause. In these two latter, however, it is more important to observe the nature of the towns, which is better known than that of any others in western Europe. Forthey embody a type of urban life which is distinct from any thatoccurs in Italy or in the better civilized districts of the Empire, and which illustrates strikingly one stratum of provincial culture. 

_Autun_ (fig. 29). 

Caesar won northern and central Gaul for the Roman Empire; it fell toAugustus to organize the conquered but as yet unromanized lands. Amongmany steps to that end, he seems to have planted new native townswhich should take the places of old native tribal capitals and shoulddrive out local Celtic traditions by new Roman municipal interests. These new towns did not, as a rule, enjoy the full Roman municipalstatus; northern Gaul was not quite ripe for that. But they wereplainly devised to help Romanization forward, and their object isdeclared by their half-Roman, half-Celtic names--Augustodunum (nowAutun), Caesaromagus (Beauvais), Augusta Suessionum (Soissons), Augusta Treverorum (Trier), and the like. [103] Of two of these, Autunand Trier, we chance to know the town-plans. The reader will notice acertain similarity between them. 

 [103] Hirschfeld, _Haeduer und Arverner_ (_Sitzungsber. Der preuss. Akademie_, 1897, p. 1102). Similar hybrid names have been created by the English in India, mostly on the North-west Frontier, where alone they have planted new inhabited sites--Lyallpur, Abbotabad, Edwardesabad, Robertsganj, and the like. But these are almost all small places or forts, and their names represent no policy of Anglicization. 

[Illustration: FIG. 29. AUTUN. After H. De Fontenay, 1889. ]

Autun stands on the site and contains the stately ruins of the RomanAugustodunum, built by Augustus about 12 B. C. He, as it seems, broughtdown the Gaulish dwellers in the old native hill-fortress of Bibracte, on Mont-Beuvray, and planted them twelve miles away on an unoccupiedsite beside the river Arroux. The new town covered an area ofsomething like 490 acres--that is, if the now traceable walls andgates are, as is generally thought, the work of Augustus. The townwithin the walls must have been laid out all at once. Quite a largepart of it, perhaps has much as three-quarters, have revealed to thecareful inquiries of French archaeologists a regular system ofquadrangular street-planning, which may very likely have extended eventhrough the unexplored quarter. The Roman street which ran through thetown from south to north, from the Porte de Rome to the Ported'Arroux, was fronted by at least thirteen 'insulae', and one of thestreets which crossed it at right angles was fronted by eleven suchblocks. They vary somewhat in size. The larger 'insulae', which liewest of the main north and south street, are oblong and measure about150 x 100 yds. (say, 3 acres); many smaller ones are more nearlysquare (98 x 98 or 109 yds. , about 2 acres). 

But the regularity of the plan is plainly the work of civilized man. When the Celts were brought to live in a Roman city, care was takenthat it should be really Roman. [104] Only we may perhaps wonderwhether the plan may not have been drawn by Augustus with an eye moreto the future than to the present and may have included more 'insulae'than there were actually inhabitants to occupy at once. That was thecase certainly in the mediaeval English town of Winchelsea, where therectangular building-plots laid out by Edward I have in great measurelain empty and untenanted to the present day. 

 [104] H. De Fontenay, _Autun et ses monuments_ (Autun, 1889), pp. 49 foll. And map (1:6, 250). The existence of a town-plan was first noticed by J. De Fontenay, _Bulletin monumental_, 1852, p. 365, but his map appears to be incorrect and his views generally are based too much on _a priori_ assumptions. 

_Trier_ (fig. 30). 

We may take another example from a northern city, Trier on the Mosel, in north-eastern Gaul (Augusta Treverorum). It was in its later days alarge city, perhaps the largest Roman city in western Europe. When itswalls were built and its famous north gate, the Porta Nigra, waserected, probably towards the end of the third century, they includeda space of 704 acres, twenty-five times as much as the originalTimgad, though, it must be added, this area may not have been whollycovered with houses. But it was then an old city. Its earliest remainsdate from the earliest days of the Roman Empire (A. D. 2), when it wasfounded, like Autun, on a spot which had (as it seems) never beeninhabited before. [105] Of this first beginning we possess vestigeswhich concern us here. Eight or nine years ago, when the modern townwas provided with drainage, the engineers of the work and the Trierarchaeologists, headed by the late Dr. Graven, combined to note thepoints where the drainage trenches cut through pieces of Romanroadway. [106]

 [105] Ademeit, _Siedelungsgeographie des Moselgebiets_, pp. 367, 431. 

