Note: As well as being available individually, this item is also included in the Antioch Gate Islamic Golden Age Library on DVDROM, priced £49.95 plus £4 P&P - see our main page for details. To purchase the DVDROM, you may use the electronic PayPal service (press the DVDROM's "Add to Cart" button below):

Or you may post a cheque, Draft Order or Postal Order made payable to Antioch Gate, addressed to:

Antioch Gate

PO Box 14507
Birmingham
B25 9ET
UNITED KINGDOM

Note: The sample page images from the text, shown above, are of deliberately reduced quality

______________________________________

MEMOIRS OF THE ARCHÆOLOGICAL SURVEY OF INDIA


No. 15

THE DRAWING OF GEOMETRIC
PATTERNS IN SARACENIC ART

 

BY

E. H. HANKIN, M.A., D.Sc.


CALCUTTA: GOVERNMENT OF INDIA
CENTRAL PUBLICATION BRANCH
1925

______________________________________

CONTENTS

Preface

Contents

List of Plates - p. i
References - p. iii
Introduction - p. 1

1. Hexagonal patterns - p. 4
2. Octagonal patterns - p. 6
3. Geometrical Arabesques - p. 10
4. Floral Arabesques - p. 23
5. Decoration of domes - p. 23

Index - p. i (new section)
Plates
 

LIST OF PLATES
 

1. Hexagonal patterns.
 

PLATE I
Figure.
1. Oblong of 60 degrees
2. Panel suitable for three repeats of 60 degree oblongs
3. Pattern of hexagons in contact
4. Pattern of six-rayed stars and hexagons
5. Pattern drawn in perspective
6. Pattern of hexagons and arrow-heads
7. Hexagonal pattern from Bourgoin

PLATE II
8. Hexagonal pattern from a tomb near Delhi
9. Pattern of large and small hexagons
 

2. Octagonal patterns.
 

10. Method of drawing octagons in contact by sides
11. Pattern formed by octagons in contact by sides
12. Pattern of superposed octagons
13. Method of drawing octagons in contact by their angles
14. Pattern formed by octagons in contact by angles
15. Method of drawing overlapping octagons
16. Pattern of octagons overlapping by two sides
17. Pattern derived from octagons in contact
18. Method of drawing pattern of Fig. 17

PLATE III
19. Pattern of trellis screen in a window of the Daftar Khana, Fathpur-Sikri
20. Construction of an octagonal star and octagons of two sizes as frequently found together in octagonal patterns
21. Octagonal pattern used in inlaid marble in the tomb of Itimad-ud-Daula, Agra
22. Pattern of pavement in the central chamber of the Taj Mahal
23. Pattern from the Court of the Lions in the Alhambra

PLATE IV
24. Pattern derived from overlapping octagons, from the screen in the tomb of Itimad-ud-Daula, Agra
25. Pattern derived from octagons, in contact, from the vestibule of Akbar's tomb at Sikandara

PLATE V
26. Pattern of a screen on the outside of the mosque at Fathpur-Sikri
27. Pattern, in room under the Khwabgah at Fathpur-Sikri
 

3. Geometrical Arabesques.
 

28. Arabesque pattern based on octagons in contact
29. Arabesque pattern based on octagons and irregular pentagons
29 (A). Design from the Taj Mahal Hotel, Bombay

PLATE VI
30. Pattern based on dodecagons, squares and equilateral triangles
31. Pattern based on dodecagons separated by equilateral triangles

PLATE VII
32. Arabesque pattern based on decagons and pentagons
33. Border of large panel in Birbal's House at Fathpur-Sikri
34. Arabesque pattern based, on 14-gons, from the entrance to the mosque in Fathpur-Sikri

PLATE VIII
35. Pattern based on 14-gons and twinned heptagons
36. Arabesque pattern based on 16-gons, 12-gons and heptagons

PLATE IX
36 (A). Arabesque pattern based on 16-gons, 12-gons and heptagons
37. Arabesque pattern based on decagons and spindle-shaped figures
38. Pattern from the soffit of an arch in the Hakim's bath at Fathpur-Sikri

PLATE X
39. Arabesque pattern containing fifteen-pointed stars
40. Arabesque pattern containing 18-pointed stars

PLATE XI
41. Pattern containing seven-pointed stars
42. Pattern based on octagons and irregular heptagons
 

4. Floral Arabesques.
 

PLATE XII
43. Floral arabesque with decagonal pattern from which it is derived
44. Design from Gismah mosque, Cairo. (Bourgoin, Pl. 44) —Pattern occurring in—
 

5. Decoration of domes.

PLATE XIII
45. Decagon and a decagon with two triangles cut out
46. Design from a dome in Fathpur-Sikri
47. Decoration of a half-dome in the Samosa Mahal, Fathpur-Sikri
48. Decoration of a half-dome in the Agra Gate of the outer wall of Fathpur-Sikri
49. Diagram of the structure of a dome of increasing curvature
50. Design from the interior of the dome surmounting the cupola on the Hiran Minar, Fathpur-Sikri
 

Agra Club.

