¹ The most recent review of the SOS type, by Strom, gives a fuller bibliography than I have selected here. The treatments by Brann, I.c. and Agora viii 32–3, Strøm, Villard, and Young are all sound and I have only a few chronological modifications to add to them. B. B. Shefton will be dealing with questions of the distribution of the amphorae and J-P. Descouedres the material from Eretria and associated matters. Excavation reports which will be of importance, especially for dating, are expected for Chalkis, Kition, Pithekoussai, Metaponto, Policoro and Kamarina; only preliminary notices of these finds could be cited below.

² I am grateful to Dr. R. Seeley for having the piece examined.

³ The vase was placed in a tank and water introduced both outside and inside to relieve pressure on the walls. Although this procedure would have kept to a minimum the amount of water absorbed by the walls from the inside we should none die less make some allowance for this in thinking of the capacity of the vase.

⁴ Despite this added frill to the decoration I have little doubt that this fragment comes from an SOS amphora; profile, size, and the rest of the decoration are sufficient to demonstrate that. It is unfortunate that it has no useful stratigraphie context.

⁵ I owe the profile drawings of Ashmolean 1954.481¹ and ² to Mrs. Pat Clarke.

⁶ ADelt xxvi (1971) B 252. I have not seen the one plece illustrated there, with Wch decoration, pl. 227a upper row, centre. No vases have yet been mended up sufficiently for the body shape to be assessed. The dating of the dumps, together with the presentation of the evidence for their interpretation as potters' waste tips, must be left to the excavators, but the vast majority of the material cannot date far from c. 700. For a description of Chalcidian fabric of this period see Boardman, , BSA lii (1957) 2Google Scholar, although I cannot agree with his words ‘rather soft in the break’, since the hardness of firing of these vases is immediately apparent when drilling.

⁷ It was found with an early rosette bowl; for the dating see Hayes, , Tocra i 46 n. 3.Google Scholar

⁸ We are most grateful to the directors of GSE and M. Vlasaki for their very generous permission to include the sherd in this study. Knowledge of it came too late for full assimilation into the text (especially in section 6, on the type and origins of the decoration of Chalcidian SOS amphorae). Analysis has given the following result:

Such a composition fits that of local LM IIIc ware quite well, but the clear difference in the colour of the fabric—it is not the buff of the local ware—means that die plece should be found a home elsewhere, and the composition is very close to the Chalkis range.

⁹ The publication of most of the amphorae from the necropolis is forthcoming in G. Buchner and D. Ridgway, Plthekoussai i. The fullest of the fleeting references made to the Mazzola and scarico material to date is Buchner, Atti xi Convegno Magna Grecia 366.

¹⁰ Much has been written recently about the dating of the vase—see the references cited. It is a small misshapen thing, rather worse potted than the published Eleusis amphora; I would not care to put a more precise date on it than 725–690, probably before 700.

¹¹ In Populi Anellenici loc. cit. the sherd is described as ‘frammento di argilla acroma’, but the photograph in Arch. Class, clearly shows the remains of glaze. We may note the mention of local imitations at Metapontum, BdA (1976) 47.Google Scholar

¹² We may note the local imitation of seventh-century date with well-spaced Sl,O,Sl,O,Sl on the neck, MEFR lxvii (1955) pl. iiia. The Attic vase mentioned in AJA lxx (1966) 361 is not yet published.

¹³ There is also exhibited in Syracuse a half-size SOS from Giardini tomb 75 (plan of the excavations, NSc (1949) 201); decoration OI,OI, and single band on shoulder. It was found with fragments of an Attic BF volute-krater of c. 535–525.

¹⁴ A poor photograph of one fragment in RM xxii (1907) 133, fig. 21, cxxviii. The plece can only be dated by the early type of foot. One fragment is embedded in a lump of metal together with a bucchero kylix; Pareti noted this and took it as part of the burial in the right niche, which Strom dates c. 625. She takes up the matter in n. 530, but does not bring the SOS fragment into consideration. One may speculate how and when the kylix and a single sherd of the amphora became engulfed in the molten metal, but the variety of possible answers precludes any sure reconstruction.

