STUDY ON THE RECONSTRUCTION OF THE FIRST IMPERIAL AUDIENCE HALL, HEIJO PALACE（2009）

Research Report Vol. 80

Nara National Research Institute for Cultural Properties

 

 

4 Roof/Rooftiles

 

English Contents and Surmmary

 

Study on the reconstruction of the primary Daigokuden Hall, Nara Palace 4 Roof/Rooftiles

 

Contents

 

Chapter I Process of review and creation of report

1 Process of review    1

2 Creation of report    7

 

Chapter II Excavated tiles

1 Comparative determination of the tiles required for the Daigokuden Hall and their form/dimensions    9

2 Surface conditioning of tiles required for the Daigokuden Hall   15

 

Chapter III Case study

　1 Noshi tiles and Mendo tiles excavated from the Nara Palace

A Introduction    18

B Noshi tiles    18

C Mendo tiles    21

D Conclusion    22

 

　2 Flat eave tiles with a vermilion line excavated from the Nara Palace as well as the projection of the eaves

A Introduction    23

B Practical examples of the Nara Palace/Heijyo-kyo    23

C Centering around the flat eave tiles with a vermilion line and the projection of the eaves    23

 

　3 Hip rafter covering tiles and hip rafters

A Introduction    26

B Outline of the of hip rafter-covering tiles    26

C Hip rafter-covering tiles of the Nara Palace    28

D Hip rafter-covering tiles of Yakushiji Temple    29

E Hip rafter-covering tiles of the Primary Daigokuden of the Nara Palace and the calculation of the width of hip rafters    30

F Reconstruction of hip rafter-covering tiles and hip rafters   31

 

　4 Rafter-end tiles and rafters

A Introduction    33

B Architectural data/information related to rafter ends    33

C Case examples of rafter-end tiles    34

D Conclusion    38

 

　5 Tile color tone

A Introduction    39

B Measurement of color tone of roofing tiles excavated from the Primary Daigokuden Hall site    39

C Surface observation (microscopic observation)    41

D Measurement of color tone of reconstructed tiles    42

E Summary    44

 

　6 Black burnished tiles of China

A Introduction    46

B Black burnished tile    46

C Dark blue glazed tile    52

 

　7 Chinese shibi/shifun

A Introduction    54

B “Shibi” and “Shifun” seen in written data/information    54

C Shibi and Shifun excavated    55

D Change in shape and manufacturing method of shibi/Shifun    60

E Summary    64

 

　8 Collecting shibi-related data in Nara, Osaka, and Kyoto    66

 

9 Ridge form in ancient China, Korea, and Japan

- Focusing on the flat descending ridge -

A Data/information of pre-Sui/Tang Dynasty of China    106

B Case examples in Korea    110

C Data/information about the pre-Middle Age of Japan    112

D Ancient flat descending ridge    114

 

10 Position and fitting of the antefix (terminal ridge-end tile) of the descending corner ridge

A Case examples in China    116

B Case examples in Korea    119

C Case examples in Japan    119

D Corner descending ridge and corner fitting   122

 

11 Shape and peripheral fitting of shibi

A Position of flat descending ridge    123

B Position of ogami-tomoe (round eave) tiles    125

 

12 Tile-molding techniques and trial production of tiles

A Surface conditioning and tile color    128

B Tile contraction percentage     132

C Tile firing using climbing kiln    133

 

Chapter IV Process of reconstruction designing

1 Allocation, dimensions, and specifications of tiles

A Distance between roofing tiles    136

B Dimensions    139

C Surface conditioning    140

 

2 Shape and dimensions of ridge    141

 

3 Fitting for anterior section of corner descending ridge    146

 

4 Fitting of keraba and shibi    148

 

5 Reconstruction of shibi

A Introduction    151

B Shape    151

C Size    151

D Design    151

E Conclusion    152

 

Chapter V Reconstruction plan    153

 

Contents and summary in English

 

Summary

In 710, the national capital was constructed in the area of the present Nara City. The Nara Palace was located at the center of the northern area of the capital, extending approx. 1.3 kilometers in the east-west direction and one kilometer in the north-south direction. The Nara Palace included not only the imperial residence, but also spaces for rituals and the affairs of State. During the first phase of construction of the Nara Palace, the Daigokuden Hall (“primary Daigokuden Hall”) was built in the center. The Daigokuden Hall was the main building of this Daigokuden Hall complex. The national Capital, however, was relocated to the Kunikyu Palace in 740. At that time, the Daigokuden Hall was also moved to the Kunikyu Palace, and, in later years, reused as the main hall of the Yamashiro-kokubunji Temple (lit. the Buddhist temple of Yamashiro district). However, Daigokuden Hall was dismantled in the intervening years. Although the national capital was relocated again to Heian-kyo in 745, the site on which the Daigokuden Hall was originally built had already been occupied by a new palace; while the platform of the Daigokuden Hall had been largely removed. In 784, the capital was relocated thrice to the Nagaoka-kyo place, and the historic role of Nara Palace terminated. Since then, in the early 9th century, another palace was transiently built at this site, but it continued for only about 15 years before being abandoned. The Nara Palace site has, as a result, reverted to agricultural land use.

