¹ Stylistic dating of the Aegean fresco material has been queried especially by L. R. Palmer (Mycenaeans and Minoans (1961); On the Knossos Tablets: The Find-Places of the Knossos Tablets (1964); A New Guide to the Palace of Knossos (1969) esp. 126–8; and The Penultimate Palace of Knossos (1969) 126 f.) and Lang, Mabel L., The Palace of Nestor at Pylos in Western Messenia, II: The Frescoes (1969) 6f., 221–7.Google Scholar Thermoluminescence analysis of burnt plasters could supply other scientific dating criteria (v. Hans, M. C. and Ralph, E. K. in Scientific Methods in Medieval Archaeology (1970) 182–293).Google Scholar See, also, on radiocarbon dating of mortar, Folk, R. L. and Valastro, S. Jr, Journal of Field Archaeology iii (1976) 203–8.CrossRefGoogle Scholar

² In addition to dating criteria mentioned in Part I, the dates of samples in this study are those of the excavation records of stratified levels and deposits or, as asterisked in the list on p. 174, those put forward in Cameron's, doctoral dissertation, A General Study of Minoan Frescoes, University of Newcastle upon Tyne, 1975.Google Scholar

³ Prolonged attempts were made, with help from Mr. Petros Petrakis (the School's restorer-technician), to join the broken pieces, from 1963 to early 1968, and again in Mar. 1971, Aug. 1973, Dec. 1974 and Aug. 1975; generally, however, the paintings were too shattered and dispersed in Antiquity to restore at all extensively.

⁴ For the date of the destruction at Knossos, see Popham, M. R., Studies in Mediterranean Archaeology xii (1970)Google Scholar; BSA lxii (1967) 337–51; BSA lxiv (1964) 299–304; BSA lxv (1970) 195–202; SMA v (1964) on LM IIIB pottery from Knossos; also a review of OKT in JHS (1965) 193–6—the list could go on. On the decline of the wall painting industry, Cameron, dissertation (n. 2) Vol. i 452–7, 463 f., 473,478 f., 581 f., with Table at 600.

⁵ The technically and pictorially identical Roman painting turned up in a late second-century A.D. context (UM Trench 15, level 4 southern extension), as its excavator, Mr. L. H. Sackett, very kindly informed us.

⁶ Journal of the Royal Society of Arts lviii (1910) 206 footnote (8 Mar. 1900). See, too, Shaw, J. W., ‘Minoan Architecture: Materials and Techniques’, Annuario della Scuola Archeologica di Atene xlix, N.s. xxxiii (1971) 209 n. 3Google Scholar, who ascribes recognition of the base for plaster on mud bricks as lime to Durm's investigations (published in 1907).

⁷ JRSA (1910) 206–12; JRIBA xviii (1911) 697–710; Tiryns ii (1912) 211–16.

⁸ JRIBA 698f.

⁹ Archaeology ix (1956) 196.

¹⁰ Tiryns ii 216.

¹¹ Profi, S., Weier, L., and Philippakis, S. E., Studies in Conservation xix (1974) 107–11.Google Scholar

¹² JRSA 210 and JRIBA 708.

¹³ BSA lviii (1963) 94 ff., in collaboration with E. E. Richards and A. E. Blin-Stoyle, with regard to composition and provenance correlations of Mycenaean and Minoan pottery.

¹⁴ Hall, E. T., Schweizer, F., and Toller, P. A., Archaeometry xv (1973) 53.CrossRefGoogle Scholar

¹⁵ The results for nos. 19, an amorphous pink lump, and 97C, the limner's putty, have also been given in TABLE 5. These two samples are excluded from all figures and calculations. For convenience, our use of the following samples should be stated:

no. 2: a mud plaster, excluded from FIGS. 2–5 and from calculations leading to TABLES 2–3;

no. 3 bis: a probable chronological ‘rogue’, included in all figures and calculations except FIG. 3 and calculations leading to TABLE 2 which involve a consideration of period;

no. 12b: the lighter of two pinkish plaster layers, included in all figures and calculations;

no. 100: a Roman sample, excluded from all figures and calculations (see note 5);

no. 101: evidently part of a painted ‘pot-stand’, included in all figures and calculations except FIGS. 5–6 and calculations leading to TABLE 3 because inevitably contaminated with clay from the sherd on which it rests;

no. 103: coarse floor-plaster lump, excluded from FIGS. 5–6.