¹⁵ The association of the amphora with the tomb is not made fully clear in the display in the museum. The skyphos is illustrated in Lerici, Nuove Testimonianze dell'Arte e della Civiltà Etrusca 34 (with wrong date).

¹⁶ I am grateful to Brian Shefton for this reference.

¹⁷ There is much uncharted territory here and the analytical compass we offer can hardly be said to be adequately boxed. For most recent bibliography see Tocra ii 62, with references to the significant material from Histria, Marseilles, and Athens; further examples have been cited above, Kerameikos and Phaleron. An intact example akin to the best preserved Marseilles fragment, Villard, La Céramique grecque de Marseilles pl. 27, I, and not far from the plece of uncertain origin, Thera ii fig. 221, is published by Lazarov, , Izvestia Varna xxvi (1975) 128–9.Google Scholar It may be pertinent to add here die neck with a probably Attic inscription from Salamis, Salamis ii 231 and 275–7. The point is made in Agora xii 192 that the type develops little at Athens in the sixth century—a point which should be taken into consideration when trying to date the Wappenmünzen with amphorae on the obverse by typological criteria (Kraay, Archaic and Classical Greek Coins 56 ff.).

¹⁸ On this criterion I have included in the lists above numerous pleces which do not demonstrably belong to SOS amphorae. Further fragments which should be taken into consideration as being on the SOS/à la brosse borderline (none of which I have seen) are: Stucchi, Cirene 1957–1966 166, fig. 188 (inscribed); Tocra ii 2265 (presumably upside down in die profile drawing; Ponsich, Récherches archéologiques à Tanger et dans sa région 185 (body sherd); the fragments from Marseilles taken as Attic rather than Ionian by Villard op. cit. 49. From die description and photographs the Marseilles fragments seem no less Attic than some of the pleces from Kition included here.

¹⁹ Reproducibility tests have indicated that the analytical precision with which these three elements may be measured is associated with standard deviations of 6, 14, and 10 percent respectively.

²⁰ Boardman, and Schweizer, , BSA lxviii (1973) 270–1Google Scholar; Schweizer, apud Prag et al., Archaeometry xvi (1974) 168–70Google Scholar; using X-ray fluorescence analysis, Stern and Descoeudres, Archaeometry xix (1977) 73 ff.

²¹ Boardman and Schweizer, loc. cit. 274, diagram X.

²² There are aspects of 19 which are not perhaps purely Attic: the clay is a full orange and the fabric very soft, while the decoration is hastily painted and of a rare type. As noted above, p. 122, the clay of 91 and 93 is not surely Attic and analysis underlines the doubt without ruling out an Athenian provenance.

²³ The difficulties posed by 92 should not be ignored and perhaps deserve more than a footnote. From all external evidence the plece seemed Attic enough to be included in the main catalogue and not the appendix on ‘à la brosse’ amphorae. The original sample, taken from the foot which was not published with the vase, gave results which were clearly not Attic; we decided to test a sample from the body of the vase, which was made available through the good offices of Dr. Karageorghis and Professor Buchholz. The result was:

amply confirming a non-Attic origin. Yet while the readings for the two samples are comparatively close in some of the elements, including the more diagnostic Mg and Ni, there are marked discrepancies in Al, Mn and to a lesser extent Cr. The Al variation may be due to the fact that the first sample 92 was drilled and the second chipped and ground, while variation in readings for Mn may be expected in the fabric of a large vase.

²⁴ I refer here mainly to David Ridgway's unpublished attempts to distinguish between local and Euboean fabrics at Plthekoussai (Papers in Italian Archaeology, I, The Lancaster Seminar (1978) 123). Elsewhere in southern Italy some deviation from the Chalcidian range is shown in Ca content (Boardman and Schweizer, loc. cit. 272), and high Ca is apparent in 34, presumably made near Policoro, though not in 32, presumably made near Sybaris.

²⁵ It is to be hoped that positive results will emerge from a programme of thin-sectioning of material from Plthekoussai, being carried out by G. Buchner and D. Ridgway at the British School at Rome.

