443. Compare this with the contemporary tower at Ghazni, in the chapters on Saracenic Architecture in India in vol. iii.

444. Numerous examples of Byzantine interlaced work of all periods will be found in Cattaneo’s work ‘On the Influence of Byzantine Art in Italy from the 5th to the 11th centuries.’

445. So much of the information regarding Spanish architecture which is contained in the following pages, is derived from Mr. Street’s beautiful work, entitled ‘Gothic Architecture in Spain,’ published in 1865, that it has not been thought necessary to refer specially to that work in the text. With one or two exceptions, all the plans are reduced from those in Mr. Street’s book, and many of the woodcuts are also his. If any one will take the trouble of comparing the very meagre account of Spanish architecture contained in the ‘Handbook,’ with what is said in this work, they will at once perceive my obligations to Mr. Street. His work is a model of its class, and has quite revolutionised our knowledge of the subject.

446. Parcerisa, ‘Recuerdos y Bellezas de España,’ Asturias, p. 78.

447. ‘Monumentos Arquitectonicos.’

448. ‘Monumentos Arquitectonicos.’

449. Ibid.

450. These external porticoes would be admirably adapted for imitation in the climate of India.

451. The Spanish arrangement has recently been adopted in Westminster Abbey, more by accident than design; with an effect as disastrous as anything in Spain, and apparently as little felt. In monastic churches the choir is always in a gallery above the west doorway.

452. The Church of St. Eustache at Paris was commenced as late as 1532, and, although its plan is almost as Gothic as those of the Spanish examples, the details of the French church are far more essentially Renaissance throughout.

453. The room called Paranimfo in the University of Alcala (see Woodcut No. 89, History of Modern Architecture, vol. i.) is of precisely similar design to this, only carried out with Renaissance instead of Moorish detail.

454. An engraving of this tower is given in Street’s ‘Gothic Architecture in Spain,’ page 225, accompanied with a very complete enumeration of all the examples of the style to be found in Toledo.

455. Another example exists at Palma, in the island of Majorca, in which there are no capitals to the columns, the ribs of the vault dying into the shaft.

456. These were destroyed by a fire which occurred between thirty and forty years ago.

457. Abulfeda, ed. Reiske, vol. i. p. 32.

458. ‘The History of Jerusalem.’ Besant and Palmer, 1888.

459. ‘The Holy Places of Jerusalem,’ by T. Hayter Lewis, F.S.A. Murray, 1889.

460. ‘Description of Syria,’ by Mukaddasi. Translated and annotated by George le Strange for the Palestine Pilgrims’ Society. London, 1886.

461. Mejr ed-Deen. ‘Fundgruben des Orients.’

462. Transactions of the Royal Institution of British Architects, 1878-79.

463. Ante, p. 228, vol. i.

464. I state these dimensions very doubtfully, the ground outside the present mosque never having been carefully surveyed by any one competent to restore the original plan.

465. ‘History of Jerusalem,’ translated by the Rev. M. Reynolds, p. 409 et seqq.

466. Translated by Jaubert, tom. i. p. 303. The particulars of the description in the text are taken from M. Girault de Prangey ‘Monuments Arabes,’ compared with M. Coste’s ‘Edifices de Caire.’

467. It should be noted that all these arcades run in the direction of the Kibleh or Mecca wall, and the same principle is observed at Kerouan, Cordoba, and other mosques built entirely for Mahomedan worship.

468. M. Coste makes all these arches pointed. M. de Prangey states that they are all circular; the truth being that they are partly one, partly the other.

469. Since then the arches have been built up, and it was for a time converted into a hospital. This now (1892) is under the care of the Commissioner for the preservation of ancient monuments, but is too far ruined to be long preserved.

470. See Coste’s ‘Edifices de Caire,’ p. 32, quoting from Makrisi.

471. ‘The Ancient Coptic Churches,’ by A. J. Butler, Oxford, 1884.

472. The marble wall decoration and the mosaics which are found in later mosques are of different design and execution from that found in Byzantine buildings; in fact as Mr. Butler remarks: “this form of art was borrowed by the Muslim builders, or rather was lent by the Coptic architects and builders, whom the Muslims employed for the construction of their mosques.” “Although the Saracens in Syria borrowed the art from Byzantium and used vitreous enamels for the decoration of their mosque walls, as well as for inlaying jewelry and steel armour on a smaller scale, yet the Mahomedans of Egypt never adopted any but the native or Coptic marble mosaic, partly because its unpictorial character suited their taste, and partly because they found, ready made, both art and artists—artists whose names have perished, but whose skill is still recorded in work of unexampled splendour which adorns the great Mosques of Cairo.”

473. The mosque cathedrals of Cordoba and Seville and the contemporary Arabic buildings. Transactions, R.I.B.A., 1882-83.

474. A view of it will be found in vol. ii. ‘History of the Modern Style of Architecture,’ 1891, p. 314.

475. To get it within the page, the scale of the plan is reduced to 200 French, or 212 English ft. to 1 in.

476. When the great national work, entitled ‘Monumentos Arquitectonicos d’España,’ is complete, this reproach will be removed, but that certainly will not be the case for ten or twelve years to come, if it ever does attain completion. The scale is too large, and the total want of principle on which it is carried out renders it useless till it is further advanced. Twenty-three numbers are published, but not one important building is complete, and, excepting a plan of Toledo, not one of the larger buildings is even attempted—Cosas d’España.