¹⁶ For example, sample 94 with 71 per cent lime content has a combined aluminium, magnesium, and iron oxide content of only 0·5 per cent; while sample 89 with a lime content of over 95 per cent has a total clay content of 4·3 per cent.

¹⁷ This figure omits nos. 2, 19, 97C, 100, 101 and 103 (see note 15).

¹⁸ This figure omits nos. 19, 97C, 100, 101 and 103 (see note 15). The samples are assigned to each group as follows:

Group I: nos. 22, 27, 30, 32–3, 39, 43, 52, 58, 63, 97B, 108, 110; also nos. 41–2, 48 and 109 from overlap with group II;

Group II: nos. 17, 38, 40, 46, 62, 78–81, 92–3, 95, 111; also nos. 36, 60, 106 from overlap with group I, and nos. 11, 15, 21, 37, 84, 104 from overlap with group III;

Group III: nos. 4–10, 12a, 13–14, 16, 19–20, 24–5, 29, 31, 34–5. 45, 47, 50–1, 53, 55–7, 61, 66, 68, 70–4, 83, 85, 87–91, 94, 99, 102–3, 105, 107; also nos. 1, 3 bis, 12b, 18, 77, 82, 98 from overlap with group II.

¹⁹ Respectively, south of the aqueduct marked in areas E/F 10, and between nos. 135–7 in areas C/D 8, of M. S. F. Hood's Archaeological Survey of the Knossos Area (1967), map.

²⁰ But cf. I. W. Cornwall, Appendix 3, in Evans, J. D., ‘Excavations in the Neolithic Settlement of Knossos, 1957–60, Part I’, BSA lix (1964) 240Google Scholar, from Stratum III (MN-early LN, ibid., p. 172), for reasonable suggestions for the possible use of such lumps.

²¹ JRIBA 698.

²² See Evans, J. D., BSA lix (1964) 146Google Scholar, 149, 153, 157, 166, 174 and pls. 59(1) and 38(1), describing EN I–II roof and MN interior wall clay plasters.

²³ As Heaton first suspected, op. cit. 697.

²⁴ Ibid. 697 and 705.

²⁵ So also Myrtos 313.

²⁶ Op. cit. 699.

²⁷ Ibid., loc. cit.

²⁸ No. 15, with an unusually low calcium carbonate figure, is excluded from this calculation which is therefore generously conservative.

²⁹ Among other considerations too detailed to list here, we may pick out the introduction of non-representational mural decoration in MM IB and of pictorial and figure depiction in MM IIIA. The issues have been discussed further in Cameron's doctoral thesis (Ch. 3) and in a paper presented to the AIA Society branches in Pittsburgh, Pa., and Providence, R.I., in Feb. 1977.

³⁰ See especially Heaton's commentaries in JRSA 208 f.; JRIBA 699 f.; and Tiryns ii 213–15, at which site alone he determined the presence of an unburnt powdered limestone ‘filler’ but only in the lower of two lime layers in three specimens—the only report known to us of the discovery of any type of ‘filler’ in any Aegean B.A. lime plasters. See TABLE I (p. 128) for references to previously published analytical results.

³¹ As Heaton first noted (JRSA 208 f., cited by Evans at PM i 535 n. 3). See, also, Mayer 353, Doerner 278, or Hale 15ff., all of whom prescribe large quantities of sand in lime mixes for fresco plasters as necessary to prevent ‘fat’(high percentage calcium) lime plasters from cracking. Discussion with a professional plasterer, Mr. Ed. Gebhardt of London, Ontario, suggests the modern plastering industry's outlook remains unchanged since the time of Heaton's inquiries. Recent experiments at the University of Western Ontario, however, have indicated several possible ways in which the Minoans, at least, may have made pure lime plasters— with or without a calcium ‘filler’ such as that proposed by Heaton as theoretically possible (if still unproven) in the case of Cretan fresco samples (note 30). In this connection a report is in preparation.