 [106] H. Gräven, _Stadtplan des römischen Triers_ in _Die Denkmalpflege_, 14 Dec. 1904 (1:10, 000); the plan has been often copied, as by Cramer, _Das röm. Trier_ (Gütersloh, 1911), and Von Behr, _Trierer Jahresberichte_, i. 1908. Compare Barthel, _Bonner Jahrbücher_, cxx. 106. Trier at some time or other became a 'colonia'. When this occurred, is hotly disputed; the evidence seems to me to suggest that it was founded without colonial status and became a 'colonia latina' in the course of the first century (see Domaszewski, _Abhandlungen_, p. 153). I have therefore inserted Trier in this chapter with Autun and not in Chapter VIII with Orange and Timgad. 

These points yielded a regular plan of streets crossing at rightangles, which in many of its features much resembles that of Autun. Thirteen streets were traced running east and west, and eight (Dr. Graven says seven but his plan shows eight) running north and south. The east and west streets, with two exceptions, lay some 320 ft. Fromone another. The north and south streets varied, some observing thatdistance, others being no more than 260 ft. Apart. As a result, therectangular house-blocks varied also in size. The largest seem to bethose which fronted a street that crossed the town from east to west, from the Imperial Palace to the Baths and the West Gate, andcorresponds roughly with the present Kaiserstrasse. This may well havebeen the _decumanus_, the main east and west street of the 'colonia', and hence the house-blocks fronting it may have been unusually large(p. 77). One of them, near the Neumarkt, reached the awkward size ofnearly 3-1/2 acres (320 x 460 ft. ). Others elsewhere were smaller, many measuring 320 x 320 ft. , and others again 320 x 245 ft. , ratherless than 2 acres. In general, the 'insulae' on the east and westsides of the town were larger than those in the centre. The whole hasa resemblance to Autun, and is more irregular than writers on Trierare ready to allow. [107]

 [107] Gräven estimated that, except in the central street, all the 'insulae' measured 300 Roman ft. (290 English ft. , 88 metres), but his plan suggests rather 100 metres. We need in reality that larger plan which he did not live to complete. 

How many houses may have occupied either a large or a small 'insula'is uncertain; indeed, we know next to nothing of the private houses ofRoman Trier. Nor can we fix the number of the 'insulae'. On the west, and still more on the east and south-east of the town, much of thearea was not touched by the drainage works and therefore wentunexplored. We have proof only of streets and buildings for a mile inlength and half a mile in breadth. 

[Illustration: FIG. 30. TRIER. From plan by the late Dr. Gräven. ]

Nevertheless we may make some guess at the original area. Thestreetage itself plainly dates from the original foundation of theRomano-Gaulish town by Augustus. There is, indeed, no other epoch inits history, so far as we know it, when a complete laying out couldhave been carried through. On the other hand, it is not probable thatthe first town was a mile long and half a mile wide. Possibly, as anacute German archaeologist has suggested, the small 'insulae' in thesouth of the town may indicate the line of an original wall and ditchwhich, like the first walls of Timgad, were overrun later by anexpanding town. Certainly, early graves found hereabouts show thatthis space lay once outside the inhabited area, and similar evidencehas been noted both on the north of the town in the Simeonstrasse, andon the west near the Mosel Bridge. If this be so, Augusta Treverorummay have at first covered only 120 or 130 acres; then, as the placespread beyond its original limits, its builders followed more or lessclosely the lines of the first streets, and, save near the PortaNigra, continued the chess-board pattern as it was continued at Turin. 

_Silchester_ (figs. 31, 32). 

