^{page 170 note 1} Sinclair, , Hesiod, pp. xxxvii, xxxviii note 1Google Scholar, argues for the Theogony as a genuine work of Hesiod. However this may be, to calculate the date of the Theogony from the geographical knowledge of the West shewn by that poem, on the assumption that this knowledge must be later than the foundation of the Greek colonies in the West, is to beg the whole question. The Western Geography of the Theogony consists of names of islands, mountains and peoples, not of Greek colonies.

^{page 170 note 2} It is possible that the Sicel and Sicily passages of the Odyssey may reflect Minoan-Mycenaean contacts with South Italy and Sicily, for which there is good archaeological evidence. Personally I do not think this view much more probable than the old one which dated those passages as later than the foundation of Naxos. The Odyssey passages seem to me to reflect ninth- and eighth-century geographical knowledge which was the result of a ninth- and eighth-century trade with the West.

^{page 170 note 3} E.g. Gwynn, , J.H.S. xxxviii 88Google Scholar.

^{page 171 note 1} Hesiod, , Works and Days 631 ffGoogle Scholar.

^{page 171 note 2} Ibid. 646–7; 618 ff.

^{page 171 note 3} See note 1.

^{page 171 note 4} Etna, Ortygia, Etruscans: Strabo i 23, ‘Hesiod’ Oxy. Pap. 1358, Fr. 2, 25. Cape Pelorus: Diod. IV. 85. Agrios, and Latinos, : ‘Hesiod’ Theogony 1011Google Scholar. Ligurians: Strabo VII 300, but Libyans in ‘Hesiod’ Oxy. Pap. 1358, Fr. 2, 15. I can see no reason for preferring the reading of the papyrus to that of Strabo.

^{page 171 note 5} I suppose the argument of ‘Phoenician Carriers’ of Greek goods as an explanation of the presence of pre-colonisation Greek pottery in the West must be met. If these Greek Geometric pots were carried to the West by Phoenician ships we should expect to find in the graves which contain them at least some definite trace of Oriental wares. There are, in fact, a few scarabs in the Sicel graves and a few at Canale. Nothing else can be referred to the Orient except, perhaps, what is just possibly a Phoenician strain in some of the pottery from the Sicel graves discussed below.

(In parenthesis it is perhaps worth saying, since some historians regard all scarabs as ‘datable documents,’ that the majority of scarabs of this period can neither be dated nor referred to any definite part of the Orient. Nor does their presence in Greek and Barbarian graves prove that those graves are later than the foundation of Naucratis; cp. the scarab finds from Eleusis, the Argive Heraeum, Knossos, Orthia, and, above all, the Heraeum at Perachora.)

Many of the Oriental imports in eighth-century Etruria are not any more necessarily Phoenician than Lydian or East Greek (cf. the eighth-century Lydian (?) Lydion from Vulci discussed below, p. 193), and the theory of Phoenician carriers leaves the activity of Greek ‘Resident Potters’ in Etruria in the eighth century unexplained and inexplicable.

In general there is very little indeed of the period c. 800 to c. 735 B.C. in Italy and Sicily that can be attributed to the Phoenicians with any sort of probability and I can see no archaeological reason why they should be given the credit of having carried the Greek Geometric pottery of the eighth century to the West. To interpret the archaeological evidence we are about to discuss by an appeal to ‘Phoenician Thalassocracy’ and ‘Phoenician Carriers’ will, of course, save trouble and doubtless thus commend itself to some historians; but it is really like an appeal to the Absolute of Schelling for an explanation of metaphysical phenomena, an appeal to the night in which all cows are black.

^{page 172 note 1} As a general explanation of the distribution of Greek pottery in Greek settlements in the post-colonisation period the ‘Travelling Potter’ hypothesis does not deserve serious refutation. ‘Bücher's discussion of this question has been wholly ignored by archaeologists’ (Hasebroek, Macgregor-Fraser translation, 50, n. 4)—a fact significant enough in itself. Are we to suppose that Clitias travelled to Gordium, painted one perfect cup there, and then, with Ergotimos, to Etruria to make the François vase—presumably carrying the distinctive Attic clay with them? The extraordinarily wide and scattered distribution of Attic Black-figure and Red-figure vases which bear the same artist's signature, let alone attributed pieces that bear such signatures as surely as if written alphabetically, is a sufficient refutation of the ‘Travelling Potter’ hypothesis in itself. And are we to suppose that these ‘Travelling Potters’ went with shiploads of the distinctive Attic or Corinthian clay? If ‘No,’ they must have been alchemists of no mean order to produce the same colour and texture from the clay of South Russia, Sicily, Etruria and Spain. If ‘Yes,’ then there was trade in bulk, to refute which the ‘Travelling Potter’ hypothesis was constructed. One is almost tempted to ask whether the engraved ostrich eggs of the various Etruscan cemeteries were laid by a travelling ostrich with a travelling egg engraver in attendance.

Local imitation of the better known Greek fabrics is, of course, a phenomenon perfectly well-known to archaeologists. Generally these classes are easily recognisable. In Etruria, for example, it is perhaps possible to distinguish between imported Protocorinthian, Barbarian local imitation of that fabric, and the local imitation—not always Cumaean—which is the work of a Greek artist. (We should, I think, be certain of this last class even without the literary tradition in Polybius vi 2; Dion. Hal. iii 46; Diod. viii 31; Cic., de Rep. ii 19–20Google Scholar; Livy i 34, iv 3; Pliny, N.H. xxv, 5, 43 (12)Google Scholar; Strabo 219, 378.) A list of local imitations of Corinthian has been compiled by Payne, (Necrocorinthia 189–209)Google Scholar. The phenomenon is not nearly so common nor so unrecognised by archaeologists as Hasebroek and his followers seem to suppose.