³² Our MM IIA samples may simply have come from technically inferior decoration of rooms of minor importance, or from lime batches from inferior limestone sources; or the excitement of decorating the new walls of the First Palace may have worn off, with the result that the industry—or its town-house branch—turned out only second-rate plasters in MM 11 A. And so on.

³³ JRIBA 699.

³⁴ Ibid. 698, in reference to EM II plasters from Vasiliki. It should be noted that alumina content remains relatively high in the present samples up to and including those of MM IB date, by later analogy.

³⁵ Ibid. 699.

³⁶ Such variations extend to (a) different samples of a single composition; (b) samples from different paintings but of the same date and site; (c) contemporary specimens from different Knossian or other Cretan sites; (d) contemporary samples from one site but of different practical or artistic functions; (e) contemporary samples from one site but from structures of different architectural or social importance; and (f) plaster samples from the same quarry sources.

³⁷ Ibid., loc. cit.: ‘the composition varies in different parts’. The sample groups from single compositions, nos. 36–41, 42–7, 55–8, 60–1 and 80–2, confirm this.

³⁸ Further in this connection, the general absence of figures of estimated accuracy from reported analyses makes the significance of such data even more uncertain.

³⁹ JRIBA 700 n. 2. PM i 448–62 more accurately defines the cists' uses, although confusing the dating of their architectural alterations (see Popham, , SMA xii (1970) 50–3Google Scholar). Further, on large lime putty containers, see Shaw, in Annuario xlix, N.S. xxxiii (1973) 213Google Scholar citing ‘a number of pithoi’ from a house to the north-east of the Palace at Zakro.

⁴⁰ JRIBA 700; so, too, Evans at PM i 532. Sinclair Hood has kindly warned us no Minoan pottery has yet been found in the Hagia Irini (Kessoboros) quarry.

⁴¹ On the chemistry of gypsum plaster, see Mayer 475 f., and on ‘gypsum walls’ Doerner 313. On its suitability for fresco reconstruction and restoration, see Cameron, Europa 46 and 50 n. 8, and Temple University Aegean Symposium 1976 (Philadelphia, 1971) 25 f.

⁴² So Duell, and Gettens, , Technical Studies in the Field of Fine Arts x (1942) 203 f. and n. 27Google Scholar, citing A. Lucas's opinion that lime burning for Egyptian plasters was unknown there until Ptolernaic times when the process was introduced by the Greeks, (Ancient Egyptian Materials and Industries (1934) 73 f.)Google Scholar, or Romans (so Mayer 350). In Crete's wetter climate, gypsum would be less desirable than lime for wall painting owing to its greater solubility under wet conditions.

⁴³ In ‘tempera’ painting, the pigments bound with an emulsion (usually egg, glue, or casein) are applied to a wholly dry, hard, surface: v. Doerner 211–43 and Mayer 231–57. ‘Fresco secco’, developed to imitate true or ‘buon fresco’, also requires binding media (either inorganic or organic) in the pigments which are applied to a hard but remoistened or soaked lime plaster, producing a painting that lacks the special durability and clarity of ‘buon fresco’: v. Doerner 297 f. and Mayer 360–2. The essential requirements of ‘buon fresco’ are simply two-fold: the pigments should be mixed only with water (though lime-water is admissable: v. n. 77) and they should be applied to a lime plaster while it is still soft and malleable, i.e. when drying out for a first time. Successful execution of this mediod of painting ensures the inseparable bonding of pigments to plaster by both chemical and mechanical means for as long as the lime plaster and its architectural support remain intact: v. Doerner 264–96; Mayer 340–60; Hale passim; Cennini, Cennino d'Andrea, Il Libro dell’Arte lxvii, trans. Thompson, Daniel V. Jr,(1933) 42.Google Scholar

⁴⁴ As Mr. Gebhardt (note 31) has informed us.

⁴⁵ Doerner 268 and cf. Mayer 353.

⁴⁶ Duell and Gettens, op. cit. 221.

⁴⁷ e.g. PM i 453; SMA xii (1970) 52 with pl. 32e; and Cameron, doctoral thesis, Vol. I 277 and 446.

⁴⁸ See Doerner 271–4; Mayer 347–50; Hale 12–14; Cennini, op. cit. clxxv (trans. Thompson, Jr. 119 f.). Also Vitruvius, , De Architectura ii, 3 and 8Google Scholar, and vii, 3–4, whether in preparatìon for ‘secco’ or ‘buon fresco’.