Silchester, Calleva Atrebatum (fig. 31), shows a different picture, which is the more interesting because the excavations carried out in1890-1909 have given us a fuller knowledge of the town than of anyother Roman site in the western provinces. [108] It was, apparently, the old tribal capital of the Atrebates and the county-town of itsdistrict in Roman days; though not possessing the full municipalstatus, it was probably the seat of local government for aconsiderable neighbourhood. In outline it was an irregular eight-sidedarea of 100 acres, defended by a strong stone wall, which was addedlong after the original foundation. Internally it was divided up bystreets which, except near the east gate, run parallel or at rightangles to one another. Its buildings are: a Forum and Basilica, asuite of public baths, four small temples, a small Christian church, ahotel, and a large number of private houses. Its area is by no meansfilled with buildings. Garden ground must have been common and cheap, and the buildings themselves do not form continuous streets; they donot even front the roadway in the manner of houses in Italian towns. In these respects Silchester differs widely from any of the exampleswhich we have already considered, so far as their internal buildingsare known to us. I will not call it a 'garden city', for a garden cityrepresents an attempt to add some of the features of the country to atown. Silchester, I fancy, represents the exact opposite. It is anattempt to insert urban features into a country-side. 

 [108] For accounts of the Silchester excavations, see _Archaeologia_, vols. Lii-lxii, and _Victoria Hist. Of Hampshire_, i. 271, 350; large plan by W. H. St. John Hope (1:1, 800) in _Archaeol. _ lxi. 

[Illustration: FIG. 31. SILCHESTER. (For detail see fig. 32. )]

Most of it must have been laid out at once. At any rate, the area ofwhich the 'insulae' numbered X, XXI, XXXV, and XIX form the corners, and the Forum the centre, must have been planned complete from thefirst. This covers just 40 acres, and is divided into rectangularplots of which the smallest covers a little less than an acre and ahalf, while the largest fall little short of 3-1/2 acres. [109] Outsidethis area, the division of the town into 'insulae' is less completelycarried through, although most of the streets run straight on as faras the walls, and one or two details may tempt us to think that thedivision into 'insulae' was at some time extended beyond the lineultimately taken by the walls. 

 [109] The plots are of three sizes, two being 3-4 acres (128 x 130 yds. ), six about 2. 4 acres (128 x 89 yds. ), and six about 1. 4 acres (89 x 80 yds. ). In the third size the dimension of 240 Roman feet (p. 79) can perhaps be recognized. 

[Illustration: FIG. 32. DETAILS OF FOUR INSULAE, THE FORUM AND THECHURCH AT SILCHESTER. (From _Archaeologia_. )]

But whatever the exact amount of Roman building and Roman street-plangiven to Silchester when it was first laid out, the place is not ineffect a real town. It is not merely that, as I have said, the housesdo not form continuous streets. A glance at the houses will show thatthey could not possibly be fitted into streets. The types of househere visible are not town houses. They are the types which appearamong the 'villas', that is, the landlords' or the farmers' dwellings, up and down the rural districts of Roman Britain and northern Gaul, and the town which they constitute is a conglomeration of countryhouses. The reverse has taken place of that which we often see to-dayin England. Our modern builders and architects had--until perhapsquite recently--only one idea of a small house, the house, namely, which to-day characterizes the monotonous streets in the poorerquarters of our new towns, with its front door and bow window on oneside, its offices behind, and its two other sides left blank for otherhouses to stand against. This is a town house. Yet our modern buildersuse it, all by itself, in the most desolate country districts. I cameacross one such not long ago, when driving over a lonely valley inExmoor. There it stood, with no other house near it, yet with its twosides blankly waiting for the street that ought to form itself to theright and left. 