^{page 173 note 1} It is, of course, possible, and in the case of Graeco-Siculan pottery of the Fourth Siculan period (c. 700 B.C. onwards) it can be regarded as certain, that local Barbarian fabrics continued to imitate Greek originals which had long passed out of fashion among the Greeks; but the fact remains that even a late seventh- or sixth-century Barbarian imitation of a Greek eighth-century original proves the importation of such an original at the time when it was still an object of commerce. Further, these local survivals of a defunct artistic tradition frequently betray their date by an eclectic adoption of contemporary forms. The Daunian, Peucetian and Messapian Schools of Geometric (with which, being all of post-colonisation date and with very little, if any, debt to eighth-century Greek Geometric, we are not concerned) are obvious examples. More important for us is the Sicel Geometric of the Fourth Siculan period which frequently betrays its date by shapes and decoration derived from contemporary seventh- and sixth-century Greek originals as well as by its association with such pottery. I have endeavoured to exclude from my lists all local pottery of seventh-century date even if it betrays a debt to eighth-century originals.

^{page 173 note 2} It is impossible for me to make anything like due acknowledgment of the extent of my indebtedness to Mr. Payne in the archaeological section of this article.

^{page 174 note 1} MacIver, RandallVillanovans and Early Etruscans 169, 170Google Scholar. See also fig. 60 (‘Dipylonic vase’), Pl. 33 and pp. 176–177. (Plate 33 contains 1. Greek imports; 2. Barbarian copies of Greek work; 3. Purely Barbarian vases.) In spite of this and other matters relating to Greek Archaeology in Dr. Randall MacIver's Villanovans and Early Etruscans and The Iron Age in Italy with which I entirely disagree, I feel in duty bound to acknowledge my great indebtedness to these works for my first introduction to a great portion of my material.

^{page 174 note 2} M.A. xix 310Google Scholar.

^{page 174 note 3} E.g. Fig. 2.

^{page 175 note 1} The archaeological evidence is admittedly very incomplete and the neighbourhood but little excavated.

^{page 175 note 2} The supposed neglect by the Greeks of the Adriatic in the Archaic period and in the fifth century, explained, as it frequently is, by geographical and climatic generalisations of dubious accuracy and validity, has been greatly over-estimated. The Attic imports to Bologna and Spina are sufficient to prove that at least during the fifth century the Northern Adriatic was anything but a ‘Mare Clausum’ for the Greeks. For the earlier period we have Corinthian pottery (Payne, Necrocorinthia 189Google Scholar) from near Venice and Ancona and the Umbrian slave depicted on a Corinthian vase (Ibid. 122 note 3). Cf. Strabo 376 on Aeginetan colonies among the Umbrians and Herodotus I, 163 on Phocaean voyages in the Adriatic. The whole question of early Greek colonisation and commerce in the Adriatic is being dealt with by Mr. R. L. Beaumont, to whom I am indebted for knowledge of statistics which shew the weakness of the traditional geographical and climatic explanation.

^{page 175 note 3} See Mayer Apulien Pls. 3 and 4.

^{page 175 note 4} Cf. Payne, B.S.A. xxix 270 nos. 7, 8 and 10Google Scholar.

^{page 176 note 1} Strabo 278.

^{page 176 note 2} Johansen, Les Vases Sicyoniens 182¹, 89Google Scholar.

^{page 176 note 3} Herodotus, vii 170. Strabo 278–9.

^{page 176 note 4} N.Sc. 1900, 419 ffGoogle Scholar. Cf. Johansen op. cit. 43. 10. No chronological conclusions can be derived from the stratification of this hopelessly unstratified deposit. The Mycenaean pottery found in it certainly demonstrates contact between the neighbourhood of Tarentum and the Aegean world in the Mycenaean period. There is, as far as I know, no certain example of Protogeometric nor of true Geometric from the Scoglio del Tonno find. The Mycenaean pottery was mixed with sherds of Protocorinthian of the first half of the seventh century.

^{page 177 note 1} Contrast the otherwise parallel native site of Torre Galli, 22 km. S.W. of Monte Leone, which has no Greek Geometric but continues into the seventh and sixth centuries, when it received a small but steady stream of Corinthian imports. The Canale settlement was at the very gates of Locri. Torre Galli was far removed from any early Greek colony.

^{page 177 note 2} Polybius describes the natives as Sicels. The marked similarity of the burial customs and tomb architecture of the dwellers on the Canale plateau to those of the Sicels of the Third Siculan period has led Orsi to identify them as Sicels.

^{page 177 note 3} The possibility that the painted, wheel-made vases are the work of a Greek ‘Travelling Potter’ working for the local market with local clay is excluded by the fact that their clay is utterly different from that of the local fabric whose texture resembles a none too happy amalgam of cow-dung and charcoal. Further, it is necessary to postulate for the Canale plateau a ‘Travelling Greek Bronze-worker’ as well as a ‘Travelling Potter’ and hence a local supply of bronze! But even if this tiresome and impossible hypothesis of ‘Travelling Potter’ and ‘Travelling Bronze-worker’ is accepted it still proves the existence of pre-colonisation contacts between the Canale plateau and the Greek world.

