Page:History of Indian and Eastern Architecture Vol 2.djvu/401

 CHAP. I. CIRCULAR PAGODAS OR CHAITYAS. 343 the result on the plan would be three angle projections ; an entrance porch D added on each face would give five angle projections, and so on. 1 The platforms on which the structure was raised would necessarily follow the same plan, and its repetition in the superstructure would result in the j-ikhara of Hindu origin which has usually three angle pro- jections. In the diagram here given the applied projecting plane on each face and its return are equal in dimension ; this is not usually the case, and sometimes the former is only about half the latter. The nearest approach to the diagram is that shown in the temple of Vat Sisavai at Sukhodaya in Siam (Plate XLVIL), where these angle projections form prominent features in the design. In the Shwe-Hmaudau pagoda (Woodcut No. 445) the plan of the platform is octagonal, and here the projecting planes, three in number on each face, have given five angles. An instance of its intro- duction in the superstructure is shown in the Abhayadana pagoda (Plate XXXVI.), where there are three project- ing angles in the elaborate cornice, carrying the finial. In ^the Seinnyet pagoda a 445 Q uart er- P lan of Shwe-Hmaudau Pagoda similar COrnice is more com- at Pegu. (FromSymes.) Sc.ale too ft. to i in. plicated, having seven project- ing angles and eight vertical fillets projecting one in front of the other. The principal variations made in the design of the Zedi are those of the relative proportion of the bell to the rest of the structure, the outline of the same and its superstructure, and the decoration employed. Thus in the Lokananda pagoda in Pagan, built by Anaurahta in 1059, the bell is of immense size, being three-sevenths of the total height of the structure, including the triple base and finial, and that is generally the characteristic of the earlier examples, but, where occasionally employed to crown the ^ikharas of the square temple, as in the Abhayadana temple, it is so small as to be scarcely recognisable. In the older pagodas of Pagan the several mouldings are all more or less convex in outline, but in later examples, and 1 These pagodas are generally described as polygonal on plan, a term which is misleading : multiplane would be more correct.