The opposite of this has occurred at Calleva; here the rural house hasbeen used, with scarcely a change, to form a town. We see the Romanstreet-plan introduced in surroundings which are not properly urban. The outward expression of the civilised municipal system jostlesagainst a provincial and rural life. Here was a premature attempt tomunicipalize the Briton, which outstripped the readiness of the Britonto be municipalized. Silchester was probably a tribal centre beforethe Roman came; for awhile it may have remained much the same underRoman rule. But forty years after the Roman Conquest, in the reign ofVespasian (about A. D. 70-85), the Romanization of the whole provinceappears to have rapidly advanced. It was, indeed, encouraged by theHome Government. Various details suggest that the laying out ofSilchester belonged to this very date. But to this the Callevan failedto rise. He learnt much from Rome; he learnt even town-life; he didnot learn town-life in its highest form. When his town had been'haussmannized' and fitted with Roman streets, and equipped with RomanForum and Basilica, and the rest, he yet continued to live--perhapsmore happily than the true townsman--in his irregularly grouped housesand cottages amid an expanse of gardens. The area of Silchesterdiffered little from that of Aosta; its population, if we may judge bythe number of dwelling-houses, was hardly as large as that of Timgad. 

_Caerwent_ (fig. 33). 

I turn lastly to another Romano-British town, Caerwent (VentaSilurum), between Chepstow and Newport in Monmouthshire. It is asmaller town than Silchester. Both towns perhaps began with the samearea, 40 or 45 acres. But Caerwent never expanded; it remained notmuch more than 45 acres within the walls. Land was probably valuablewithin it; certainly its houses are packed closer, and its gardenground is smaller than at Silchester. Its general type is, however, the same. It has a very similar Forum and Basilica, Temples, anAmphitheatre, and a large number of private houses which resembleclosely those of Silchester. It has, moreover, at least in the partsthat have been so far excavated, distinct traces of a rectangularstreet pattern, which, if it was carried through the whole town, wouldprovide (including the Forum) twenty 'insulae'. The size of theseblocks cannot be determined with any precision. Indeed, in some casesthe houses seem to have encroached on and distorted the street-plan. Probably it would be true to say that the average block covered anacre and a half or an acre and two-thirds. [110] We do not know enoughof the history of Caerwent to do more than guess how this street-plancame to it. Very likely the same process of establishing aRoman-looking town for a local capital was adopted here as atSilchester. Very likely the step was taken in the same period as atSilchester, that is, in the last thirty years of the first century. Its occurrence is significant. Caerwent lay remote in the far west, with nothing but garrisons beyond it. It was the outpost of Roman citylife towards the Atlantic. It was the only town of Roman municipalplan in Britain which was swept by Atlantic breezes. [111]

 [110] The three best defined examples measure about 260 x 260, 260 x 280, 275 x 275 ft. (1. 55, 1. 61, and 1. 73 acres respectively). The unit of 240 Roman feet (p. 79) does not appear at Caerwent. 

 [111] Accounts of the Caerwent Excavations, 1899-1910, will be found in _Archaeologia_, vols. Lvii-lxii. A good plan of the whole town, from which fig. 33 is taken, was issued in vol. Lxii, plate 64, by Mr. F. King, architect to the excavations (scale, 1:900). 

[Illustration: FIG. 33. CAERWENT. (Reduced from plan by F. King. )]

Silchester and Caerwent did not stand alone in Britain. At Wroxeter, the ancient Viroconium, tribal centre of the Cornovii and aRomano-British country-town much like Silchester, though somewhatlarger, oblong 'insulae' have recently been detected by Mr. J. P. Bushe-Fox which measure 103 x 126 yds. (2-2/3 acres). AtCirencester, the Romano-British centre for the canton of the Dobuniand a still larger town than Wroxeter, the 'insulae' near the Basilicaseem to have measured as much as 120 yards in length, though fulldetails have not yet been obtained. Both these towns may be ascribedto the later years of the first century and to the same civilizingprocess as Silchester and Caerwent. As further Romano-British townsare uncovered, we may therefore hope for more examples. Howeverimperfectly the inner meaning of town-planning was understood, it wasplainly common in the south of Roman Britain. 

NOTE. THE EASTERN PROVINCES. 