^{page 178 note 1} Orsi's suggestion that it may come from the Argolid I cannot understand unless it is a reference to the Argivo-Cycladic category of Dugas, which has been demolished by Payne and Buschor. See, however, No. 11, p. 179.

^{page 179 note 1} I am not certain that all the Canale Greek Geometric comes from the same centre of manufacture.

^{page 179 note 2} Johansen op. cit. 182.

^{page 180 note 1} The complete absence of Protocorinthian of the Globular Aryballus period and of other contemporary Greek pottery from Greek Locri points to an interval of at least fifty to seventy-five years between the earliest graves at Greek Locri and the earliest Greek Geometric imports to the Canale plateau. The more developed examples of the latter may be as late as c. 700 B.C. I do not think that any of them belong to the seventh century.

^{page 180 note 2} See Johansen op. cit. passim and particularly chapter 6. Compare, however, the fragment from the Fusco cemetery discussed below. Certainly this one fragment, which may or may not have come from a grave, does not invalidate Johansen's main chronological conclusions.

^{page 180 note 3} Orsi's interpretation of this stratum varies. At one time he considered it to have been produced by a levelling of the ground and so of no chronological value. For the Sicels on Ortygia see Thuc. vi. 3. 3. There is no archaeological evidence to support the theory that Syracuse was originally a Euboic or, more specifically, an Eretrian settlement. The so-called literary evidence for this theory (e.g. Strabo 449, Schol. Ap. Rhod. i, 419, Schol. Iliad, ix 557) is contemptible.

^{page 182 note 1} I have not seen this vase nor is it illustrated in N. Sc. 1925.

^{page 182 note 2} For a somewhat sceptical discussion of the ‘Parian’ category see p. 183 note 4 below.

^{page 182 note 3} I am inclined to think that the ‘Argive’ character of these kraters is due to the fact that they are ‘Provincial Protocorinthian,’ that is Syracusan. I have had no opportunity to examine their clay since 1926.

^{page 183 note 1} There is an exact parallel to the decoration in an amphora from Thera, , Dragendorf, , Thera iiGoogle Scholar fig. 107 = fig. 379a. Cf. Rubensohn, , A.M. 1917, 76Google Scholar fig. 83 from Paros, , and Pfuhl, A.M. 1903Google Scholar Pl. xxxiv, no. 3 (Handle-zone decoration); also Louvre A 266 from Thera.

^{page 183 note 2} The two Cretan examples must be used with caution as chronological evidence, for in Crete some Geometric motives last into the early years of the seventh century. On the other hand, these two examples do not seem to belong to the latest phase of the style and are earlier than the contents of the oldest graves from the Gela, cemeteries, e.g. the pithos in M.A. xviiGoogle Scholar, Pl. 5 centre, which Payne has identified as a Cretan import of the early seventh century. Compare also ibid. figs. 409, 410, 412—also identified by Payne as Cretan imports —which come from the upper strata at Bitalemi and are undoubtedly of a date later than the foundation of the colony.

^{page 183 note 3} Payne, Necrocorinthia 5, note 1Google Scholar.

^{page 183 note 4} I have used the term ‘Parian,’ which Buschor has affixed to this class, with a certain amount of misgiving. I confess that Buschor's arguments for Paros (e.g. the presence of numbers of miniature dedicatory vases, which are according to him of this class on the island of Paros) do not seem to me convincing. The fabric is undoubtedly of the general class known to archaeologists as Cycladic and so I should prefer to describe it, with the addition of a non-local label.

Rubensohn and an earlier school named the class Euboic on account of its affinities with Boetian Geometric. Against this view can be urged the known and very distinctively different character of Eretrian Geometric. Nevertheless, I have often been tempted by the presence of this class and its derivatives in Sicily and Etruria to revive the old heresy in my own mind and to think of the class as ‘Chalcidian Geometric.’ The a priori historical reasons for such a label for a class of Geometric well represented in the West are of course obvious. That Ghalcis should share in the material culture and artistic tradition of the Cyclades is exactly what one would expect. Cf. Thuc. iv, 84, 109, Plutarch, Q.G. xxxGoogle Scholar. I have, however, no intention of complicating the issue by suggesting this new name for the class without evidence of the presence of this pottery in large quantities in Euboea, and specifically at Chalcis; and such evidence is entirely lacking.

^{page 184 note 1} The material from the rest of the Greek colonies in Sicily can be dealt with very briefly. I know of no evidence from Naxos, Catana, Leontini, Himera, Megara Hyblaea. Geometric vases from Sicel settlements near Catana, Taormina and Leontini are discussed below.

Zankle. A Geometric vase associated with a globular Protocorinthian aryballus and later pottery is illustrated in N. Sc. 1929, 40 fig. 2. It is impossible to state definitely whether or not this vase is of pre-colonisation date because of the lack of trustworthy literary evidence for the exact date of the Greek colony and because of the lack of archaeological material from the colonial necropolis. The circumstances and place of the find seem to me to indicate that this vase, as well as the globular aryballus, belong to the colonisation rather than to the pre-colonisation period. If this be admitted the date of the foundation of Zankle must be regarded as probably close to that of Syracuse and Megara Hyblaea and certainly not later than the date of the latter colony. The Geometric vase is probably Cycladic, cf. Dugas, Délos xvGoogle Scholar Pls. 3, 4 and No. 29 from Syracuse. Stylistically it belongs to the second half of the eighth century.