The above note was written twenty-five years ago and is true now, except that the twenty-three must be now eighty-nine, where it stopped nine years ago.

477. Alcazar = el-Kasr, “the Castle.”

478. A perfect copy of this court was reproduced by Mr. Owen Jones at the Crystal Palace in 1854. Except being slightly curtailed in plan, every detail and every dimension is identical with the original.

479. Nothing need be said here of La Cuba and La Ziza, and other buildings in Sicily, which, though usually ascribed to the Saracens, are now ascertained to have been built by the Normans after their conquest of the island in the 11th century. They are Saracenic in style, it is true, and were probably erected by Moslem artists, but so were many churches and chapels in Spain, as mentioned above; and I am not aware of any building now extant there which can be safely ascribed to the time when the island was held by the Moslems, or was then erected by them for their own purposes. Till that is ascertained, Sicily of course does not come within the part of our subject which we are now considering.

480. Plate lxxxii.

481. For the plan and section of this mosque I was indebted to the kindness of my friend, the late M. C. Texier, who placed his MS. plans at my disposal for the purpose of being engraved for this work.

482. For the plan of this building I am indebted to the unpublished drawings of the late M. C. Texier.

483. The steps by which the transformation may have been arrived at, passing through the traditional method of constructing vaults in plaster, which is still practised in Persia, were suggested in an article contributed to the Proceedings of the R. I. B. A., 1888, vol. iv., new series.

484. Both the plan and view are taken from Baron Texier’s ‘Arménie et la Perse,’ which gives also several coloured plates of the mosaic decorations, from which their beauty of detail may be judged, though not the effect of the whole.

485. The earliest attempt in this direction that I am acquainted with is the great portal of the palace at Mashita (Woodcut No. 268).

486. Texier, from whose work the illustrations are taken, ascribes the building to another Khodabendah of the Sufi dynasty, A.D. 1577-85. Our knowledge, however, of the style is sufficient to show that the monument must be 200 or 300 years older than that king; and besides, the Sufis, not being Tartars, would not build tombs anywhere, much less in Sultanieh, where they never resided.

487. ‘Travels,’ vol. i. p. 277.

488. Ker Porter’s ‘Travels,’ vol. i. p. 432 et seq. I cannot help suspecting that there is some mistake about these dimensions—they seem excessive. The Piazza of St. Mark’s at Venice, which resembles it more than any other area, is only 560 ft. long, with a mean breadth of about 250 ft. Probably 1500 feet by 500.

489. ‘Views of monuments in Central America, Chiapas, and Yucatan.’ 25 plates, folio. London, 1844.

490. ‘Incidents of Travel in Central America and Yucatan,’ by J. L. Stephens. 1st and 2nd series, 4 vols. 8vo. Murray, 1841, 1843.

491. The evidence collected by the Abbé Brasseur de Bourbourg, ‘Voyage de Tehuantepec,’ seems, if it can be depended upon, to confirm this idea.

492. Ausland, 1845, Nos. 165, 168.

493. D’Eichthal, ‘Revue Archæologique,’ vol. x. 1864, p. 188, and following numbers.

494. Sir Stamford Raffles’s ‘History of Java,’ vol. ii. p. 51.

495. ‘Anahuac,’ by Edward B. Tylor, 1861; pp. 188, 194.

496. The plate published by Humboldt, representing one of the bas-reliefs, is so incorrect as to be absolutely worthless.

497. There is a celebrated bas-relief on the back wall of a small temple at Palenque, representing a man offering a child to an emblem very like a Christian cross. It is represented in the first series of the ‘Incidents of Travel,’ vol. ii. p. 344. None of the sculptures have given rise to such various interpretations; but nothing would surprise me less than if it turned out to be a native mode of representing a Christian baptism, and was therefore subsequent to the conquest.

498. Since the first edition of this work was published, a folio work has appeared in Paris, entitled ‘Les Ruines de Palenque,’ illustrated by plates, made under the superintendence of M. de Waldeck, with text by the Abbé Brasseur de Bourbourg. The text is certainly not to be trusted. The plates add little to what we learn from Catherwood’s drawings, and I do not feel sure how far that little is to be depended upon.

In so far as they go they confirm the idea of the famous cross bas-relief being of Christian origin.

499. It is only fair to state that Mr. Markham (Journal Roy. Geo. Soc., vol. xli. p. 307) denies the Aymara origin of the Tia Huanacu ruins, and ascribes them to the Incas, and consequently disputes the distinction pointed out above. The truth seems to be that, until we get more photographs or detailed drawings, all conclusions regarding Peruvian architecture must be considered as more or less hypothetical.

500. For the principal part of this information I am indebted to Mr. William Bollaert and the photographs of the Messrs. Helsby, of Liverpool, and also to a paper on the Aymara Indians, by Dr. David Forbes, communicated to the Ethnological Society of London in June 1870.