²⁶ Such inclusions have been mentioned more frequently of late; Kerameikos vi 2, 144, Eretria v 22. We can single out from the mass of vases which display red inclusions a trio in the National Museum in Athens: 221, the early black-figure Siren olpe (Shefton-Arias-Hirmer pl. 21); 2226, late Protoattic sherd with fragmentary inscription (BSA xxxv (1934–5) pl. 54. f; Beazley, , AJA xxxix (1935) 475, 1)CrossRefGoogle Scholar; 18772, fragment of plaque from Aegina, c. 700 (Jeffery, LSAG no, pl. 16, 1; the drawing is misleading since it tends to dissimulate the scar by the crucial antepenultimate letter—pi or gamma; the start of a downstroke, giving a pi, seems just visible).

²⁷ Descoeudres mentions these inclusions with respect to three of his sub-groups of Eretrian fabric, 3d, 6c and to a lesser extent 9 (Eretria v 21–2); no pleces of the first two subgroups are included in Eretria v and in Archaeometry xix (1977) l.c. only one of each is analysed, 26 and 76; purely on grounds of analysis 76 could well be Attic. However, there is on display in Eretria Museum a sherd from a large ‘Dipylon’ type amphora, from Papadimitriou's excavations, showing a warrior and (?) charioteer painted in the typlcal Eretrian white-on-glaze technique, and the fabric includes several red stones.

It is not easy to incorporate the results of the Eretrian analyses here, principally because the two elements which we have found most diagnostic, Cr and Ni, were not measured. Some distinction between Eretrian and Chal-cidian fabric in the concentration of those elements that both programmes have in common is observable, but the ranges are not widely separated. Mg content is the most significant available with the mean figures of: Eretria 3·2; Athens 4·2 (SOS), 5·2 (Stern/Descoeudres); Chalkis 1·6 (SOS), 2·4 (Boardman/Schweizer). On available evidence none of the problem pleces from Plthekoussai seems to have an Eretrian origin.

²⁸ The various oplnions held concerning the marks are summarized by Bravo, Hoz, Madr. Mitt, xi (1970) 104 ff.Google Scholar and esp. n. 5.

²⁹ They are: 1, 2, 21, 80, Syracuse 21210 (Gela), Villa Giulia (Cerveteri tomb 6, 12) and Louvre D33, D34, and D35.

³⁰ Unpublished sherds of an amphora of the LG I period from Plthekoussai (necropolis sporadico) of the same fabric as the vase cited in n. 38 below, and probably Leukandi, Preliminary Report fig. 78, which is in turn of similar dark brick-red coarse clay. See now PdP 33 (1978) 136.

³¹ For examples in Attica and some from elsewhere see Hansen, Glotta liv (1976) 31–2 (with regard to his remarks concerning Nestor's cup, it should be noted that simple is found on the Plthekoussai sherds mentioned in the previous note). While is used occasionally in Euboea, Boeotia, and Sicily, it is far rarer in Ionia; to add to the examples cited by Hansen, there are six or seven to place beside the overwhelming majority of at Naukratis, one on a Chiot chalice from Aegina, Furtwängler, Aegina 456, no. 244, and LSAG 343, 29 from Miletus, and 372, 61C, from Borysthenes island. is found in Attica, but rarely: the Burgon amphora, sherds from the Acropolis, Graef-Langlotz ii 1369, 1370 and Agora xxi F 63 and F 65.

³² The one Attic companion cited by Pape-Bennseler suffers from being a variant reading at Dem. xxi 182.

³³ Few examples are given by Pape-Bennseler, but they are well scattered.

³⁴ Perhaps he is rather Charoplas, who has a namesake, Charoples, in the early fifth century at Styra, , IG xii 9, 56 (432).Google Scholar Charoplnos is a sixth-century Parian, , LSAG 103, 4Google Scholar, while Charops can be found in Athens in the fourth century, Bull. Ep. (1950) 72a.

³⁵ For the north Aegean, but not necessarily non-Greek origins of the name see Bull. Ep. (1974) 142.