Akragas. I have omitted a discussion of the pre-colonisation material from this site for two reasons: 1. Quantities of Greek seventh-century pottery from Akragas are reported by Marconi and Hardcastle, but no examples are illustrated and the material is as yet inaccessible (I have only been able to inspect two sherds of, possibly late seventh-century, Corinthian). 2. The pre-colonisation material at Akragas belongs, apparently, to the seventh century, and in this article I am concerned only with the pre-colonisation contacts which belong to the period before the first great wave of Greek colonisation to the West in the years c. 735 B.C. to c. 690 B.C. Some early Geometric material from a Sicel site in the neighbourhood of Girgenti is discussed below.

Selinus. I know of no pre-colonisation evidence from Selinus. The early seventh-century Daedalic marble lamp, M.A. xxxii plate xxiii 1, must certainly have been brought with them by the colonists. The problem is complicated by the dual literary tradition as to the date of the colony. There is certainly no eighth-century pottery from this site.

^{page 184 note 2} Class ii is rare in the Sicel cemeteries of the Third Siculan period. It is only in the Fourth period (after c. 700 B.C.) that the class becomes at all common.

^{page 185 note 1} It is important to realise the distinction between Class iii and Class iv, though naturally there are some examples of which it is difficult, if not impossible, to say whether they are bad direct copies of Greek originals or true Barbarian vases with a more than ordinarily heavy debt to Greek designs.

^{page 185 note 2} I have sometimes thought that Class iv contained examples with decorative elements which are neither purely Greek nor purely Siculan, nor yet an amalgam of the two, and which might be referred to the Phoenician settlements on the coast of Sicily mentioned by Thucydides (vi 2, 6). This theory might account for a certain resemblance to the Iron Age pottery of Cyprus which is apparent in some of the amphorae which are illustrated in R.M. xiii, e.g. 343Google Scholar figs. 59 and 60 (see Fig. 9). Orsi detects Cypriot influence in the askos No. 51 (ibid. 355 fig. 70) of which he says: ‘Malgrado esso abbia tutta l'apparenza di un vaso importato e Ciprioto io propendo a crederlo di fabrica Sicula imitante un articolo estero.’ However, Payne has convinced me that this vase has no affinities with Cypriot pottery and is a normal example of Class iv. The evidence is far too scanty and indefinite to justify a subdivision of Class iv into Phoenician influence and Greek influence, or indeed to justify the recognition of Phoenician influence as more than a possibility.

^{page 185 note 3} This tomb contained two burials, one of which at least belongs to the colonisation period; for with the Sicel vases were found fragments of Protocorinthian cups of the Globular Aryballus period. It is, of course, probable that both burials belong to that period. R.M. xv 70Google Scholar fig. 11 seems to me a Sicel imitation of a Greek original of colonisation date, but Nos. 35, 36 probably derive from Greek models of an earlier period.

^{page 186 note 1} There is, however, a parallel for the shape (save for the handles) in a Geometric cup from Tiryns, (Tiryns i Pl. xv 10Google Scholar).

^{page 186 note 2} This vase, so far as I know, unpublished, is in the Museo Preistorico in Rome.

^{page 186 note 3} The Knossian vases have two handles. The parallel in decoration is only close for the first two Cava di S. Aloe examples. There is a less close parallel in a bowl from Leprignano in the Villa Giulia.

^{page 187 note 1} From a tomb containing a continuous series of burials from the Protogeometric to the Polychrome style. The vases were found associated with a Geometric burial which contained a large number of Cypriot imports.

^{page 188 note 1} I.e. the style of these vases is not represented in any of the cemeteries of the Greek colonies in Sicily.

^{page 188 note 2} The fragments must be dated entirely on stylistic grounds. Nothing can be deduced from the contents of the tomb.

^{page 189 note 1} If I am right in identifying this vase as derivative from Greek Protogeometric art, it is an interesting and isolated example of Aegean contacts with Sicily in the dark period between the end of the Mycenaean period and the arrival of the first Greek traders in the first half of the eighth century.

^{page 189 note 2} The majority of the Sicel tombs from near Girgenti belong to the Fourth Siculan period.

^{page 190 note 1} This exclusion probably involves the loss of a certain amount of pre-colonisation material, but the residuum of unquestionable evidence is sufficiently impressive in itself.

^{page 190 note 2} E.g. Graves Nos. 1 and 38 in the West Group.

^{page 190 note 3} The general type of this skyphos was certainly in existence before the Globular Aryballus period. Cf. the two kylikes from the Prehellenic native settlement at Cumae M.A. xxii Pl. xvii, nos. 7 and 9. Some of the Finocchito examples stand half-way (stylistically) between the type Johansen op. cit. Pl. ix. 4 and the Cumae examples and are probably genuine pre-colonisation material. I have nevertheless, for safety's sake, excluded them from my list.

^{page 190 note 4} E.g. Graves Nos. 5, 51, 61 and 71 in the West Group.