To complete the survey of Roman provincial town-planning, we mustglance briefly at the East. Here towns of Roman origin were few, andof those few scarcely any are well known. But they do not lackinterest. For example, take Antinoê, built by Hadrian in memory of hisfavourite Antinous, on the banks of the Nile. It was a parallelogrammore than 3 miles round, which covered an area of 360 acres. Two mainstreets, each colonnaded, crossed at right angles and cut it into fourparts. Of the other streets, nothing certain seems to be known. Butreferences to the town in papyri denote four quarters of it by variousletters, Alpha, Beta, Gamma, Delta, and distinguish its house-blocksby the term Plintheion with a numeral attached. Thus, a house isdescribed as lying 'in the letter Delta and the Plintheion 7'. Ourdocuments show that there were in Antinoê at least eleven of thesePlintheia. [112] It is fairly plain that they are rectangular'insulae', of either Roman or Hellenic type, while the general fashionof the town and of its monuments suggest a Greek rather than anItalian city. 

 [112] _Exploration des ruines d' Antinoe_, by A. C. Gayet (Annales du Musée Guimet, xxvi, Paris, 1897); _Grundzüge der Papyruskunde, _ Wilcken, i, pp. 49, 50. Professor A. S. Hunt refers me to the following papyri:--Reinach, 49. 11; Oxyrhynchus, 1110. 9-10 and note there; Brit. Mus. 1164 (c) 12. The numeration of the divisions of the town by letters was borrowed from Alexandria, where the five parts of the city were known as A, B, C, D, E. For plans see the Napoleonic _Description d'Égypte_ iv (Paris, 1817), plate 53, and E. Jomard, _Antiquités d'Égypte_ (1818), chap. Xv. 

[Illustration: FIG. 34. BOSTRA. (After Baedeker. )]

Another instance may be found still further east, in the land beyondJordan, at the capital of the Haurân, Bosrâ, anciently Bostra. Littlehas been achieved in the way of exploration of this site beyondstudies of the stately ruins of theatres, palaces, temples, triumphalarches, aqueducts. Little can therefore be said as to the date of itsground-plan. But it was rectangular in outline, or nearly so; and itsstreets crossed at right angles and enclosed rectangular insulae. [113]The place owes all its greatness to Rome. During the second century itwas the fortress of the Legio III Cyrenaica, which guarded this partof the eastern Roman frontier. About A. D. 225 it became a 'colonia, 'and perhaps we should date from this the town-plan just described(fig. 34). 

 [113] Baedeker, _Palestine and Syria_ (1906), p. 162. 

This rectangular planning remained long in use in the Eastern Empire. When in A. D. 705 (as it seems) the town of Chersonnesus in the Crimeawas rebuilt after a total destruction, it was rebuilt on a symmetricalplan of oblong 'insulae' (25-30 by 60-70 yds. Area). Its streets weremean and narrow. But their plan at least was apparently more regularthan that of their predecessors. [114]

 [114] Minns, _Greeks and Scythians_, pp. 493, 508, and references there given. 

CHAPTER X

ROMAN BUILDING-LAWS

Archaeology tells us that the western half of the Roman Empire andmany districts in its eastern half used a definite town-plan which maybe named, for brevity, the chess-board pattern. It remains to askwhether literature, or at least legal literature, provides any basisof theory or any ratification of the actual system which archaeologyreveals. Of augural lore we have indeed enough and to spare. We knowthat the _decumanus_ and the _cardo_, the two main lines of the Romanland-survey and probably also the two main streets of the Romantown-plan, [115] were laid out under definite augural andsemi-religious provision. We should expect to find more. A system oftown-planning that is so distinctive and so widely used mightreasonably have created a series of building-laws sanctioning ormodifying it. This did not occur. Neither the lawyers nor even theland-surveyors, the so-called Gromatici, tell us of any legal rulesrelative to town-planning as distinct from surveying in general. Thesurveyors, in particular, are much more concerned with the soil of theprovince and its 'limitation' and 'centuriation', than with thearrangements of any individual town, and, whatever their value forextramural boundaries, [116] throw no light on streets and 'insulae'. 

 [115] See p. 73. 

 [116] Schulten, _Hermes_, 1898, p. 534. 

