Journal of Archaeological Science 40 (201 (2013) 3) 1158 158e e1164
Contents lists available at SciVerse ScienceDirect
Journal of Archaeological Science j o u r n a l h o m e p a g e : h t t p : / / w w w . e l s e vi vi e r . c o m / l o c a t e / j a s
Quarries Quarries and transporta transportation tion routes of Angkor Angkor monument monument sandstone sandstone blocks blocks Etsuo Uchida a , Ichita Shimoda b ,
*
a Department of Resources and Environmental Engineering, Faculty of Science and Engineering, Waseda University, Ohkubo 3-4-1, Shinjuku, Tokyo 169-8555, Japan b Department of Architecture, Faculty of Science and Engineering, Waseda University, Ohkubo 3-4-1, Shinjuku, Tokyo 169-8555, Japan
a r t i c l e
i n f o
Article history:
Received 24 May 2012 Received in revised form 21 September 2012 Accepted 26 September 2012 Keywords:
Angkor monuments Sandstone Quarry Transportation Transportation route Canal Cambodia
a b s t r a c t
Sandstone blocks are the main construction materials used in the Angkor monuments in Cambodia. However, a thorough study of the quarries has not yet been carried out. We conducted a � eld investigation of sandstone quarries from the Angkor period at the southeastern foot of Mt. Kulen, which is approximately 35 km northeast of the Angkor monuments. As a result, we discovered more than 50 sandstone quarries. On the basis of the measurements of magnetic susceptibilities and thicknesses (step heights), we found that they were quarried at different times. These four quarrying areas were identi �ed as the quarries D to G inferred by Uchida et al. (2007). (2007) . In addition we investigated a canal that was � identi ed on satellite images, connecting quarry sites at the foot of Mt. Kulen to the Angkor monuments. The �eld investigation suggests a high probability that the canal was used for the transportation of sandstone blocks from Mt. Kulen. 2012 Elsevier Ltd. All rights reserved.
1. Introduction
Sandstone, laterite, and brick are the main construction materials in the Angkor monuments. Three types of sandstones are found in the monuments, but a gray to yellowish brown sandstone (feldspathic (feldspathic arenite) arenite) is most common among them ( Uchida et al., 1998). 1998 ). In this study, we conducted an investigation of quarries and transportation routes of the sandstone. The gray gray to yellow yellowish ish brown brown sandst sandstone one consist consistss mainly mainly of quartz, plagioclase, potassium feldspar, feldspar, biotite, muscovite, and rock fragments of approximately 0.2 mm diameter. No differences were found in constituent minerals and bulk chemical compositions of the sandstone among the monuments ( Uchida et al., 1998; 1998; Kucera et al al.,., 200 20088). Howeve Howeverr, the sandsto sandstone ne showe showed d differe difference ncess in magnetic susceptibility. On the basis of the magnetic susceptibility, Uchida et al. (1998, 2003, 2007) inferred 2007) inferred seven different sandstone quarri quarries es from from the Angkor Angkor period. period. In additio addition, n, the magnet magnetic ic susce suscept ptib ibili ility ty meas measur urem emen ents ts made made it possib possible le to estim estimat atee construction periods and sequences of the monuments. The gray to yellowish brown sandstone was derived from the Terrain Rouge Formation (the Phu Kradung Formation in Thailand) of the early to middle Jurassic (Garnier, ( Garnier,187 18733; Saurin, 19 1954 54;; Delvert, 1963;; Meesook 1963 Meesook et al., 2002). 2002). The Terrain errain Rouge Rouge Format Formation ion is distributed widely in eastern Cambodia in addition to Thailand,
*
Corresponding author. author. Tel.: þ 81 3 5286 3318; fax: þ 81 3 5286 3491. E-mail address: weuchida@w
[email protected] aseda.jp (E. (E. Uchida).