^{page 191 note 1} I suppose that eighth-century Greek writing may be regarded as suspect by some historians and, since ProfessorCarpenter's, Rhys article A.J.A. 1933, 8Google Scholar, by a few archaeologists. But the Hymettus Geometric inscriptions (Blegen, A.J.A. 1934, 10Google Scholar), if not the Dipylon jug, prove beyond question that Greek writing not only existed in the eighth century but could be used for purely frivolous purposes. See also Stillwell, A.J.A. 1933, 605Google Scholar, eighth-century inscriptions from Corinth (this last piece of evidence seems to me doubtful).

^{page 191 note 2} I am not competent to give a detailed archaeological analysis of the evidence for Greek pre-colonisation contact with the West provided by fibulae from Sicilian and South Italian sites. In general the material falls into three classes, native fibulae, Greek imports, and local imitations of Greek types, and points to the same general conclusion as the pottery evidence. There is, however, so far as I know, less variety in the type of imports.

I must confess that I am satisfied neither with the local nor with the chronological classification of Greek fibulae that is at present in vogue. For example, the finds at Olympia, the shrine of Artemis Orthia at Sparta, and the Heraeum at Perachora seem to indicate that fibulae of the types Nos. 58A, B are not Italiot and Sicilian (see Blinkenberg, Fibules Grecques et Orientales 197 ffGoogle Scholar.) but mainland Greek (Peloponnesian? Corinthian??).

^{page 192 note 1} The archaeological evidence for eighth-century Greek contact with Etruria is both complex and difficult to interpret. I have confined myself exclusively to the evidence of pottery and to a comparatively small though representative selection of that evidence.

^{page 192 note 2} Barbarian copies of this style may, of course, belong to a later date. They are none the less valuable evidence for the importation of eighth-century Greek originals at a time when they were still objects of commerce.

^{page 192 note 3} The terms ‘Italo-Geometric’ and ‘Italian Geometric’ are often loosely used to cover all four classes. I propose to confine ‘Italo-Geometric’ to Class D and to use ‘Italian Geometric’ for native work which shews no sign of Greek influence. See p. 174.

^{page 193 note 1} Gsell's statement that at Vulci ‘Geometric vases’ were found side by side in the same tomb with Orientalising ware (of the late seventh century) is still used by some people as evidence of the impossibility of dating ‘Italian Geometric’ more precisely than ‘ninth to seventh centuries B.C.’ But it is quite clear from an examination of Gsell's burial groups that this so-called ‘Geometric’ which lasts into the late seventh century consists entirely of examples of the Late Protocorinthian and Transitional Linear style (and Italian copies of those styles)—that, in fact, it is not Geometric at all. There is little true Geometric in Gsell (e.g. Gsell Pl. 1, No. 2 from Tomb 75, which, though of purely local shape, shews Greek Geometric influence in its decoration, and is a fair example of Class D, Italo-Geometric), and what little there is has no Orientalising bed-fellows.

^{page 193 note 2} Necrocorinthia p. 4 note 2. He compares K.G.I.B. 116 No. 3 (here Pl. 28, x).

^{page 193 note 3} E.g. N. Sc. 1914, 323 fig. 16Google Scholar, grave 25 (Vetralla); N. Sc. 1914, 334 fig. 25Google Scholar, grave 10 (Vetralla); J.d.I. 1900, 167 fig. 9 (Pitigliano)Google Scholar; and two vases from Chiusi, Montelius op. cit. Pl. 214.

^{page 194 note 1} I am, of course, not concerned with the question of the supposed Lydian origin of the Etruscans.

^{page 194 note 2} Both clay and paint are characteristic of many purely Barbarian vases from the Etruscan cemeteries.

^{page 194 note 3} Necrocorinthia loc. cit.

^{page 194 note 4} I at one time thought that this vase belonged to Class A or B. An examination of it in the Museo Preistorico last August has convinced me that it is Class C.

^{page 195 note 1} I have not seen these vases. My attention was first drawn to their publication in the Notizie degli Scavi by Miss S. Benton.

^{page 195 note 2} Necrocorinthia loc. cit.

^{page 195 note 3} There are several examples of Italo-Protocorinthian of the Globular Aryballus period from Leprignano in the Villa Giulia. From this period onwards the importation of Greek pottery and the copying of Greek originals seems to have been continuous.

^{page 195 note 4} There is, I understand, a large quantity of Geometric material from this site in the Villa Giulia which is as yet unpublished, inaccessible and invisible.

^{page 196 note 1} There are two tomb-groups in the Villa Giulia which must be mentioned here (First and Second rooms in the South wing). Tomb 779 contains a Geometric cup of greenish-yellow clay, brown paint with tinges of purple. Handle zone decorated with vertical lines and herringbone pattern. The vase is undoubtedly Class A. The colour and texture of clay and paint seem to me to indicate Corinth as its place of manufacture. Tomb 785 contains a flat cup of good purified pinkish clay decorated with narrow horizontal bands and a handle zone of concentric circles joined by tangents. Good brown paint. Both shape and drawing are sure and refined. Class A; possibly Corinthian (cf. Johansen op. cit. Pl. 3). With it were two cups of local friable white clay from which all painted decoration has disappeared but whose shapes betray the Greek Geometric original. There is also a vase in the Musée Scheurleer to which Payne has called my attention. No. 70; C.V.A. Musée Scheurleer i, ‘Style Italo-Géometrique,’ ivb–ivc, pl. 1, 1, from ‘Rome.’ This is either a Cretan import or local work closely modelled on a Cretan original.