In 1922, the Nara Palace site was designated as a national historic site. Since 1952, subsequently, the Nara National Research Institute for Cultural Properties (“NNRICP”) has continuously undertaken excavation survey research. Furthermore, from 1963, public ownership of the land has proceeded in order to improve the site as a historic site park. With regard to the primary Daigokuden Hall, the NNRICP has implemented related excavation surveys several times, and published an excavation survey report in 1982. In addition, as the part of the total improvement project of the Nara Palace site, the reconstruction of the Daigokuden Hall was planned. After 1989, the Daigokuden Hall reconstruction plan commenced its specific review; and, in 1993, the reconstruction principles for the Daigokuden Hall were determined. Since then, after the completion of studies and designs for reconstructing the Daigokuden Hall, reconstruction commenced in 2001. This work will be completed in March, 2010.

This report explains the studies implemented by the NNRICP including the grounds for reconstruction designs for the Daigokuden Hall. The report comprises four volumes, explaining the following items:

1 Platform/Foundation

2 Wooden parts

3 Coloring/clasps and clamps

4 Roof/tiles

　While, each volume includes the following content:

I Process of review and creation of report

II Direct data/reference materials relating to Daigokuden Hall of the Nara Palace

III Case study

IV Process of reconstruction

V Reconstruction plan

　he content of the case study indicate details of the basic research implemented in parallel with the reconstruction review. In addition, some findings of studies made after the completion of the reconstruction design are also included. Incidentally, please note that the reconstruction plan shown in the report only indicates the theoretical form of the ancient architecture, and that the building as it will actually be reconstructed has been changed slightly, taking into consideration factors of structural reinforcement and maintenance.

 

　This report is the fourth of four volumes, explaining mainly the results of studies on the form, size, and color of roof tiles, as well as tiling method and ridge types.

Since tiles presumed to be those actually used for Daigokuden Hall in ancient times were found during excavation surveys, it is possible to reconstruct such tiles, using their basic characteristics, such as shape, size, etc. Nevertheless, as the roofs of all the existing ancient architecture have since been retiled, how the tiles were arrayed and how the ridges were laid out in ancient times remains obscure. Therefore, we figured out the tiling techniques of those ancient times by collecting and summarizing indirect historical data. With regard to the surface finish of the tiles, we reviewed the surface-finishing and firing methods by researching excavated tiles and similar examples. Although the shibi, ornamental roof-ridge talisman-tiles placed on both edges of a ridge, can be presumed to have been gold copper, no shibi have been found at the Nara Palace site. Accordingly, we reconstructed the shibi of the Nara Palace, consulting designs of shibi found at other ancient Buddhist temple sites.

We studied similar examples not only for the subject matter directly leading to the design of the roof tiles but also for understanding the entire range of techniques of that time. The objects of our study and a brief outline of each follows.

　1. Noshi tiles (flat ridge tiles stacked on the ridge) and mendo tiles (gap-filing tiles) excavated from the Nara Palace

We reviewed these two types of the tiles excavated from the Nara Palace: noshi tiles (used on the ridge) and mendo tiles (used under the ridge). Through these reviews, we found that: (i) there are two types of Noshi tiles; the first are those originally produced as Noshi tiles, and the second are those converted from flat tiles divided into halves, and used as Noshi tiles; (ii) there are two types of Mendo tiles; the first are those produced by reshaping round tiles to Mendo tiles immediately before firing, and the second are those converted from ordinary round tiles divided into halves, and used as Mendo tiles.

　2. Flat eave tiles with a vermilion line excavated from the Nara Palace, as well as a projection of the eaves

We reviewed excavated flat eave tiles, picked out those upon the reverse side of which a vermilion line remains, and checked the positions of the vermilion lines. The vermilion lines were made when red paint applied to the tile fillets (on which the tiles were laid) adhered to the tiles. Accordingly, a vermilion line on the reverse side of a tile shows the position of the tile fillet, proving how long the flat eave tiles projected from the position of the tile fillet in ancient times. After reviewing the position of such vermilion lines, we found that the distance from the tile fillet to the flat eave tiles was approx. 9.5 centimeters.

　3. Hip rafter-covering tiles and hip rafters

In the neighboring areas of the Daigokuden Hall in the Nara Palace, hip rafter-covering tiles were found. However, as hip rafter-covering tiles conceivably used for the Daigokuden Hall have not been found, we reconstructed the hip rafter-covering tiles by reviewing such tiles excavated from neighboring areas of the Daigokuden Hall and from other Buddhist temples.