0305-4403/$ e see front matter 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jas.2012.09.036
Laos, and Vietnam. In Thailand, there are no Khmer monuments which used the gray to yellowish brown sandstone as a construction materi material. al. Howe However ver,, it was was used for Wat Wat Phu in Laos Laos and My Son Son in Vietnam. In northern Cambodia, the gray to yellowish brown sandstone was used for major monuments such as Koh Ker, Preah Khan of Kompong Svay, and Banteay Chmar as well as the Angkor monume monuments nts (Uchida et al., 2010 2010). The The nearest nearest outcrop outcrop of the the gray to yellowish brown sandstone to the Angkor monuments is located at the southeastern foot of Mt. Kulen. Several sandstone quarries of the Angkor period have been previously discovered there ( Garnier, 1873;; Saurin, 1954; 1873 1954; Delvert, 1963; 1963; Carò and Im, 2012). 2012 ). However, a thorough study of the quarries has not yet been done. We therefore therefore conducted conducted a �eld invest investiga igation tion at the southea southeaste stern rn foot foot of Mt. Kulen. In addition, on the basis of satellite images (Google earth), traces of a canal (consisting of bank, canal and river) connecting quarry sites at the southeastern foot of Mt. Kulen to the Angkor monuments were identi �ed. Under the hypothesis that the canal was used for the transportation of sandstone blocks, we conducted a �eld investigation of this canal. This paper reports results of the � eld investigation. 2. Sandstone Sandstone quarrie quarriess
Garnier (1873) � rst postulated that the gray to yellowish brown sandstone sandstone used in the Angkor monuments monuments was supplied from the southeastern foot of Mt. Kulen. Afterwards Delvert (1963) conducted a �eld investigation of sandstone quarries. He found several
E. Uchida, I. Shimoda / Journal of Archaeological Science 40 (2013) 1158 e1164
quarries at Trapeang Thmar (Thmo) Dap, Au Mealea, Phnom Bey (Bei), and other areas. The distribution of sandstone quarries is roughly shown in a map by Boulbet (1979). Recently Rocks (2009) and Carò and Im (2012) reported several sandstone quarries from the Angkor period. However, the number of sandstone quarries discovered until now is not enough for the construction of all the Angkor monuments. Thus it is expected that there are abundant undiscovered sandstone quarries. In this context, we conducted a � eld investigation of sandstone quarries at the southeastern foot of Mt. Kulen. As will be discussed later in detail,an eastewest linear structure 2.4 km long was observed at the southeastern foot of Mt. Kulen on satellite images. Under the hypothesis that this linear structure may be a transportation route for the sandstone blocks, we conducted a � eld investigation around it. We discovered more than 50 sandstone quarries of various scales in the area. The locations of each quarry are indicated in Fig. 1. Photos of representative quarries are shown in Fig. 2. Most of the quarries discovered to date are distributed in the eastern to northern areas of the linear structure mentioned previously. The scale of the quarries ranges from 10 to 50 m in width. As mentioned previously, there are no differences in constituent minerals and bulk chemical compositions of the gray to yellowish brown sandstone among the Angkor monuments, but the average magnetic susceptibilities changed over time as elucidated by Uchida et al. (1998, 2003, 2007). Therefore a magnetic susceptibility measurement was conducted in each quarry in order to deduce the period in which each quarry was used. The magnetic susceptibilities were measured randomly at 5e25 places in each quarry depending on its scale. The measurements were carried out on surfaces as �at as possible, avoiding weathered surfaces. The average values are summarized in Table 1. In addition to the
1159
magnetic susceptibilities, shape, especially thickness, of the sandstone blocks also changed over time (Uchida et al., 2005). Therefore step heights of the quarry sites instead of thicknesses of the sandstone blocks were also measured (Fig. 1 and Table 2). Before the Angkor Wat period, sandstone blocks had a square end and a thickness of 40e50 cm, except for some sandstone blocks in the Preah Ko and Bakheng periods. After that, the thickness of the sandstone blocks tended to thin over time, that is, to 30 e35 cm in the Angkor Wat period except for Angkor Wat itself and Wat Athvea (40e55 cm), and to 20 e30 cm in the Bayon period (Uchida et al., 2005). Although these values are thicknesses of the sandstone blocks used in the monuments, the thickness of the sandstone blocks cut from the quarries is estimated to have been at least 10 cm thicker than these values. On the basis of the average magnetic susceptibilities and thicknesses (step heights), the investigated sandstone quarries can be classi�ed into four areas, shown in Fig. 1. Fig. 3 shows histograms of magnetic susceptibilities of the sandstones for Areas A to D. Area A is along the eastern part of the previously mentioned linear structure, and corresponds to the areas Toek Lick and Thma Andong identi�ed by Carò and Im (2012). In Area A, the average magnetic susceptibilities and thicknesses (step heights) of the sandstone blocks range from 1.6 to 2.5 103 SI units and from 50 to 60 cm, respectively. These data suggest that the quarries in Area A were used in the Preah Ko or Khleang to Baphuon periods (including Pre Rup) and correspond to quarries A or D of Uchida et al. (2007) (Table 2). Taking into consideration that Bakong is the only large scale monument in the Preah Ko period, it is quite possible that the quarries in Area A were used in the Khleang to Baphuon periods and correspond to quarry D of Uchida et al. (2007).