^{page 196 note 2} That this friable white clay and red paint are local is shewn by their use on vases which betray not a trace of Greek influence, e.g. Villa Giulia Nos. 4503, 1585.

^{page 196 note 3} Falerii is rich in Classes B and C in the period c. 735 to c. 650 B.C., when its imports save for one Rhodian Bird Bowl of early type (Montelius op. cit. Pl. 323, 7) seem to have been almost exclusively Corinthian. To Class B (with a few exceptions which are certainly Class C) I should refer the amphorae and oinochoai which resemble the examples from the Fusco Cemetery at Syracuse as well as many of the vases of the type Johansen op. cit. Pl. ix 5, 6, e.g. Villa Giulia Nos. 5012, 5013, 5014 to Class C. Villa Giulia Nos. 4971, 4742, 4942 to Class B. Contrast Nos. 5006 and 5008, which are Protocorinthian imports.

This period, I think, demands a fivefold classification into the work of—

1. Protocorinthian potters.

2. Cumaean (?) potters.

3. Greek potters at Falerii.

4. Barbarian potters at Falerii imitating Greek work.

5. Barbarian potters at Falerii borrowing a few Greek designs.

It is possible that in the pre-colonisation period some of the vases which I have placed in Classes A and B were made at Cumae, but we know as yet so comparatively little of the character of Cumaean pottery even in the period c. 735 to c. 650 B.C. that it would be more than hazardous to attempt to identify Cumaean pottery in the preceding period.

^{page 197 note 1} The general shape of this vase is given in Randall MacIver op. cit. Pl. xi 12, but the drawing is too small to shew the character of the design in the horizontal handle zone. The bodies and wings of the birds are filled with parallel upright zigzags and the same decoration is used as filling ornament in the panel above and below the birds' bodies.

I can see no trace of the slip reported by Randall MacIver.

^{page 197 note 2} To this general heading we must probably add Campana 18 and Campana 20 (Louvre A 490, 491), which, as they were in the Campana Collection, were almost certainly found in Italy. These two vases are Cycladic. I have omitted a Bronze Age Cycladic Oinochoe of unknown provenance (Pottier, Vases antiques du Louvre Pl. xxix D 5Google Scholar) and a Mycenaean Stirrup Vase (ibid. D 1) found at San Cosimo near Oria in the district of Otranto, as being not strictly relevant as pre-colonisation material.

^{page 198 note 1} No. 114 is illustrated in Pottier op. cit. Pl. xxxii D 114.

^{page 199 note 1} I must repeat once more that the lists I have given are not exhaustive though representative.

^{page 199 note 2} The foundation of Marseilles is dated 600 B.C. by Pseudo-Scymnus 209 ff. quoting Timaeus. Eusebius (Jerome) dates it 598 B.C. (593 B.C. Armenian Version). Athenaeus 576 A citing Aristotle indicates that the Phocaeans had traded with the country before founding the colony. The much-discussed Thucydides passage i 13, 6 implies that the colony received an additional body of settlers after the fall of Phocaea. This seems certain from Strabo 252 citing Antiochus (the emendation which converts Marseilles into Alalia can only be described as wanton). Thucydides does not imply that Marseilles was founded in the second half of the sixth century.

The c. 600 B.C. date is consistent with the archaeological evidence from the site (see Jacobsthal op. cit. passim); there is very little pottery earlier than the Middle Corinthian Period. For pre-colonisation imports of the seventh century to the south of France see Vasseur Pls. v 12, vi 10, x 2, xi 9 (?); Jacobsthal op. cit. fig. 40b (Hyères). Corinthian; Vasseur op. cit. Pl. vi 12 (Cretan). Possibly also some of the Rhodo-Milesian in Vasseur op. cit. Pl. v.

^{page 200 note 1} I should like to take this opportunity of expressing my thanks to Professor Jacobsthal for his great kindness and assistance to me at Oxford in the spring of 1934.

^{page 200 note 2} See note 3 below.

^{page 200 note 3} Cumae is dated traditionally to the middle of the eleventh century B.C. So far as this can be taken at all seriously, it must be regarded as recording some westward movement of the Greeks in the period of migrations. It certainly cannot be regarded as the date of the Greek colony. Archaeologically the colony is dated earlier than Syracuse by the presence of Protocorinthian pottery of the Globular Aryballus period in greater quantities than at Syracuse, where it only occurs in the earliest graves. It is dated after c. 800 B.C. by the native settlement which then occupied the site and which imported two Greek Geometric vases which, stylistically, cannot be much earlier than that date. A date c. 775–c. 750 B.C. seems to be indicated, and this agrees well enough with Strabo's statement that Cumae was the oldest Greek colony in Sicily and Italy (Strabo 243). The first Greek settlements on the islands in the bay of Naples (Strabo 247) were possibly earlier, as Livy (viii 22) indicates. There is no archaeological evidence for this unless we assume, what is probable enough, that the presence of Greek pottery and intensive Greek artistic influence in Etruria of a date stylistically earlier than the contents of the earliest graves at Cumae demands a Greek centre of distribution in the neighbourhood of an earlier date than that of Cumae.

^{page 200 note 4} After c. 735 B.C.–c. 690 B.C. and throughout the seventh century her Greek imports have the same character as those of all cities, Greek and Barbarian, in the western Mediterranean during that period; that is to say, there is a marked preponderance of the pottery of Corinth.