PLATE XIV
51 and 52. Interior of the Agra Club, showing Saracenic designs adapted to modern decoration

______________________________________

From the INTRODUCTION

ONE of the main characteristics of Saracenic art is its universal employment of geometric patterns often of amazing complexity. A reason for the employment of such patterns is that the portrayal of living things was forbidden by the Muhammadan religion. That a geometric pattern readily becomes monotonous by repetition goes without saying. Variety, therefore, was indispensable, and this desire for variety inevitably led to the discovery of new and complicated designs, so subtly complicated that it is hardly credible in some cases that the ordinary person could appreciate their nicety or distinguish, for example, between a pattern that contained 15-pointed and one that contained 18-pointed stars, or understand the purpose of changing a scheme of 10-pointed stars to one of eleven ... dating from the earliest period we have Qusair 'Amra found by Musil. This building, a royal bath and resting house, is of simple plan; it is roofed with tunnel vaults and a small dome, and, most remarkable of all, is decorated with figures, painted by Byzantine artists, including one of the Khalif enthroned; there are inscriptions in Greek and Arabic (date 712-715 A.D.) but everything is quite Byzantine. Slightly earlier is the Dome of the Rock at Jerusalem, which was once decorated without as well as within with Byzantine gold, mosaics, and the great mosque at Damascus was decorated in the same way ... Ziad ibn Abihi, Governor of Basra ... about 60 A.D. employed Persian workmen, and among them a man who had been a builder to Chosroes, who recommended the cutting of columns from Ahwaz marble ... The only truly Arab period was the Umayyad, and this dynasty came to an end in 750 A.D. After this the Abbasids came and founded Baghdad, the whole centre of gravity was displaced ... After the fall of the Umayyad dynasty Persian influence dominated ...

It is surprising that, despite the complexity of Saracenic patterns, the geometrical knowledge required either for drawing or for designing them is small. Any one who can draw one line perpendicular to another, who can describe an equilateral triangle, and who can bisect an angle, is capable of copying these patterns and, with the methods about to be described, of designing new ones. Formal methods of making pentagons or heptagons, such as may be found in works on geometrical drawing, probably were not employed. If it is required to draw such figures, or any other polygons, all that is necessary is to describe a circle and to divide its circumference into the requisite number of equal parts, by trial and error, with the help of ordinary dividers.

That the builders of the Taj at Agra were incapable of drawing the particular class of pattern about to be described as "geometrical arabesque" seems probable. The method of drawing such patterns is quite unknown at the present day in India and during a visit to Cairo, some years ago, I found no evidence that it was known to the Egyptian workmen. They were making beautiful products of Saracenic art, but appeared never to attempt to reproduce the more complicated patterns that had been used by their predecessors. Lack of knowledge of the methods appears also to handicap European artists when copying the more elaborate achievements of Saracenic art ...

Of Saracenic patterns those made on hexagonal and on octagonal bases are drawn with ease. Any one who will devote a little trouble to the subject can without difficulty find methods as good as or better than those described below. But there is another class of patterns of the first importance in Saracenic art that is peculiar to this school, and which I propose to designate by the name "geometrical arabesque". The original method of constructing these patterns has long been forgotten and in its absence the work of reproducing them is most laborious and difficult. During visits to Fathpur-Sikri many years ago, I spent much time in measuring the angles and making tracings of these designs but always failed to find any rational scheme by which they could be constructed. At last, by good fortune, I happened to enter a small Turkish bath attached to Jodh Bai's Palace. It had previously been inhabited by Indians, who had only just been evicted, and I was probably the first Englishman to visit the place. In one of the rooms of this bath was a half-dome decorated by a straight line pattern. In addition to the pattern, some faint scratches were discovered on the plaster. Obtaining a table and chair and a piece of tracing paper I succeeded in making a copy. On closer examination these scratches were found to be parts of polygons, which, when completed, surrounded the star-shaped spaces of which the pattern was composed, and it turned out that these polygons were the actual construction lines on which the pattern was formed. As will be explained in detail below, in making such patterns, it is first necessary to cover the surface to be decorated with a network consisting of polygons in contact. Then through the centre of each side of each polygon two lines are drawn. These lines cross each other like a letter X and are continued till they meet other lines of similar origin. This completes the pattern. The original construction lines are then deleted and the pattern remains without any visible clue to the method by which it was drawn.

After considerable labour in working out details I have at length elaborated the method by which one can draw complicated arabesque patterns and even design new ones. The room in which this clue was found is very dark, so perhaps this is a reason why the artist carelessly forgot to obliterate his construction lines, which have lasted for three and a half centuries and now give us an insight into a forgotten art.

______________________________________

QUOTES

p. 4: The space to be decorated must be so proportioned that it will contain a whole number of such oblongs. For instance, the panel shown in Fig. 2 contains three of them. If the space should be a little wider or a little higher, so that the three oblongs do not exactly fill it, it will be unsuitable for a hexagonal pattern, unless a large number of sixty-degree oblongs drawn to a smaller scale can be made to produce an exact fit.

p. 8: An example of the use of large and small octagons together is shown in Fig. 21. The panel is divided into squares in which are drawn octagons in contact by their sides; these form the large octagons of the pattern and are shown at A in heavy lines. Alternate octagons are omitted by leaving out some of the lines, which are indicated by dots.

p. 12: That this treatment is possible is due to the fact that in these designs, "dead ends" rarely occur because, as a rule, each line zigzags its way through the pattern till it ends at the margin. Thus each line takes a part in the formation of several pattern-spaces. In a pattern containing many "dead ends" each line only circumscribes one pattern-space and those with which it is immediately contiguous, thus causing the design to look like a number of pieces put together rather than one continuous pattern. In Saracenic art the artists endeavour to decorate a surface by modifying it, rather than by concealing it with struck-on ornamentations. This probably is one reason for their convention that a whole number of repeats of the pattern must be used in a decorated panel.

An interesting example of how arabesque patterns should not be drawn is shown in Fig. 29A. The pattern occurs in each of several panels in the ceiling of a room in a large hotel in Bombay. The designer, with misapplied ingenuity, had attempted to make the pattern last described fit a space for which it was quite unsuited and in this attempt made the following mistakes ...

p. 15: In earlier Saracenic work the pattern is drawn as described above, but, in later work lines such as OP and XZ are not drawn exactly parallel but approach each other outwardly. This variant of the pattern, so far as my experience of it goes, is that met with in India.

Click here to return to the main page