³⁶ The very frequency of names in Pet- in Egypt makes a Greek explanation unlikely. I would not wish to advance the possibility of Pet(rie), and would like to take the the opportunity of at least querying the possibility of Bil(iotti) on the cup from Rhodes which I suggested in BSA lxx (1975) 164. Names in … are reviewed by Robert, , Bull. Ep. (1974) 142Google Scholar and include an example from lasos. In addition, I have very rarely come across any short graffiti which seem of doubtful authenticity.

³⁷ On early acrophonic numerals see PdP xxx (1975) 365–6. Delta is also found on Ashmolean 1956.507 and Salamis tomb 10, 15.

³⁸ Garbini, , PdP 33 (1978) 143 ff.Google Scholar We also find one mark inscribed in part over another on 92 (FIG. 7(b)); I would read an original graffito ΦΕ (the following strokes are very faint, possibly accidental), over which has been cut part of the alpha of the retrograde mark,

³⁹ It should be noted, however, that repetitions are not common among mercantile diplnti on Corinthian and early Attic BF vases; Greece and Rome xxi (1974) 141 and BSA lxx (1975) 149.

⁴⁰ See n. 38 for the probability that it was not inscribed early in the career of the vase.

⁴¹ The word does occur in a mercantile context, but clearly with reference to plain, i.e. unribbed, black-glaze vases, probably of the fourth century, Hackl, Münchener Archäologische Studien dem Andenken Adolf Furtwänglers gewidmet 56, no. 607.

⁴² Epsilon after a vowel in the first declension is found in Tataie's inscription on the aryballos from Cumae, , LSAG 240, 3.Google Scholar On the other hand on a fragment of a local (?) skyphos from Plthekoussai, Mazzola 70–C–1050, is the snatch to all appearances in the local script. From this evidence of the earliest period there would seem to appear a non liquet about the ‘proper’ Euboean usage.

⁴³ It would be controversial to introduce such a ‘red’ xi to Megara Hyblaea (or any neighbouring state) however; the ‘blueness’ of the Megarian script has been champloned, with substantial new evidence, by Manni Plraino, xxi (1975) 121 ff.

⁴⁴ See Young, Brann, and Villard, BAM 11. cc. Young does not begin the series early enough, for the Plthekoussai and Kerameikos evidence points to its inception in the Attic LG lb period; Brann terminates the series too early unless we interpret very broadly her words (Agora viii 32) ‘except for a few late stragglers this series ceases at the end of the seventh century’; the material from Vulci and Kamarina at least belies this. Villard too assumes that the ‘à la brosse’ type replaced the SOS around 600 rather than overlapplng it during the following years.

⁴⁵ The evolution of the lekythos at Athens is readily judged from Haspels, Attic Black-figure Lekythoi pls. 1–10. Vallet has argued for the use of the SOS as an oil container in a fundamental article, Hommages à Grenier 1558 ff.

⁴⁶ Such a convex bulge to the neck is typlcal of Chiot wine amphora of the sixth and later centuries, BSA xlix (1954) 169, V. Grace, Amphoras and the Ancient Wine Trade fig. 44, Histria ii pls. 52–3. Bulgy necks are rare earlier and one may ponder the possible connection of the Chalcidian amphorae with Cypro-archaic I oinochoai in this respect, e.g. SCE iv 2 fig. XXIX 13, XXXIV 16, Ant. K. x (1967) pl. 38,1.

⁴⁷ For profiles of Attic neck-amphorae of the seventh century see G. Mylonas, 9–16

⁴⁸ The early Panathenaic amphorae have round handles (with which we may compare 91) and the neck and lip profile is far closer the ‘à la brosse’ than late SOS type; see AJA xlii (1938) 495 ff. There is no observable difference in size between late SOS and early ‘à la brosse’ amphorae.

It is a nice question whether the SOS was still being made at Athens at the time Kleitias painted the François vase, even nicer whether he intended it as a wine jar. On the first question it would be best to await the publication of the Kamarina material, on the second we are faced by the alleged Solonian prohibition of Attic wine exports. If Dionysos is carrying oil do we have a precocious use of the ‘political’ use of mythology at Athens by vase-painters, champloned by Boardman (RA (1972) 57 ff.; J HS xcv (1975) I ff.)? See further p. 140.