The nearest approach to building-laws which occurs is a clause whichseems to be a standing provision in many municipal charters andsimilar documents from the age of Cicero onwards, to the effect thatno man might destroy, unroof, or dismantle an urban building unless hewas ready to replace it by a building at least as good or had receivedspecial permission from his local town council. The earliest exampleof this provision occurs in the charter of the municipality ofTarentum, which was drawn up in the time of Cicero. [117] It isrepeated in practically the same words in the charter of the 'coloniaGenetiva' in southern Spain, which was founded in 44 B. C. ; it recursin the charter granted to the municipality of Malaga, also in southernSpain, about A. D. 82. [118] Somewhat similar prohibitions of theremoval of even old and worthless houses without special leave areimplied in decrees of the Roman Senate passed in A. D. 44 and A. D. 56, though these seem really to relate to rural rather than to urbanbuildings and were perhaps more agrarian than municipal in theirobject. [119] Hadrian, in a dispatch written in A. D. 127 to an easterntown which had lately obtained something like municipal status, includes a provision that a house in the town belonging to oneClaudius Socrates must either be repaired by him or handed over tosome other citizen. [120] Similar legislation occurs in A. D. 224 and inthe time of Diocletian and later. [121]

 [117] Mommsen, _Eph. Epigr. _ ix, p. 9; Dessau, _Inscr. Sel. _ 6086; 'nei quis in oppido quod eius municipi erit aedificium detegito neive demolito neive disturbato nisei quod non deterius restiturus erit nisei de senatus sententia. Sei quis adversus ea faxit, quanti id aedificium fuerit, tantam pequniam municipio dare damnas esto eiusque pequniae quei volet petitio est. ' (English translation in E. G. Hardy's _Roman Laws and Charters_, p. 101. )

 [118] Dessau, 6087, 6089; Hardy, _Roman Laws_, part 2, pp. 34, 108. 

 [119] For these decrees, which are practically equivalent at this date to laws, see _CIL_. X. 1401 = Dessau 6043, and de Pachtère in _Mélanges Cagnat_, p. 169. 

 [120] For the letter of Hadrian see _Bulletin de Corresp. Hell. _ x. 111; it is quoted by Bruns, _Fontes_, 1909, p. 200. Compare the _Historia Augusta_, Life of Hadrian, ch. 18. 

 [121] Mommsen, _Eph. Epigr. _ iii, p. 111 and _Ges. Schiften_, i. 158, 263, 371; Liebenam, _Städteverwaltung_, 393. 

Rules were also laid down occasionally to forbid balconies and similarstructures which might impede the light and air in narrow streets, andit was a common rule that cemeteries and brickyards must lie outsidethe area of inhabitation. At Rome too, efforts were made by variousemperors to limit the height of the large tenement houses which thereformed the 'insulae'. These limits were, however, fixed haphazardwithout due reference to the width of the streets; they do not seem tooccur outside of Rome, and even in Rome they were very scantilyobserved. 

But in general no definite laws were framed. Probably themunicipalities were somewhat closely tied in the administration ofmunicipal property and had to refer schemes for the employment even ofthe smallest bit of vacant space to the 'patron' or the _curator_ ofthe town. But, apart from the provisions mentioned above, they had nospecific rights, that are recorded, against private owners orbuilders. It was only once, after Rome itself had been burnt out, thatan imperial order condemned landowners who 'held up' their groundinstead of using it, to forfeit their ownership in favour of any onewho offered to build at once. 

CHAPTER XI

THE SEQUEL

What was the sequel to this long work of town-planning? Two factsstand out distinct. First, the Roman planning helped the towns of theEmpire to take definite form, but when the Empire fell, it too met itsend. Only here and there its vestiges lingered on in the streets ofscattered cities like things of a former age. But, secondly, from thisdeath it rose again, first in the thirteenth century, withever-growing power to set the model for the city life of the modernworld. 