　4. Rafter-end tiles and rafters

At such excavation sites as ancient Buddhist temples, the rafter-end tiles fixed at the rafter edge have been found. However, such rafter-end tiles have not yet been found around Daigokuden Hall, and, the rafters of Daigokuden Hall were conceivably fixed with some metal ornament. In any case, it was possible for us to calculate the shape and dimensions of the rafters of this ancient architecture by reviewing the rafter-end tiles. Thus, we successfully collected the data/information for the reconstruction of such rafters after evaluating and reviewing the rafter-end tiles of the period from the seventh to eight centuries, and provided this data/information to the public agencies that are engaged in physical reconstruction work.

　5. Color tone of tiles

Since the color tone of the tiles, found through excavations and conceivably used for the Daigokuden Hall, vary considerably from one to the next, we examined what tile color the ancient peoples intended to prepare when building Daigokuden Hall. We examined many tiles including those excavated through surveys not only from Daigokuden Hall but also from ancient Buddhist temples and even from ancient China, in order to review color patterns based on a scientific index. As a result we theorize that ancient people intended to prepare black color tiles for Daigokuden Hall.

　6. Black burnished tiles of China

Based on our hypothesis that the tiles of Daigokuden Hall were intended to be black, we researched and reviewed the production method by which the black color of tiles used for high-ranked buildings in ancient China was produced.

　7 Chinese Shibi/Shifun (shifun: lit. “owl mouth” originating from one of the Chinese nine young dragons. Another ornamental roof-ridge talisman-tile to swallow all evil influences)

For the reconstruction of Daigokuden Hall’s shibi, we studied the history and changes in the shape of a Chinese shibi, consulting excavated articles, artifacts, and paintings as basic data.

　8 Collecting shibi-related data in Nara, Osaka, and Kyoto

For the reconstruction of Daigokuden Hall’s shibi, we collected data/information of shibi found in Nara, Osaka, and Kyoto, and listed it.

　9 Ridge form in ancient China, Korea, and Japan

How roofs were tiled and how the tiles were laid out when forming ridges in ancient Japan remain obscure. Accordingly, we analyzed the form of ancient ridges based on excavated articles, artifacts, and paintings of China. Korea, and Japan, and hypothesized the ridge form of ancient Japanese architecture.

　10 Position and fitting of the antefix (terminal ridge-end tile) of the descending corner ridge

A tiled roof can cause rain leaks or the fall of flat eave tiles on corners, unless each composition unit of the roof corner is well fitted. Accordingly, in the same way as Paragraph 9 above, we analyzed the ancient ridge form and tile roofing based on excavated articles, artifacts, and paintings of China, Korea, and Japan, and arrived at a number of conclusions.

11 Shape and peripheral fitting of shibi

　The shibi is fitted with the ridge and tiles from many directions. The original architectural function of the shibi is conceived to have been covering the ridge and tiles fitted from many directions. Accordingly, we created drawings, and made a model of the shibi fitted with the ridge and tiles in order to review them, while calculating the size of the shibi. The research results indicated the possibility that the width of the shibi does not become larger in proportion to the scale/size of a building. We also examined the fitting of the peripheral areas of the shibi.

12 Tile-molding techniques and trial production of tiles

We executed a trial production of tiles, using tile-molding techniques hypothesized from excavated tiles. We used both modern tile kilns and reconstructed ancient kilns for the trial production. In addition, we checked how the pre-firing, conditioned surface of the tiles changes after firing. We also used several types of tile clay, in order to check the color development of each type of clay. Furthermore, we numerically converted dimensional changes (contraction percentage) and weight changes in each process of tile manufacturing. Subsequently, we backwards calculated the tile dimensions based on the contraction percentages, and figured out the dimensions of ancient tiles in the clay-molding step.

 

　　The shape/dimensions of the tiles used for the Daigokuden Hall was calculated, consulting tiles excavated from the area of Daigokuden Hall (Fig. V-1). Subsequently, using such tiles, we reviewed the distance between/overlapping of the tiles. In the following step, we reviewed the ridge shape based on the case studies, and determined the ridge shape and the tiles used for the ridge. With regard to the tiles used for reconstruction, we established the principle that only the tiles with the same form and dimensions of those found through excavation surveys should be used. Therefore, we used techniques that varied from the traditional techniques remaining today. For instance, in respect of the ridge, we adopted the ridge form comprising flat tiles laid on the top of the ridge, with round tiles overlaying the flat tiles. In addition, with regard to the corner part of the roof, we adopted the method in which flat tiles are laid upon round eave tiles at the corners. The quantity of the tiles required for reconstructing the roof was approx. 99,000.

 

2009年3月23日　発行

平城宮第一次大極殿の復原に関する研究

4      瓦・屋根

奈良文化財研究所学報　第80冊