Fig. 1. Topographic map showing the bank with an east-west orientation at the southeastern foot of Mt. Kulen and the distribution of sandstone quarries. The bank is considered to have been used as a transportation route. The numbers of quarries correspond to those in Table 1. On the basis of the magnetic susceptibilities and thicknesses (step heights) of the sandstone blocks, the quarries can be classi �ed into four areas (Areas A to D). MS: a range of magnetic susceptibilities ( 103 SI unit), and TH: a range of thicknesses.
1160
E. Uchida, I. Shimoda / Journal of Archaeological Science 40 (2013) 1158 e1164
Fig. 2. Photos of the sandstone quarries located at the southeastern foot of Mt. Kulen. (a) quarry No. 22 in Area A, (b) quarry No. 10 in Area B, (c) quarry No. 45 in Area C, and (d) quarry No. 31 in Area D.
Table 1
Positions (longitude and latitude) of the sandstone quarries at the southeastern foot of Mt. Kulen and average magnetic susceptibilities of sandstones at each quarry. For the average thickness (step height) of the sandstone blocks see Fig. 1 and Table 2. Quarry No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 a
Latitude 13 29 09.32 13 29 29.46 13 29 30.26 13 29 33.54 13 29 32.80 13 29 33.93 13 29 20.95 13 29 20.10 13 29 19.36 13 29 18.58 13 29 18.08 13 29 15.40 13 29 14.93 13 29 15.25 13 29 14.48 13 29 09.50 13 29 00.30 13 29 01.20 13 28 39.00 13 28 43.61 13 28 46.14 13 28 48.03 13 28 50.19 13 29 04.63 13 29 07.86 13 29 09.35 13 29 03.79 13 29 17.83
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
Longitude 104 12 19.53 104 12 18.80 104 12 19.56 104 12 08.89 104 12 07.86 104 12 07.24 104 12 16.45 104 12 18.05 104 12 18.42 104 12 17.06 104 12 17.86 104 12 16.76 104 12 18.39 104 12 19.39 104 12 18.34 104 13 21.50 104 13 04.50 104 12 17.40 104 12 25.75 104 12 31.80 104 12 33.44 104 12 41.82 104 12 43.86 104 12 42.55 104 12 42.55 104 12 41.04 104 12 27.66 104 11 05.23
N N N N N N N N N N N N N N N N N N N N N N N N N N N N
Average magnetic susceptibility ( 10
3
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
SI unit).
E E E E E E E E E E E E E E E E E E E E E E E E E E E E
Av. M.S. a 0.68 1.28 2.70 2.67 2.82 4.53 2.09 3.31 2.99 3.57 3.38 3.50 1.05 1.98 1.15 1.99 2.46 0.97 2.49 1.61 1.92 2.14 2.62 1.90 2.21 4.26 3.64 1.70
Quarry No. 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
Latitude 13 29 25.21 13 29 26.14 13 29 25.50 13 29 24.04 13 29 23.55 13 29 23.83 13 29 24.55 13 28 52.70 13 29 00.63 13 29 01.72 13 29 02.77 13 29 04.57 13 29 02.44 13 29 03.49 13 29 02.70 13 29 19.20 13 29 20.17 13 29 18.17 13 29 17.14 13 29 20.35 13 29 09.82 13 29 03.63 13 29 09.20 13 29 02.50 13 29 05.20 13 29 04.71 13 29 00.99 13 29 01.99
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
N N N N N N N N N N N N N N N N N N N N N N N N N N N N
Longitude 104 11 09.24 104 11 09.68 104 11 11.94 104 11 13.38 104 11 14.77 104 11 16.71 104 11 16.75 104 13 33.00 104 12 07.77 104 12 10.33 104 12 08.18 104 12 06.35 104 11 58.62 104 11 57.54 104 11 55.06 104 11 52.17 104 11 51.42 104 11 46.78 104 11 37.99 104 11 35.15 104 11 50.83 104 12 16.13 104 13 18.50 104 12 57.90 104 12 59.90 104 12 06.97 104 11 54.57 104 11 53.27
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
0
00
E E E E E E E E E E E E E E E E E E E E E E E E E E E E
Av. M.S. a 2.08 2.01 2.00 1.76 1.63 1.68 2.05 0.77 1.44 1.43 0.75 0.49 0.80 1.09 0.33 0.77 0.99 0.54 2.09 1.79 0.88 0.71 e e e e e e
E. Uchida, I. Shimoda / Journal of Archaeological Science 40 (2013) 1158 e1164
1161
Table 2
Correlation of Areas A to D of sandstone quarries discovered in this study to the Quarries A to G inferred by Uchida et al. (2007). Period