I. The value of town-planning to Roman civilization was twofold. Itincreased the comfort of the common man; it made the towns strongerand more coherent units to resist the barbarian invasions. When, after250 years of conflict, the barbarians triumphed, its work was done. Inthe next age of ceaseless orderless warfare it was less fit, with itsstraight broad streets, for defence and for fighting than the chaos ofnarrow tortuous lanes out of which it had grown and to which it nowreturned. The cases are few in which survivals of Roman streets haveconditioned the external form of mediaeval or modern towns. We inEngland tend perhaps to overrate the likelihood of such survivals. Ourclassical education has, until very lately, taught most of us more ofancient than of mediaeval history, and when our antiquaries find townsrectangular in outline and streets that cross in a Carfax, they givethem a Roman origin. 

Such a tendency is wrong. Plentiful evidence shows that even in Italyand even in towns where men have dwelt without a break since Romandays, the Roman streets, and with them the Roman town-plans, have faroftener vanished than endured. Rome herself, the Eternal City, useshardly one street to-day which was used in the Roman Empire. Some fewItalian towns, described in detail above, have a better claim to becalled 'eternal'; half a dozen in northern Italy retain their ancientstreets in singular perfection. Yet even there cities like Padua andMantua, Genoa and Pisa, have lost the signs of their older fashion. So, too, in the provinces. In the Danubian lands only one town caneven be supposed to preserve a few of its Roman streets. In all theonce great cities of that region, Sirmium and Siscia, Poetovio andCeleia and Emona, they have wholly gone; you may walk across the sitesto-day and seek them in vain in modern street or hedgerow or lane. InGaul there were many Roman municipalities in the south; there weremany towns of lesser rank but equal wealth in the centre and west andnorth. But we owe our knowledge of their town-plans to an inscriptionfrom Orange and to some excavations at Autun and Trier. Cologne andTrier alone, or almost alone, keep Roman streets in modern use, andthey are significant. Both became Roman towns in the first century;both held colonial rank; both have lived on continuously ever sinceand hardly changed their names. Yet both bear to-day the stamp of theMiddle Ages, and the Roman streets which they use are small and nearlyunrecognizable fragments. 

There is, indeed, no law of survivals. Chance--that convenient ancientword to denote the interaction of many imponderable forces--has ruledone way in one place and otherwise in another. Sometimes monumentshave alone survived, sometimes only streets, and we can seldom givereasons for this contrast of fates. At Pola, gates, temples, andamphitheatre still tell of the Roman past and the modern town-squarekeeps so plainly the tradition of the Forum that you cannot walkacross it without a sense of what it was. Yet not a single streetagrees with those of the Roman 'colonia'. In the Lombard and Tuscanplains, at Turin and Pavia and Piacenza, at Florence and Lucca, theRoman streets are still in use, just as the old Roman field-ways stilldivide up the fertile plains outside those towns. But, save in Turin, hardly one Roman stone has been left upon another. In the no lessfertile plain of the lower Rhone, at Nîmes and Arles and Orange, thestately ruins wake the admiration of the busiest and least learnedtraveller; of the Roman streets there is no sign. 

Britain has enjoyed less continuity of civilization than any otherwestern province; in Britain the survivals are even fewer. In London, within the limits of the Roman city, no street to-day follows thecourse of any Roman street, though Roman roads that lead up to thegates are still in use. At Colchester the Roman walls still stand; theplaces of the Roman gates are known; the masonry of the west gate isstill visible as the masonry of a gateway. But the modern and ancientstreets do not coincide, and the west gate, which has so wellwithstood the blows of time, can hardly be reached by road from withinthe city. At York the defences of the legionary fortress have stilltheir place in the sun, but the 'colonia' on the other bank of theOuse has vanished wholly from the surface, walls and streets together, and the houses of the citizens of Eburacum are known solely by findsof mosaic floors. At Lincoln the Roman walls and gates can easily betraced and one gate rears its arch intact, but the Bailgate alonefollows, and that erratically, the line of a Roman street. The roadfrom the Humber, thirty miles north of Lincoln, runs to-day, as it hasrun for eighteen centuries, under the Newport arch and through themodern town and passes on southwards. That long straight road hasgiven a feature to Lincoln, but it is a feature due to the Romanhighway outside the town, not to the streets within it. Lincoln itselfis as English as Cologne and Trier are German. 