^{page 201 note 1} Pseudo-Scymnus 238. Strabo 243, 247. Dion. Hal. VII 3. Livy viii 22. Thuc. vi 4. The fact that Thucydides only mentions Chalcis is not as serious as the uncritical exponents of ‘Quellenkritik’ would like us to think. The reference is only incidental and does not describe the foundation. Further, there are reasons for believing that Cumae became the exclusive property of Chalcis in the period c. 735–c. 690 B.C. (see below, p. 206).

I can see no real reason why we should suppose that the tradition of the Aeolic element at Cumae rests on nothing better than the local patriotism of Ephorus and so conclude that the name of the colony was derived from Euboic Cyme. Is it possible (assuming that the Theogony is a work of Hesiod) that Hesiod's western geography was derived from his father, the merchant of Aeolic Cyme?

^{page 201 note 2} These inscriptions consist of–1. An incised inscription on a Protocorinthian Ovoid aryballus (c. 675 B.C.) in the British Museum (Johansen op. cit. Pl. xv 5). 2. An incised inscription on the base of a Protocorinthian conical oinochoe (M.A. xxii Pl. 27; the beginnings of two alphabets, one of which includes the distinctive Corinthian B).

These inscriptions form the basis for the study of the early seventh-century alphabet of Cumae. To class them as Protochalcidian is to beg the whole question. Many letters differ in several important respects from the sixth-century alphabet of Chalcis as known to us from the ‘Chalcidian’ pottery of that period, and the differences are not merely the differences between seventh- and sixth-century letters of the same alphabet.

^{page 201 note 3} It is, of course, possible that some of these examples from Cumae belong to the period after c. 735 B.C. The criterion of date that I have applied is, in fact, purely stylistic and in itself does not support a wholesale reference of the Cretan Geometric to the period before c. 735 B.C. (see p. 183 note 2). It would, I think, be safe to say that none of the examples discussed above are much later than c. 690 and that many are earlier than c. 735 B.C.

^{page 202 note 1} Some of the Corinthian of the Globular Aryballus period must be earlier than the date of the foundation of Syracuse, but it would be quite useless to attempt to make even a tentative list.

^{page 203 note 1} The majority of the imports were undoubtedly of Cycladic type, but Cycladic, like East Greek, is a term which covers the products of a number of Greek states. I have little doubt that in time we shall be able to classify and localise the various fabrics within this general class; but at present it is impossible to do so with any certainty.

^{page 204 note 1} Throughout the period c. 690 to c. 600 B.C. I know of but one example of Cycladic pottery from any western site. Cretan exports are confined to Gela and Veii (this last is an inference from the earliest Etruscan mural paintings at Veii; see Rumpf Wandmalereien in Veii) and possibly Cumae (see above, p. 201). East Greek fabrics are rare till the last quarter of the century and in the period c. 690 to c. 630 are only found at Gela, Syracuse and Megara Hyblaea, with scattered finds in Italy and Etruria. There is a little, very little, seventh-century Laconian from Tarentum and a little which probably comes from Etruria. There is no Attic till the last quarter of the century. These exceptions only serve to mark the contrast with the great quantities of Corinthian which are found on every site and often in themselves point to a connection between colonisation and commerce. For example, the Cretan and Rhodian pottery at Gela is only what we should expect of a joint colony of Crete and Rhodes, and the same might be said of the Laconian at Tarentum. Indeed at Tarentum one is only surprised at the rarity of this fabric in the seventh century. Further, at Gela there is at least as much Corinthian of the seventh century as of the combined products of Crete and Rhodes, and at Tarentum there is that overwhelming preponderance of Corinthian which we find on every other western site in this period. This Corinthian domination of the western pottery trade is shewn not merely by the vast numbers of the Corinthian imports but is also reflected in the local pottery of the period. Cumaean, Italo-Corinthian, Etrusco-Corinthian in all their varieties and classes, are one and all imitations of the pottery of Corinth, and the Corinthian model can often be detected in the more crude work of the Siculan potters of the seventh century. For resident Corinthian potters in Etruria and the Demaratus tradition see the literary references at the end of note 1 p. 172, see also note 3 p. 196 for the seventh-century archaeological evidence from Falerii and its neighbourhood.

^{page 204 note 2} For the original debt of Protocorinthian pottery to Cretan models see Payne, Necrocorinthia 5Google Scholar, Johansen op. cit. 45, Payne, Protokorinthische Vasenmalerei 11Google Scholar.

^{page 205 note 1} See Payne, Necrocorinthia 186Google Scholar. Johansen's lists (with a few additions from the reports of excavations published since Les Vases Sicyoniens), which cover the period c. 800 to c. 640 B.C., give us the following:—Rhodes (Lindos, Kamiros, Vroulia and elsewhere), Kalymnos, Crete (Praesos, Afrati, Knossos, Dikte, Arkades and elsewhere), Ephesus, ‘Smyrna,’ Hissarlik, Lemnos, Delos, Thera, Cyprus (Amathus, Limassol).

^{page 205 note 2} I am indebted for this information to Professor Jacobsthal.

^{page 205 note 3} Thuc. vi 3, 2. Strabo 380. Athenaeus iv 167dGoogle Scholar; date from calculations based on Thuc. vi 4. The so-called literary evidence for a previous Euboic settlement is, I think, most unsatisfactory, e.g. Strabo 449, Schol. Apollon. Rhod. i 419Google Scholar, Schol. Iliad ix 557Google Scholar.