⁴⁹ Annuario op. cit. 214.

⁵⁰ Contra Shaw, ibid. 215; cf. Doerner 294 f., referring to Adolf Wilhelm Keim's trowelling experiments, to which we may add supporting evidence forthcoming from the recent Minoan fresco duplication project at the University of Western Ontario.

⁵¹ Michalowski, Kazimierz, The Art of Ancient Egypt (1969) 169 f.Google Scholar and Col. pls. 68–9 with p. 383, no. 337 (Senmut), for example.

⁵² See, for example, Procacci, Ugo in Frescoes from Florence (Hayward Gallery, London, 1969) 1569.Google Scholar and 46 ff. nos. 1–3 5–6, 8–9, 18, 21–2, 24, 26–8, 32–3, 35–6, 38, 40–1, 43, 47–8, 51, 67.

⁵³ Swindler, Mary, Ancient Painting (1929) 74.Google ScholarCameron, in Kadmos vii (1968) 53–8Google Scholar; Europa 46, 57–61, 63–5 and figs. 5–6; BSA lxiii (1968) 1 ff., nos. 5–6, 40–6; and Temple University Aegean Symposium 1976 (Philadelphia, 1976) 39 n. 50. Coleman, Katherine, Hesperia xlii (1973) 298.Google Scholar We define a slip as a finely ground, usually white, lime layer of a regular and visibly distinct physical substance (i.e. capable of flaking from the supporting main body of plaster as a thin wafer-like layer), measuring not less than d. 0·3 mm. to about 2·0 mm., which provides the surface for the painting proper.

Several samples were tested chemically for the presence of gypsum because certain samples appeared unusual under the stereomagnifier. The tests showed that gypsum was either absent or present in very low concentrations and in the latter case evidently accidentally so.

⁵⁴ See Frescoes from Florence 28 fig. 4 (‘pontata’) and cover illustrations (contours of heads and drapery especially, for ‘giornata’); ‘giornata’ appear abundantly in, for example, Michelangelo's frescoed Chapel, Sistine (The Complete Work of Michelangelo, New York, ed. Salmi, Mario III, Painting, by Roberto Salvini 158 ff. with col. pls. 5–22).Google Scholar

Professor Mabel Lang has suggested that background areas in different colours, especially where over-painted with wavy lines or bands, may correspond to different days' work (Pylos ii 14). Even so (though this interpretation may be queried), such ‘divisions’ are not true ‘giornata’, whose contours are defined by cutting off any waste plaster at the end of the day, along which lines distincdy visible ridges or ‘seams’ appear where the next day's plaster adjoins.

⁵⁵ Thus Heaton, while believing the Aegean method was almost universally ‘buon fresco’ lists objections arising from the absence of certain Italian fresco features (see JRSA 211 and JRIBA 709); see, too, Duell and Gettens, op. cit. 193–5, 218 (3) and 219 f.—‘From the few facts available it cannot be concluded that these Aegean wall paintings were done in true fresco after the manner of the Italian painters.’ The issues are further confused, particularly in accounts which reject a ‘buon fresco’ classification, by a lack of descriptive clarity: thus we read such phrases as ‘wet’ or ‘damp plaster’ in contexts that leave it uncertain if the lime is wet because it is soft and malleable (i.e. freshly laid so still drying out for a first time when painted) or because, having already hardened by once becoming dry, it has been remoistened or even freely soaked with water before painting. The difference is crucial, being that between ‘buon fresco’ and ‘fresco secco’, if we leave aside here the issue of any binding agent in the pigments (never satisfactorily detected).

⁵⁶ Similar to Hood, Sinclair, The Minoans (1971) 38 fig. 14 (MM IA).Google Scholar

⁵⁷ See Annuario op. cit. 218–21 and figs. 250–3, on ‘tarazza’ floors.

⁵⁸ Ibid., loc. cit.

⁵⁹ Ibid. 210 fig. 238, from Gournia.