⁴⁹ The diameters of 50 cm. or more that I have available are for: 2, 27, 68, 72, 74, Agora P7185, Salamis tomb 10, 15 and Louvre D34.

⁵⁰ I have applied several formulae to 2 and others of the nineteen vases, all based on the kotyle size of 273 cc. used by Lang, Agora × 44, which in turn is very close to the chous size used by Grace, Hesperia xl (1971) 85. The formula V = r(internal)² × body height (i.e. less foot and neck) gives a range from 28,500 to 69,000 cc, or 104 to 253 kotylai for the nineteen vases; the two extreme examples stand rather apart (Louvre D39 and 2), but discounting them the average capacity using this formula is 191 kotylai. A simpler formula is V = 14 (more or less the neck diameter of most amphorae) × D × height less foot; this gives virtually the same result for Louvre D39 and only 46,155 cc for 2, with an average without these two of 146 kotylai. The formula adopted by Lang of height less foot (but note that the ² is omitted, ibid. 59; correctly given in Sov. Arch. (1976) 3, 93) gives 72,370 cc, while the formula preferred in Sov. Arch, ibid., height less foot × ((D+neck D))² gives 53,525. One further method of calculating the capacity of 2 which was tried was to cut out of cardboard a half-section of the vase (internal); the centre of gravity of the section was found and the distance from it to the vertical axis used as r in the formula V = area of half-section × 2πr; this gave 1,117 × 2πr, = 1,117 × 2π × 9–85 = 69,095 cc = 253 kotylai. It is clear that the first and last of the methods gives the best results for 2, especially when taking into consideration the fact that I may have overestimated the internal measurements of the vase; however, we still have to make allowance for the fact that a proportion, perhaps up to two litres of the 63,750 cc of water taken to fill 2 will have been absorbed by the walls. The most striking result is that none of the formulae are tolerably close to the actual measurement.

⁵¹ The empty tare of 17 kg. (or a little less—the vase was still a little damp when weighed) plus 61–2 litres of oil at 920 gr. per litre.

⁵² For capacities of Panathenaic amphorae see Edwards apud Agora × 39, n. 9 and CVA Metropolitan Museum 3 32 ff.

⁵³ 64×44×14 = 39,424cc = 144·4 kotylai. Such calculations are of course based on near complete uncertainty over the size of the Attic foot in the eighth to seventh century (see further, section 7); I only intend the equations 64 cm. = 2 feet and 44 cm. = 22 fingers to be roughly approximate. No doubt other hypotheses could be shown to be acceptable if a different foot is used. Here 44 is probably excessive for the average internal diameter, but 64 too little for the average internal height, since the SOS has a deep base (Grace, V., Hesperia xl (1971) 72.Google Scholar

⁵⁴ Although the neck diameter scarcely seems of great relevance to such calculations it is used in formulae for calculating capacities propounded in antiquity (Hero, Stereometrica 21–5); for its application see also n. 50. If the potters were using some such rule of thumb method it does not of course imply any considerable mathematical acumen on their part, although it does suggest that a value for π was known in Athens at this date. It may have been a further aspect of the orientalizing period, although the computation involving 22 and 7 (or 11 and 14) does not seem attested earlier in the Near East; in the Egyptian Rhind papyrus we find (O. Neugebauer, Vorlesungen über Geschichte der Antiken Mathematischen Wissenschaften i 122 ff.).

⁵⁵ On one of the Phaleron vases the neck ornament is said to have been incised, ADelt. ii (1916) 29, tomb 37.

⁵⁶ Bands above the panel are rare on Attic amphorae: 55, Eretria inv. 4738a and b, and Syracuse, Arch. Sic. S-O no. 280 are the only assured examples known to me. This feature seems to be the only one to suggest a non-Attic origin for the Eretria sherds, since both the other pleces are also anomalous, 55 in being the only Attic plece in the Chalkis deposit and having an unusual form of O decoration, and the Syracuse sherd with its elaborate Ob decoration and rather squat profile. Three lines below the panel are found on 45; in this respect and in the forms of O and T used 45 is close to the two LG I vases cited at the end of n. 59.

