1 The more important of these proposed reconstructions may be summarized as a basilica with a single apse in the position of the present east end extending to the present west end (Cattaneo, cited in Forlati, F., La Basilica di San Marco attraverso i suoi restauri (Trieste, 1975), 47)Google Scholar and a triple apsed single domed church extending over the present ground plan of the interior (Forlati ibid., 52). Both these hypothetical churches are believed to have been destroyed.

2 This paper is presented in the form of an hypothesis, providing an interpretation of the building on the basis of available and visible evidence. Before its conclusions can be entertained with certainty the hypothesis will need further proof by opening up the structure to reveal construction joints and evidence of construction sequence.

3 Wherever possible reference is to the works of Otto Demus whose extensive summaries of the church and the mosaics embrace other available sources. An important subsequent publication, though not available on public sale is San Marco, the mosaics, the history, the lighting (Milan 1990).Google Scholar Other significant recent documents include Herzner, V., ‘Die Baugeschichte von San Marco und der Anfstieg Venedigs zur Grossmacht’, Wiener Jahrbuch für Kunstgeschichte, 38 (Koln 1985).Google ScholarDeichmann, F. W., Corpus der Kapitelle der Kirche von San Marco zu Venedig (Wiesbaden 1981).Google Scholar

4 The nordiernmost of the western doors, so named because within stands a mosaic panel of the saint, a stylite. His name is accorded also to the dome within the portal.

5 Now in the gallery of the Accademia in Venice.

6 Probably 1267, Demus, O.The Church of San Marco in Venice (Washington, 1960), 103, 104, n. 170.Google Scholar

7 On the first Sunday of Lent. Mango, C., ‘The Liquidation of Iconoclasm and the Patriarch Photios’ in A., Bryer and J., Herrin (eds.) Iconoclasm. (Birmingham 1977), 133–40.Google Scholar

8 There is a significant difference in the painted representations of the earlier mosaic sopraporte (Bellini). Only the St Alipio mosaic is in a conch, the remainder being on flat panels which suggests, among other things, that the first mosaics were on flat lunettes; and they have, therefore, been restored in this way. Non-figurative designs are also shown on the upper levels of this façade. They are of a vegetal (vine rinceaux) type found from Roman times and used in the Byzantine and early Muslim era. Equivocally they may be taken to indicate that the western superstructure was higher than is here taken to be the case, or that they are part of the eleventh-century rebuilding, although by that time such designs were much less used. Alternatively there is an argument for giving the lower lunettes a later date than is postulated here.

9 G. Gambosi, II Portico, 163, cited in Forlati op.cit. (note 1), 67 footnote and Demus op.cit. (note 6), n.71, n.94.

10 The dates of this phase of building derive largely from the sixteenth-century historian Magno, StefanoLa Basilica di San Marco Documenti, F., Ongania (ed.) (Venice, 1881–1888) 30–37, 39–41, 46, 811. Precise dates remain uncertain, but Doge Domenico Contarini died in 1070, having initiated the work some years earlier and it was completed, i.e. dedicated in 1094 in the Dogeate of Vittorio Falier which ended in 1096. See Forlati, op. cit. (note 1).Google Scholar

11 O. Demus op. cit. (note 6), 77. ‘Startling as it may appear, it seems to be the only possible conclusion.’

12 Creswell, K. A. C., A Short Account of Early Muslim Architecture, 2 vols. (Oxford, 1940), 11, Chapter XIV, 289–307.Google Scholar

13 White, H. E., The Monasteries of the Wadi 'n-Natrun (Cambridge, 1926).Google Scholar

14 O. Demus, op. cit. (note 6), 104–5.

15 Ibid., 66, 88.

16 Pointed out by Dr Rowland Mainstone F.S.A., to whom I am deeply indebted for helpful discussions during extended joint visits to the church. Also see O. Demus, op. cit. (note6), 85.

17 This and other structural arguments are being examined further by Dr Mainstone and publication is in prospect. It is fair to conclude at this point, that the visible deformations of the central dome are consistent with those to be expected of a single dome built in isolation.

18 The current photogrammetric survey of the church when available will perhaps provide information of sufficient precision to allow conclusions to be drawn.

19 Among the documents of the church, published as La Ducale Basilica. See O. Demus op. cit. (note 6) bibliography.

20 O. Demus op. cit. (note 6), 69, ‘… redintegrare, redifichar, reparare, restaurare …’.

21 The chronicles do, however, refer to the incorporation into the new building of the old church of St Theodore.

22 O. Demus op. cit. (note 6), 73. Windows of St Theodore.

23 S. Magno op. cit. (note 10), 811

24 O. Demus op. cit. (note 6), 64.

25 The word ‘exactly’ needs qualification. By minimizing the width of the arch before the apse the architects made a small saving on the eastward dimension although by so doing they seriously reduced the lateral strength of the arch, which bowed forward in its plastic state and has solidified in this deformed way.