⁶⁰ PM iii 319–21, 333–8 and 356; Palmer, L. R., A New Guide to the Palace of Knossos (1969) 85–8Google Scholar; Shaw, Annuario op. cit. 213 with n. 4. But it is uncertain what the origin of the calcium deposits actually was, whether freshly quarried limestone, fragments of old wall plasters, or building materials (limestone and/or gypsum), whether calcined in the final great destruction by fire or by actual industrial processes: clearly scientific analysis of any available calcium samples (as from the material said to have preserved the shape of a throne in the Hall of the Double Axes: PM iii figs. 222–3) might resolve present uncertainties on the matter.

⁶¹ See Cameron, and Jones, , in BSA lxxi, 15–19Google Scholar, ‘A note on the identification of fresco material from the British campaigns at Palaikastro, 1902–1906’, for further discussion of sample no. 114.

⁶² JRSA 209 and JRIBA 705. Heaton suggested bone black at Tiryns (Tiryns ii 215), while Profi et al. have suggested three possible blacks at Mycenae, (Studies in Conservation xix (1974) 107–11).Google Scholar Analysis of no. 115 here seems to rule out grey-black dolomitic rock as a possible source of black pigment at Knossos.

⁶³ See footnote to TABLE 6, and Studies in Conservation xxi (1976) 34–9.

⁶⁴ Schubert, W. and Seidel, E., Z. Deutsch Geol. Gesell schaft cxxiii (1972) 371.Google Scholar

⁶⁵ Evans reported a ‘kyanos’ (Egyptian) blue painted fragment from a MM IIA deposit below the 10th step of the eastern flight of the Theatral Area (PM iii 249 f. with fig. 172). For riebeckite, no. 17 in the present study.

⁶⁶ As the Vat Room (MM I) and Loomweights Base ment (MM IIB) faience deposits at Knossos may suggest: PM i 490–4 with Index 47 f.; Cadogan, G., paper to the Temple University Aegean Symposium 1976 (Philadelphia, 1976) 18 f.Google Scholar

⁶⁷ S. E. Philippakis et al., ‘X ray Analyses of Bronze Age Pigments from Thera’, submitted to Studies in Conservation (1976). Glauconite, a related iron mineral, has been identified as a green pigment by Giavanoli in Roman murals from Switzerland (Archaeometry xi (1969) 53 ff.).Google Scholar

⁶⁸ Lucas, A., Ancient Egyptian Materials and Industries (1948) 342.Google Scholar Charles Kiefer and A. Allibert's experiments on selfglazing processes, however, suggest 870–920 °C (Archaeology xxiv (1971) 107 ff., esp. 112 f.).

⁶⁹ Tiryns ii 216. For malachite at Early Neolithic IA Knossos, see BSA lix (1964) 238.

⁷⁰ Green, like blue, seems to have been traditionally among the more difficult of colours to make adhere. Strangely, although good blues and yellows were known in Crete by MM IIA, no evidence for their mixing as a green is yet known before MM IIIA. Some greens, especially from the compositions from the House of the Frescoes (MM IIIB/LM IA) now look somewhat faded; and after this date there seems gradually less and less use of this colour, at least in the Cretan fresco record.

⁷¹ The white ‘impasto’ bands imitating rectangular stone floor slabs on a red-painted MM I stucco floor in situ in a building to the east of House E at Mallia (BCH lxxxix (1965) 1000 f.Google Scholar, figs. 1–2 and ADelt xx (1965) B'3, 571 pl. 722b); also on MM IB stucco imitating a speckled stone (liparite ?) from a small house on the southern slope at Phaistos (MA xii (1903) 20Google Scholar with pl. VIII 6 and p. 83 fig. 22—top row, no. 1, for a mention and illustration of associated polychrome pottery; see, too, PM i 180 with n. 1).

⁷² From the MM IIA Theatral Area context, cited at note 45 above (PM iii 250 n. 3). Grey is also found on the Loomweights Basement dadoes, from a MM IIB context (PM i 251 f. fig. 188 a–b; DM/DB 5 Apr. and 21 May 1902 (I); also Fyfe, Theodore, JRIBA x (1903) 109 figs. 1–2).Google Scholar

⁷³ From the Loomweights Basement deposit (see n. 72 for references); and on the ‘Sponge-print dado’ from below the west threshold and wall of the N.W. Portico (AE/NB 1929, 11; PM iii 361 ff., figs. 238–40, the latter a comparable MM IIA cup—to which period Evans ascribed the fresco; and PM iv 109). Somewhat indefinite hues of ‘ochre’ have occasionally been noted at earlier dates, as at EM II Myrtos, ‘Phournou Koriphi’ (Myrtos 306).