Inferred quarries A to Ga Quarry A Quarry B Quarry C Quarry D
Average magnetic susceptibilities in quarries A to G (10 3 SI unit)a 1.1 e2.3 1.0e9.0 2.3 e3.0 1.1e2.4
Thickness in centimeters of sandstone blocks used in the monumentsb 30e50 20e45 40e45 40e45
Quarry E Quarry F Quarry G
2.8 e4.3 0.7 e1.4 1.2e3.1
30e50 25e35 25e35
Areas of quarries found in this study
Average magnetic susceptibilities in Areas A to D (10 3 SI unit)
Thickness in centimeters of sandstone blocks (step height) cut from Areas A to D
Area A
1.2e2.5
50e60
Area B Area C Area D
2.6e4.5 0.3e1.4 1.6e2.1
45e60 30e50 30e40
Preah Ko Bakheng Transitional Bakheng to Baphuon Angkor Wat Early Bayon Late to post Bayon a b
From Uchida et al. (2007). from Uchida and Ando (2001).
Area B is in the northeastern part of the linear structure. In Area B, the average magnetic susceptibilities are characteristically high, 2.6e4.5 103 SI units (except for quarries Nos. 2 and 7, 1.3 and 2.1 103 SI units, respectively), and the thicknesses (step heights) are 45e60 cm. Judging from these facts, it is considered that the quarries in Area B were used in the Angkor Wat to early Bayon periods and correspond to quarry E of U chida et al. (2007) (Table 2). Area C is located in the north of the east part of the linear structure. The magnetic susceptibilities in this area are as low as
0.3e1.4 103 SI units except for a quarry No. 14 (1.98 103 SI units) and thethicknesses (step heights) arealso relatively low, 30e 50 cm. These characteristics suggest that the quarries in this area were used in the main Bayon period and correspond to quarry F of Uchida et al. (2007) (Table 2). The quarries in Area D (Don Enn) are in the northwestern part of the linear structure. The average magnetic susceptibilities range 1.6e1.2 103 SI units and the block thicknesses (step heights) are 30e40 cm. The average magnetic susceptibilities in Area D are similar to those in Area A. However, the thickness (step height) of the sandstone blocks is much thinner in Area D compared with Area A. On the basis of these facts, it is considered that the quarries in Area D were used in the late Bayon period and correspond to quarry G of Uchida et al. (2007) (Table 2). These results suggest that the sandstone quarrying activity moved over time from the eastern area to the northwestern area of the liner structure. It seems that the quarries were concentrated in certain areas during certain periods, and that sandstone blocks were supplied from multiple quarries in a certain area. The depth of the quarries varies from place to place, and ranges from less than 1 m up to 6 m. A depth of 1 e3 m is most typical. The largest quarry found to date is quarry No. 45, which reaches 50 m in width and 6 m in depth. The sandstones of the quarries in Area B have high average magnetic susceptibilities corresponding to the sandstone blocks used in the monuments of the Angkor Wat to early Bayon periods. However, their scale seems not to be suf �cient to supply all the sandstone blocks used in these periods when the representative large scale monuments of Angkor Wat, Ta Prohm, and Preah Khan were constructed. Thus it is expected that many other large scale quarries remain undiscovered around Area B. The total amount of sandstone blocks cut from the quarries in Areas A to D is not estimated yet. This kind of surveying is thus indispensable to estimate the total volume of sandstone blocks supplied from these quarries. 3. Transportation route of sandstone blocks
Fig. 3. Histograms of magnetic susceptibilities of the sandstone in Areas A to D. MS: magnetic susceptibility, and SD: standard deviation.