⁵⁷ Agora P7185 is an exception.

⁵⁸ Desplte the ubiquity of concentric circles in Late Geometric ornament, the spacing of the Och type is not readily parallelled. Gjerstad rightly questions any close relationship between the Cypriot and Greek usage of concentric circle ornament (SCE iv 2, 301), but the closest parallels for this particular Chalcidian variety are found in the sets of mainly vertical circles on Cypro-geometric III and Cypro-archaic I flasks and oenochoai, e.g. SCE iv 2 figs. XIX, XXI–XXIII, XXV and more especially the Bichrome Red I vases fig. XLI; here I follow the dating for the start of CA I around 740 B.C. proposed by A. Dimitriou at the Mycenaean Seminar in London, June 1977 (see now AA 1978 12 ff.). Cypriot contacts with Euboea at this time are discussed in Ant. K. x (1967) 133 ff., to which we can add Coldstream's reallocation of the Cesnola group to Euboea, (BICS xviii (1971) 1 ff.Google Scholar) although it has not been welcomed by all those working on particular Geometric schools (a selection of reactions: Buchner, Atti xi Convegno Magna Grecia 371–2; Walter-Karydi, , AA (1972) 408ff.Google Scholar; Descouedres, , Eretria v 57 n. 344.Google Scholar)

An alternative explanation would be to derive Och from Wch in splte of the overwhelming number of Och at present known; much depends on how short-lived the potters’ dumps at Chalkis were and whether the sherd 10 is demon-strably earlier than them. If the wheel was the original motif it should have some more than purely decorative significance; dare one connect it with the type on early Chalcidian coins?

⁵⁹ Taking the preponderance of T at Plthekoussai and the general statement about the Phaleron material, AE (1911) 248, into account, the ratio of T to O at Athens in the early period seems roughly even; O on early pleces: 7, 8, 45, 75, 78, Phaleron tomb 47. More complex O motifs are found on some Attic LG I amphorae with neck decoration, e.g. ADelt. xxviii (1973) A pl. 16α–β, 26ζ, Dipylon grave XIII (the grave of the ivories), Jdl xiv (1899) 191 fig. 48, and Athens NM 12895 from the Rousopoulos collection, an interesting vase since it has most characteristics of an SOS of the early period save its short body with striped decoration; the neck is ridged and the handles rounded, striped; the neck decoration is Ob,Tb,Ob, divided and framed by single long zig-zags.

⁶⁰ e.g. ADelt. xxviii (1973) A pls. 3a, 8a, and 21a; Dipylon grave X, Jdl loc. cit. fig. 49. See also Young 211.

⁶¹ Eight-spoked wheels at Athens, Boardman, , JHS lxxxvii (1967) 3Google Scholar; it is more frequent unconnected to a chariot, as a shield blazon (Tolle, , Antike Welt v (1974) 3Google Scholar, 29, fig. 10 various) and in particular as the core decoration in the LG II Concentric Circle group (GGP 74–5). The lack of significance in the normal SOS decoration is stated by Brann on F41.

⁶² The many-stroke sigma is found sporadically through out the Greek world in the seventh century, but only persists at Sparta. Four-bar sigma is a common enough alternative to three-bar at Athens in that century but becomes something of a rarity after. See LSAG 34 and 67, BSA lxviii (1973) 184 n. II and Hesperia suppl. xvi 44. 82 and Phaleron tomb 47 show that a definite four-bar version can be found on quite early SOS.

⁶³ The horizontal wavy line is too common a motif in the LG period for us to pln down its origin here.

⁶⁴ The lozenge, which is the equivalent of T as filling ornament, dies out during the period of the Chimaera-and-Nettos painter; see Kubier, Altattische Malerei figs. 14, 18, and 72.

⁶⁵ Ta×3 on the amphora ADelt. xxiii (1968) B pl. 28; T × 3 on Kerameikos inv. 3249, AM lxxxi (1966) Beil. 65,3; Tb on Athens NM 12895 (n. 59); T with triple outline and flanked by birds on ADelt. xxviii (1973) A pl. 24a.