26 O. Demus op. cit. (note 6), 63.

27 Demus, O., The Medieval Mosaics of San Marco, Venice, a color archive, 2 vols. in 4 (Chicago, 1985), 11, 45. Demus describes the pattern in t he eye of the cupola in the north-east pier as a ‘wheelshaped ornament in the zenith’. This is a curious description from a scholar of mosaic of a pattern regularly used, particularly but not only in the Iconoclast period, representing a Greek cross with radial rays. He refers also to the discovery of a twelfth-century coin in the mortar backing of the lower gold ground to the mosaic which may not be quite the firm evidence that it first appears. Repairs have visibly been made in sections of this mosaic as can be seen from its background. Recognizing this distinctly earlier style Demus concedes ‘It is not impossible that the original decor of the two easternmost tribunes was made in the twelfth century and only restored or remodelled in the thirteenth.’ By this argument an even earlier date may be considered on similar terms.Google Scholar

28 i.e. the domes known as the first, the second and the third dome of Joseph, the dome of the Creation and of Abraham and the Moses vault:

29 I am indebted to Ernest Hawkins F.S.A. for help in discussions and material bearing on this problem.

30 A caution must be issued: in later repairs some iron reinforcement was introduced into the structure and these elements may prove to be wholly or in part of later date than the Second Church.

31 Some half-dozen columns of similar diameter and similar marble are incorporated structurally on the south face, and a careful analysis of similar columns, when present protective screens are removed, may yield important information on the reuse of the columns of the narthex and their number.

32 Harrison, R. M., A Temple for Byzantium (London, 1989), 103.Google Scholar

33 Purple porphyry is used in this church extensively in conjunction with green (Greek) porphyry as inlay. This is a combination typical of early Byzantine churches—San Vitale, Ravenna; Haghia Sophia and St Polyeuctos among them. Its placement and the sources may yield valuable information.

34 F. W. Deichmann, Corpus der Kapitelle von San Marco Zu Venedig ref. capitals 1 and 2 and 5–16.

35 On Islamic material in the church see Howard, Deborah ‘Venice and Islam in the Middle Ages’ Architectural History, 34 (1991), 61–3.CrossRefGoogle Scholar

36 O. Demus op. cit. (note 6) Book II. Architecture, 82–5. The alteration is attributed to the first half of the thirteenth century on stylistic grounds.

37 Ibid., 22–5. The foundation of the Second Church is attested by later chroniclers referring to twelfth-century sources.

38 Such an example is the Nilometer on Rhoda Island in Cairo, where a later Caliph had his predecessor's name taken out of the foundation inscription.

39 These foundations, recently exposed (and inspected by the author) are of the same squared limestone. The entire crypt wall is laid upon them and it may be that in the Great Rebuilding foundation stone was salvaged from the removed lateral walls and reused in the deepest of the new foundations, i.e. the crypt. The original church would in this sense indeed have been ‘removed to its very foundations’ and this removal would account for the failure of Sig. Forlati to find foundations when he excavated for them in the transept. F. Forlati op.cit. (note 1), 49 et seq. and 74. The absence of such foundations caused Sig. Forlati to reject the proposal that the First Church was a basilica.

40 O. Demus op. cit. (note 6), 65. It is noteworthy that the building of one dome of this size took over two years. Given the scale of the work this is a reasonable time scale, the limiting factor being the slow process of carbonation in so great a mass of mortar. It is corroborative that the building of the structure of the Second Church took eight years, a similar time for each of the four domes assuming them to have been built successively.

41 Ibid., 71.

42 Ibid., 65.

43 In presuming the destruction of its predecessor historians have been forced to explain the retrospective character of the Second Church as imitative of antique forms of Byzantine Architecture, e.g. Mango, C., Byzantine Architecture (New York, 1976). To a lesser degree this argument holds even for the First Church.Google Scholar

44 O. Demus op. cit. (note 6), 73. ‘Its architects were able to use not only the foundations but parts of the walls of the Participaci church …’. The reconstruction, by Sig. Forlati, construes the original church as having been of cruciform construction having a single dome and extending over the whole plan area of the present church.

45 F. Forlati op. cit. (note 1), 46. A bridge over this canal was excavated and described after its discovery. Its interlocking voussoirs may be late Roman.