Previously it was believed that sandstone blocks of the Angkor monuments were transported from the southeastern foot of Mt. Kulen to Tonle Sap Lake via canal, and then supplied to the Angkor monuments by going up the Siem Reap River via Tonle Sap Lake (Lunet de Lajonqière, 1911: Boisselier, 1952). In this scenario, the transportation distance reaches 90 km, and also it is obliged to go up the Siem Reap River. This is not an easy transportation route. On the basis of satellite images, we found a waterway (consisting of canals and partly of rivers) connecting the southeastern foot of Mt. Kulen to the Angkor area, and also an east ewest linear structure 2.4 km long at the southeastern foot of Mt. Kulen. The existence of
1162
E. Uchida, I. Shimoda / Journal of Archaeological Science 40 (2013) 1158 e1164
the waterway was suggested by Evans et al. (2007) and Hendrickson (2007). We conducted a � eld investigation to con�rm whether these structures were used for the transportation of sandstone blocks. 3.1. Sandstone block transportation route at the southeastern foot of Mt. Kulen
are distributed around the bank in the eastern and northern areas. Several small sandstone quarries exist adjacent to the bank. Thus, it is highly possible that this bank was used as a transportation route for sandstone blocks. As the relatively old quarries (Area A) are located near the east end of this bank structure, it is supposed that this bank was constructed in the early Angkor period. 3.2. Waterway connecting Mt. Kulen to the Angkor monuments
An eastewest linear structure was observed on satellite images of the southeastern foot of Mt. Kulen as shown in Fig. 1. As it was hypothesized that this linear structure was a transportation route for sandstone blocks, we conducted a � eld investigation to con�rm this. The east end of this linear structure is located at N 13 280 5800 and E 104 120 1500 withan altitude of 93 m, and the west end is at N 13 290 0500 and E 104 100 5900 with an altitude of 90 m. The length of the linear structure is 2.4 km. The central part has a slightly lower altitude and the lowest point is 82 m in altitude. The �eld investigation revealed that the linear structure forms a bank which is about 2.5 m higher than the surrounding area. Because an 800 m long depression was observed adjacent to the bank to the north in the eastern part and some water remains in the depression, it is possible to consider that it was a waterway. However, the depression is not observed in the western part of the bank structure and the central part of the bank has not yet been investigated due to dense forest. As mentioned previously, many sandstone quarries
On satellite images, we found a line starting from Point 1 (N 13 28 2200 , E 104 100 4700, and 81 m in altitude) 1.3 km south of the west end of the aforementioned bank and continuing to the Angkor monuments (Fig. 4). Because this line was hypothesized to be a canal used for transportation of sandstone blocks from the southeastern foot of Mt. Kulen to the Angkor monuments, we carried out a � eld investigation. The route is shown in Fig. 4 with a red line. Judging from the satellite images, it seems to consist mainly of canals and partly of rivers. The result of the � eld investigation will be described in what follows. Photos are also shown in Fig. 5. Point 1 to Point 2 : The 2 km area between Point 1 and Point 2 (N 13 270 5100, E 104 090 3900 ) forms an embankment 2 m high. A depression 20 m wide is adjacent to this embankment to the north. It was assumed that this depression was used as a waterway, but no water remains in this depression. The altitude of Point 1 is 0
Fig. 4. The transportation route of sandstone blocks, connecting the quarry sites at the southeastern foot of Mt. Kulen to the Angkor monuments, shown with a red line. (For interpretation of the references to colour in this � gure legend, the reader is referred to the web version of this article.)
E. Uchida, I. Shimoda / Journal of Archaeological Science 40 (2013) 1158 e1164
1163
Fig. 5. Photos of the transportation route of sandstone blocks. The location of each point is shown in Fig. 4. (a) The starting point (Point 1) of the transportation route to the Angkor monuments. An embankment about 2 m high is accompanied by a depression to the north. (b) The turning point (Point 2) of the embankment into a canal. (c) Appearance of the canal at Point 3. The height from the bottom of the canal to the top of the embankments reaches 6 m. (d) Appearance of the canal at Point 7. The cross section of the embankment is exposed by a road cut. (e) Appearance of the canal at Point 11. The canal is about 20 m wide with an embankment 2 m high to the south, and (f) appearance of the canal at Point 12.