II. But if Roman streets have seldom survived continuously to moderndays, if Roman town-planning perished with the western Empire, it hasnone the less profoundly influenced the towns of mediaeval and modernEurope and America. Early in the thirteenth century men began torevive, with certain modifications, the rectangular planning whichRome had used. Perhaps copying Roman originals seen in northern Italy, Frederic Stupor Mundi now built on a chess-board pattern the TerraNova which he founded in Sicily. Now, in 1231, Barcelonette was builtwith twenty square 'insulae' in south-eastern France. Now, too, the'Bastides' and 'Villes Neuves' of southern France and towns likeAigues-Mortes (1240) were built on similar plans. [122]

 [122] For the Bastides and Villes Neuves see Dr. A. E. Brinckmann, _Deutsche Bauzeitung_, Jan. -Feb. , 1910, and, for an example, fig. 35. Many of them may be earlier than 1200 (A. Giry, _Bibl. De l'École des Chartes_, xlii. 451), but those with more or less chess-board plans seem later. 

[Illustration: FIG. 35. PLAN OF A BASTIDE TOWN, SAUVETERRE-DE-GUYENNENEAR BORDEAUX (A. D. 1281). (By Dr. A. E. Brinckmann. )]

Soon after, the chess-board pattern came to England and was used inEdwardian towns like Flint[123] and Winchelsea; then, too, it wasadopted at the other end of the civilized world by German soldiers inPolish lands. Cracow, for example, owes to German settlers in themid-thirteenth century that curious chess-board pattern of itsinnermost and oldest streets which so much puzzles the modernvisitor. [124] It is unnecessary here to follow further the renaissanceof town-planning. By intervals and revivals it continued to spread. In1652 it reached Java, when the Dutch built Batavia. In 1682 it reachedAmerica, when Penn founded Philadelphia. In 1753, when Kandahar wasrefounded as a new town on a new site, its Afghan builders laid out aroughly rectangular city, divided into four quarters meeting at acentral Carfax and divided further into many strangely rectangularblocks of houses. [125]

 [123] Compare E. A. Lewis, _Medieval Boroughs of Snowdonia_, pp. 30, 61 foll. 

 [124] So, too, Lemberg. Compare R. F. Kaindl, _Die Deutschen in den Karpathenländern_, i. 178, 293; ii. 304; he does not, however, deal with the actual plans. 

 [125] I have to thank the late Sir Alfred Lyall for a sight of a survey made by English engineers in 1839. 

But in growing, the old town-planning has passed into a new stage. TheRomans dealt with small areas, seldom more than three hundred acresand often very much less. The town-plans of the Middle Ages and evenof modern times affected areas that were little larger. Only the lastdays have brought development. Till the enormous changes of thenineteenth century--changes which have transferred the termination ofancient history from A. D. 476 to near A. D. 1800--the older fashionsremained, in town-life as in most other forms of civilized society. Towns were still, with few exceptions, small and their difficulties, if real, were simple. Save in half a dozen abnormal capitals, theyhad, even in relatively modern days, no vast populations to be fed andmade into human and orderly citizens. They had no chemical industries, no chimneys defiling the air, or drains defiling the water. Now, builders have to face the many square miles of Chicago or BuenosAyres, to provide lungs for their cities, to fight with pollutedstreams and smoke. Their problems are quite unlike those of theancients. When Cobbett, about 1800, called London the Great Wen, hecontrasted in two monosyllables the ancient ideal of a city with theugly modern facts. 

It is not, therefore, likely that modern architects or legislatorswill learn many hints from plans of Timgad or of Silchester. There arelessons perhaps in the growth of Turin from its little ancientchess-board to its modern enlargement, but such developments are rare. The great benefit to modern workers of such a survey as I haveattempted is that it shows the slow and painful steps by which mankindbecame at last able to plan towns as units, yet inhabited byindividual men and women, and that it emphasizes the need for definiterules and principles. Nor is it perhaps quite superfluous to-day topoint out how closely, even after the great upheaval of the nineteenthcentury, the forms of modern life depend on the Roman world.