⁷⁴ Evans, J. D., BSA lix (1964) 238Google Scholar; BSA lxiii (1968) 271, mentioning white and more rarely ochre ‘pastes’ (impasto) for pottery decoration.

⁷⁵ JRSA 207ff.; JRIBA xviii (1911) 708; Tiryns ii 216; Duell and Gettens, op. cit. 218 (4); Majewski, Lawrence J. and Reich, Marjorie in Hesperia xlii (1973) 299Google Scholar, agree with three ‘possible exceptions’. Young, W. J. reports at Pylos ii 230Google Scholar: ‘Unfortunately, many of the mediums would be leached out as a result of burial’, to which we should add ‘if any binder were ever added at all’. Most recently, Profi, et al., Studies in Conservation xix (1974) 107 (4.1) and 110Google Scholar, concur on this point.

⁷⁶ Technical Studies in the Field of Fine Arts x (1942) 212.

⁷⁷ Archaeology ix (1956) 196. But the latter statement has no scientific basis in proven fact; likewise the authors'theory that ‘the color was always kept wet with lime water (calcium hydroxide)’ [i.e. in preparatory suspension?]. And if the paints were applied when Levi says they were, was the plaster soft, or hard but dampened? From its context it is clear that this argument has been offered to account for the apparent lack of pigment fusion with or penetration into the lime plaster. This is a feature which those authors assume—incorrectly in the present view— to be a necessary requisite for identification of ‘buon fresco’ (see pp. 163 f., below). Moreover, both modern and Renaissance ‘buon fresco’ painters have made good whites from and modified other colours with ‘well-seasoned pit lime’: Doerner 282 and cf. 297 f.; also Mayer 360; Cennino d'Andrea Cennini, II Libro dell'Arte Ch. 58. And, not least, Heaton, Tiryns 215, specimens 7 and 10 (modified greys) and specimen 5 (white pigment), for Aegean Bronze Age examples.

It is incorrect, therefore, to imply that mixing some lime in coloured pigments or suspending others in lime water is impermissible in ‘buon fresco’ painting, as Levi's suggestion of a ‘lime tempera’ method of painting intends (loc. cit.). In any case, calcium hydroxide will always ‘contaminate’ even solely water-mixed pigments in ‘buon fresco’ simply because they are laid on a soft plaster whose calcium hydroxide converts to calcium carbonate at the painted surface. So how can some lime in the pigments possibly be avoided?

⁷⁸ Pylos ii 10 ff. (with n. 15) and 229 f. Yet this view suffers from much the same (and other) objections as those already made to Levi's propositions (note 77), and in any case seems at odds with what her scientific analyst actually reported: ‘All the pigments appear to contain a low percentage of lime. However, this could be the result of burial and not be from a direct mixture with the pigment itself’ (ibid. 230). Clearly the percentage lime content in the Pylos pigments was too slight to draw the conclusion in question.

⁷⁹ Hesperia op. cit. 300.

⁸⁰ Oxford Companion to Art (1970) 444.

⁸¹ Surface disturbance from use of a float seems best indicated where the marks go in any counter-direction to painted brushstrokes where these have not been obliterated by polishing; and brushmarks (as against painted brushstrokes) if the direction of marks and strokes is identical because both features were caused simultaneously, as seems true in nos. 52 and 99.

⁸² Procacci, Ugo, Frescoes from Florence (1969) 15.Google Scholar

⁸³ Since Heaton's publication of a microphotograph (no scale) of a paint-film (Tiryns ii 214 fig. 79), Duell and Gettens have emphasized micron measurement of pigments (Technical Studies in the Field of Fine Arts x (1942) 208 with fig. 11, 215–18 (specimens catalogue) and 219 (10), for a summary. So, too, Hesperia op. cit. 299 f. (Kea frescoes).