⁶⁶ ADelt. xxviii (1973) A pl. 21 a; it only differs from the SOS in the filling ornament and its wholly glazed body.

⁶⁷ Corinth: e.g. Corinth vii 1, 162; vii 2, An 248.

⁶⁸ The published photograph of 36, from Incoronata, suggests it has a single band, although an early plece; in front of the vase itself, I was not sure whether or not there was a second band below the one given prominence in the photograph.

⁶⁹ Vertical wavy lines are found on the neck of an oenochoe of late MG date, Mylonas, pl. 397, 867. Coldstream, GGP 195, has occasion to remark ‘the vertical wavy lines are hardly to be expected before LG’. The most consistent users of them on the necks of amphorae are Euboeans, Boeotians, and islanders of the Cyclades, the earliest group being perhaps Delos group Aa which takes up the motif towards the end of Attic LG I (GGP 180). Certainly later is a squat neck-amphora, claimed to be Cycladic, with multiple O,Sc decoration on the neck: Boston 61.388, Class. J. Ixix (1963) 193–4, fig. 3.

⁷⁰ Amphorae in Leiden and the Agora, Davison, YaleClSt. xvi figs. 99 and 100 (= GGP 55, 5–6), descendants of the Leiden amphora fig. 94 (= GGP 55, 1).

⁷¹ On the Rhodian Kreis- und Wellenband aryballoi see Ridgway, , ‘The First Western Greeks’, Greeks, Celts and Romans 15Google Scholar, with bibliography.

⁷² Striped handles are found combined with O types of decoration on 6, 8, 45, and 69. The later member of the group with Tis Adiens 14489 (PLATE 18a); the rest are 5, 35, 47, and Villa Giulia, Cerveteri tomb 5, 11; the Eretria sherds published in AE loc. cit. should also be included and are apparently all of the early period. Megara Hyblaea tomb 209 has the same Sl,T,Sl neck but has not got verticals beside the handles, nor, apparently, striped handles.

⁷³ Both types are discussed in Beazley and Magi, Raccolta Guglielmi 50–2; for the amphoriskoi see also Agora xii 155–6 and for BF amphorae Jackson, East Greek influences on Attic vases 71–2. A direct echo of the SOS decoration in Ionia (an area where few SOS have yet been found) is the Clazomenian amphora from Olbia with sphinx between wavy lines on one side of the neck, wavy lines on the other, Olbia (1964) 155 fig. 23. Less likely to have been influenced from Athens are Chiot amphora with an O type motif on the neck, Lambrino op. cit. 139, Actes xii Con. Int. Ét. Class. 617, pl. 9, 2.

⁷⁴ Plut. Solon xxiv 1 and Ar. Ath. Pol. x.

⁷⁵ Hommages à Grenier 1560–1.

⁷⁶ A summary of Klein's thoughts on the same subject is in AJA lxxv (1971) 206.

⁷⁷ The basis of modern discussion of the reforms is Kraay's article in Essays presented to E. S. G. Robinson 1 ff.; most subsequent comment is listed by Rhodes, , Num. Chron. 1975 1 ff.Google Scholar The evidence for a single Greek mina weight, with minor variations, remains a little scattered; see in particular Crawford, , Eirene x (1972) 5–8Google Scholar and supporting evidence added by Kroll, Studies presented to George Hanfmann 92 and Johnston, atti xvii Convegno Magna Grecia.

⁷⁸ Gruben, AA (1972) 325–6. he postulates a pre- and post-Solonian Attic foot but can cite no actual use of either in Attica.

⁷⁹ Since Ath. Pol. does not tell us, there is no way we can say precisely what Solon is supposed to have increased beyond the Pheidonian. However, since Man can measure all things, we may assume that both linear and capacity measures come under this heading. is regularly used in both senses from Homer to Aristotle and beyond, although, as we have seen, it is difficult to decide whether this meant that in ‘Homer's day’ of the later eighth century the metretes was arithmetically linked with the foot or finger measure. We can be more confident that such a correlation had been made by the 590s, and so if we can discern no change in the capacity measures then we may suspect that the linear were not changed either.