81m and thatof Point 2 is 78m. The central partis lowerthan Point 1 and Point 2, and the lowest point is 72 m in altitude. As water cannot �ow from Point 1 to Point 2, the sandstone blocks may have been transported by a road on the embankment. No arti�cial structure related to stone transportation was observed between the west end of the bank at the southeastern foot of Mt. Kulen and Point 1. Point 2: The depression continuing from Point 1 deepens suddenly at Point 2, and changes into a canal, in which water remains. It is postulated that the sandstone blocks were transported by a road on the embankment to Point 2 and then by a canal from Point 2. Several sandstone blocks are scatted at Point 2. Point 2 to Point 3 : The area between Point 2 and Point 3 (N 13 270 5300 , E 104 080 4000 ) is a canal with embankments on both sides, where water remains in most places. The embankments have a height of 2 e3 m, and the canal has a depth of 5 e7 m from the top of the embankments to the bottom of the canal and a width of 20 e 30 m. Because the altitude at Point 3 (81 m) is slightly higher than at Point2 (78 m), the canal around Point 3 was dug deeper than the surrounding areas. Point 4: Instead of a canal, several rivers run in the area around Point 4 (N 13 270 3500 , E 104 060 4800 ). Point 5 toPoint 6 : There is a canal between Point 5 (N13 270 1600 , E 104 050 3700 )andPoint6(N13 270 1800 ,E104 050 2200 ), similar to that in the area between Point 2 and Point 3. The height of the
embankments is 1.5 m and the width is around 20 m. The canal is destroyed in the west part of this area and is used as a � sh farm at present. There is a river instead of a canal in the west end of this �sh farm. Point 7 : A canal was con �rmed at Point 7 (N 13 270 4000 , E 104 0 04 4900 ). A cross section of the embankments can be observed at this point because of the construction of a road through the embankment. The height of the embankments is about 1.5 m. Point 8: At Point 8 (N 13 270 0700, E 104 020 2300 ), an embankment with a height of about 2 m is observed on the south side, but no embankment on the north side. We can see a depression with width of around 20 m at this point, but no water was observed there. On satellite images, a river can be seen south of this point, and water runs in this river at present. Point 9: At Point 9 (N 13 250 5500 , E 104 00 0 1400 ), there is a meandering river instead of a canal. Point 10: The river joins the Roluos River at Point 10 (N 13 250 2800 , E 103 590 2900 ). The Roluos River runs southward and reaches Tonle Sap Lake through the Roluos monuments. Chau Srei Vibol, which is 3 kmeast of Point 10, is connected to Point 10 with a canal. A long canal continues westward from Point 10 to the Angkor monuments. Point 10 to Point 14 : There is a canal between Point 10 and Point 14 (N 13 250 4100, E 103 530 2600 ). The canal shows an east ewest orientation in the area between Point 10 and Point 13, then turns
1164
E. Uchida, I. Shimoda / Journal of Archaeological Science 40 (2013) 1158 e1164
to the northwest, and continues to the west of Banteay Kdei (Point 14). The 2 m high south embankment is used as a road at present, but the north embankment is not clearly observed. The width of the canal is 20e25 m. The canal is �lled with soil in many places, although water remains in several places up to 2 m deep.
(E. Uchida)). The manuscript was improved by useful suggestions by two anonymous reviewers. References Boisselier, J., 1952. Ben M ala et la chronologie des monuments du style d Angkor Vat. In: Bulletin de l École Française d Extrême-Orient, vol. 46, pp. 187e238. Boulbet, J., 1979. Le Phnom Kulen et sa region. In: Collection de texts et documents sur l Indochine, IXI. École Française D Extrême-Orient. Carò, F., Im, S., 2012. Khmer sandstone quarries of Kulen Mountain and Koh Ker: a petrographic and geochemical study. J. Archaeol. Sci. 39, 1445 e1454. Delvert, J., 1963. Recherches sur l érosion des grès des monuments d Angkor. In: Bulletin de l École Française d Extrême-Orient, vol. 51, pp. 453 e534. Evans, D., Pottier, C., Fletcher, R., Hensley, S., Tapley, I., Milne, A., Barbetti, M., 2007. A comprehensive archaeological map of the world s largest preindustrial settlement complex at Angkor, Cambodia. PNAS 104, 14277e14282. Garnier, F., 1873. Voyage d exploration en Indochine, effectué pendant les année 