⁸⁴ Hesperia loc. cit. The Kea ‘controls’ are presumably ‘Renaissance type’ fresco controls, i.e. with admixtures of some kind of traditional, non-calcium, filler. It is, unfortunately, not recorded whether the reference samples in ‘fresco secco’ were done with more than one pigment. It cannot be assumed that the properties of blue pigment on dry plaster are the same as those of red pigment.

⁸⁵ Heaton (JRSA 211) refers to Professor Baldwin Brown's notes to ‘Vasari on Technique’ 288, stating:‘It is a mistake to speak of the pigment sinking into the wet plaster. It remains on the surface, etc. …’, a view which Heaton explicitly says he cannot endorse in the light of the Minoan material he studied. But a case could be made, with reference to recent practical experimentations, to show that both types of result are equally compatible with ‘buon fresco’ painting, and so their evidence—contrary to general supposition—need not be thought to be mutually exclusive. The real value of micron measurement of Aegean pigments thus far seems to be that it shows objectively that a uniform or commensurate result is not obtained for similar hues of colour, let alone for all the colours the Aegean artists employed (references in note 81).

⁸⁶ e.g. Heaton, JRIBA 709. Another instance Heaton convincingly excluded as probably contaminated by the conservator's wax (JRSA 210), such as was definitely applied to the surfaces of several Knossian frescoes (see Cameron in Temple University Aegean Symposium 1976 41, n. 61, for examples and references).

⁸⁷ Art Bulletin xlvi, no. 3 (1964) 378.

⁸⁸ See n. 50 above for references to practical experiments in 1911 and 1976. If there had been an organic lubricant used in polishing, its decay since the Bronze Age would now call for an explanation for the excellent state of preservation of many, some very highly, polished painted surfaces to this day.

⁸⁹ e.g. the ‘Embossed Band’ series of relief stucco frag ments (PM iii 37 fig. 20; Knossos Fresco Atlas (1967) col. pl. E, fig. 2a–c and Catalogue 23).

⁹⁰ JRIBA loc. cit.

⁹¹ Lang, , Pylos ii 10 with n. 14.Google Scholar

⁹² Ibid., loc. cit. with n. 15, apropos Duell and Gettens, op. cit. 209–12 and 219 (6).

⁹³ It is ‘prodigal’ indeed that this flow of calcium ‘fixed’ the ground, but not secondary, pigments to their plasters (thus Pylos ii 10 ff.); the ‘fugitive’ or‘friable’ paints (cf. Levi, , Archaeology ix (1956) 196Google Scholar) may have bonded ineffectually simply because they were applied where the plaster had perhaps become too dry to take the paints, or because the painted surface was not polished hard or long enough to attract additional calcium hydroxide to the surface where it would carbonate. See PLATE 16b. There is no need to invoke a ‘lost binder’.

⁹⁴ The fresco project at the University of Western Ontario's Department of Visual Arts, spring 1976, produced six small works for study. But Keim had reached similar conclusions (e.g. a 14-day painting period on soft plaster at Pompeii in the tropical July heat of 1911—under a sheet-metal roof: Doerner 295) sixty-five years ago: Doerner goes on to mention that this is why ‘giornata’ or ‘tareas’ (similarly add ‘pontata’, too) were unnecessary in ancient painting.

⁹⁵ HM tray 47 Zeta XVIII, I ‘Area of North Foundations’ (Knossos). The sequence of coloured bands, defined by black lines, from top to bottom is: white, black, pale blue with brushmarks, white, and then a dark red, much scratched, area.

⁹⁶ The UWO fresco project (n. 94) provided interesting information on this point

⁹⁷ e.g. the famous Tiryns bull with two sets of forelegs and three tails (Schliemann, Heinrich, Tiryns (1886) Col. pl. 13Google Scholar), and evidently the floor in the Throne Room at Pylos (Pylos ii 11), to which more examples from both other Mycenaean and Minoan sites could be added.

⁹⁸ HM tray 162 Gamma VII, 6 (‘Ohne Angabe der Fundstelle’)—of which Dr. Helga Reusch rightly notes ‘sicherlich HOF’ (House of the Frescoes). Cameron's interpretation of this fragment's technique at BSA lxiii (1968) 3 is now corrected at Bulletin of the Institute of Classical Studies xvii (1970) 166.