^{page 205 note 4} Plutarch, Quaestiones Graecae xiGoogle Scholar. Date, , Strabo, 262Google Scholar. The synchronism with Syracuse is supported by Schol. Apollon. Rhod. iv 1212Google Scholar and Plutarch, Amat. Nan. 772Google Scholar. Doubts have recently been cast on an Eretrian colony in Corcyra by Halliday in his edition of the Quaestiones Graecae. Against this the following arguments may be urged:

1. A Euboic tradition in Corcyra is supported by her sixth-century coinage, which alone of those of Corinth's North-West colonies does not bear the Corinthian Pegasus but the Euboic device of the cow suckling her calf.

2. The suggestion that this device was Macedonian and acquired by Corcyra via the Egnatian Way route has been made improbable by the discovery on the island of Zacynthus of what is apparently a seventh-century votive disc bearing the cow and calf. This seems to shew that the device was in existence in the Ionian Islands before the Macedonian coinage. See B.S.A. xxxii 214Google Scholar, but read Corcyrean for Syracusan.

3. The standard of the coinage of Corcyra is the same as that of the earliest coinages of the Euboic colonies in Sicily, Naxos, Zancle, Himera. That is to say, it is Proto-Euboic (I am indebted for this last piece of information to Mr. E. S. G. Robinson; he cannot, however, be held responsible for the use I have made of it).

^{page 206 note 1} The Aethiops story, which rests on the almost contemporary evidence of Archilochus, shews, I think, that the division of the land at Syracuse had been made before the expedition started. This can, of course, be used to maintain that the colony was purely agricultural, but it also shews how deliberately the expedition had been planned.

^{page 206 note 2} Epitaph of Orsippus, , Pausanias i 44, 1Google Scholar. Plutarch, Quaestiones Graecae xviiGoogle Scholar seems to me probably to refer to the same struggle. See, however, Halliday's edition ad. loc.

^{page 206 note 3} Thuc. vi 3, Steph. Byz. s.v. Chalcis. Ps.-Scymnus 270. Strabo 267, cf. Strabo 262. Date from calculations based on Thuc. vi 4 (Eusebius 737 B.C.).

^{page 206 note 4} Thuc. vi 3–4. Strabo 257. Ps.-Scymnus 283. Pausanias iv 23, 7Google Scholar.

^{page 206 note 5} Thuc. vi 4. Polyaenus v 5, 1–11Google Scholar.

^{page 206 note 6} Chalcis and Eretria were presumably friends at the date of the foundation of Cumae. See above, p. 201. Strabo 448 refers their enmity to a quarrel over the Lelantine plain. I do not propose to enter into a discussion of the date of the so-called Lelantine War nor into its possible connection with the colonisation of the West.

^{page 207 note 1} I do not propose to discuss the evidence for Chalcis' share in the western trade in the seventh century. No Chalcidian pottery is known to exist of earlier date than the sixth century, and even this fabric, as Smith has shewn (California Publications in Classical Archaeology vol. i. no. 3Google Scholar), is very possibly a product of a Chalcidian colony in the West. The fame of Chalcidian swords is, however, as old as Alcaeus (Diehl 54, 6), and it is possible enough that her western exports consisted of metal-work which has not survived or which cannot be attributed to her on archaeological grounds. The main evidence for the fame of Chalcidian bronze is still literary rather than archaeological; see, however, Neugebauer, R.M. 1923–1924Google Scholar.

^{page 207 note 2} There is, I think, reason for believing that Corinth retained a large portion of the westward commerce till a much later date, but her trade from c. 575–550 B.C. onwards was, I think, mainly a carrying trade. I hope to discuss this matter in the near future.

^{page 207 note 3} Payne, Necrocorinthia 25, 184, 187Google Scholar.

^{page 207 note 4} Strabo 801 and Herodotus ii 178Google Scholar.

^{page 207 note 5} Herodotus i 20 and v 92Google Scholar. Aristotle, Politics 1284 A and 1311 AGoogle Scholar. Perhaps also Frontinus iii 9, 17Google Scholar. For the particular debt of Corinthian art to that of the Near East during the last quarter of the seventh century see Payne, Necrocorinthia 43–54Google Scholar.

^{page 207 note 6} Herodotus vi 21Google Scholar and Athenaeus 519 B seem to imply a commercial understanding between Miletus and Sybaris (their friendship is often referred to the eighth century, but there is not a shred of evidence for such an early date). It can hardly be earlier than the last quarter of the seventh century, when we have the first concrete archaeological evidence of extensive East Greek commercial interests in the West both in the Greek cities and in Etruria. If Miletus had a commercial understanding with Sybaris c. 600 B.C., it is more than probable that her friendship with Corinth of the same date likewise had a commercial aspect.

^{page 208 note 1} I am greatly indebted to Prof. Beazley in the archaeological section of this article, but neither he nor Mr. Payne can be held responsible for any errors there may be in my attributions. To Prof. Myres I am indebted for a suggestion as to the character of the decoration of Sicel pottery, and to Mr. Wade-Gery and Mr. R. H. Dundas for advice and encouragement in the study of the historical problems of Greek colonisation in the West. Finally I should like to thank Mr. T. J. Dunbabin for much valuable information on the subject of Sicel pottery and Dr. J. G. Milne for help with my plates.