1866 et 1868, Paris. Hendrickson, M., 2007. Arteries of Empire. An Operational Study of Transport and Communication in Angkorian Southeast Asia (9th to 15th centuries CE). Ph.D thesis, University of Sydney. Kucera, J., Novák, J.K., Kranda, K., Poncar, J., Krausová, I., Soukal, L., Cunin, O., Lang, M., 2008. INAA and petrological study of sandstones from the Angkor monuments. J. Radioanal. Nucl. Chem. 278, 229 e306. Lunet de Lajonqière, E., 1911. Inventaire descriptif des monuments du Cambodge, vol. 3. Publication de l EFEO 9, Paris. Meesook, A., Suteethorn, V., Chaodumrong, P., Teerarungsigul, N., Sardsud, A., Woongprayoon, 2002. Mesozoic rocks of Thailand: a summary. In: Proceedings of the symposium on Geology of Thailand, pp. 82 e94. Rocks, D., 2009. Ancient Khmer quarrying of arkose sandstone for monumental architecture and sculpture. In: Kurrer, K.E. (Ed.), Proceedings of the Third International Congress on Construction History. Brandenburg University of Technology Cottbus, Germany, Neunplus 1. 1253 e1242. Saurin, E., 1954. Quelques remarques sur le grès d Angkor. In: Bulletin de l École Française d Extrême-Orient, vol. 46, pp. 619 e634. Uchida, E., Ando, D., 2001. Petrological survey 2000. In: Japanese Government Team for Safeguarding Angkor (Ed.), Annual Report on the Technical Survey of Angkor Monument 2001, pp. 225 e247. Uchida, E., Cunin, O., Shimoda, I., Suda, C., Nakagawa, T., 2003. The construction process of the Angkor monuments elucidated by the magnetic susceptibility of sandstone. Archaeometry 45, 221 e232. Uchida, E., Cunin, O., Suda, C., Ueno, A., Nakagawa, T., 2007. Consideration on the construction process and the sandstone quarries during Angkor period based on the magnetic susceptibility. J. Archaeol. Sci. 34, 924 e935. Uchida, E., Ogawa, Y., Nakagawa, T., 1998. The stone materials of the Angkor monuments, Cambodia. The magnetic susceptibility and the orientation of the bedding plane of the sandstone. J. Mineral. Petrol. Econ. Geol. 93, 411 e426. Uchida, E., Ito, K., Shimizu, N., 2010. Provenance of the sandstone used in the Khmer monuments in Thailand. Archaeometry 52, 550 e574. Uchida, E., Suda, C., Ueno, A., Shimoda, I., Nakagawa, T., 2005. Estimation of the construction period of Prasat Suor Prat in the Angkor monuments, Cambodia, based on the characteristics of its stone materials and the radioactive carbon age of charcoal fragments. J. Archaeol. Sci. 32, 1339 e1345.
4. Considerations and summary
’
’
’
’
On satellite images, we found a linear structure 2.4 km long at the southeastern foot of Mt. Kulen. As the result of a � eld investigation,it was proved to be a bank. Abundant sandstone quarries are distributed around the bank. On the basis of the measurements of magnetic susceptibilities and thicknesses of sandstone blocks (step heights), the quarries can be classi�ed into four areas, and it was revealed that the quarrying activity moved from the eastern area to the northwestern area of the bank over time. These facts suggest that the bank was used as a transportation route for the sandstone blocks from the beginning of the Angkor period. Judging from its location, it is considered that the canal (partly a river) starting from a point 1.3 km south of the west end of the aforementioned bank (Point 1) and continuing to the Angkor monuments was used for the transportation of sandstone blocks quarried from the southeastern foot of Mt. Kulen. This canal is almost the shortest route from quarry sites to the Angkor monuments, and the distance is 34 km. Sandstone blocks can be easily transported by this canal from quarry sites to Phnom Bok and Chau Srei Vibol situated on the way to the Angkor monuments; to the Roluos monuments by going down the Roluos River; and also to Phnom Krom by going down the Siem Reap River. Because it is not necessary to go upstream in the canal and river, it seems that sandstone blocks were transported ef �ciently by this route. From such a point of view, the transportation route suggested previously from the southeastern foot of Mt. Kulen, via a canal to Tonle Sap Lake, and going up the Siem Reap River to the Angkor monuments seems less probable than that proposed in this paper. Acknowledgments
This study was supported �nancially in part by a Grant-in-Aid for Scienti�c Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (Project No. 23401001
’
’
’
’
’
’
’
’
’
’
’
