KANJI – Tsuglag-khang
Architectural measures – Kanji Tsuglag-khang
by John Harrison (JH)
First Phase
Materials on site: mud bricks, yamang/slate copings, and timber. Nearly enough keru/pine in 4x4in/10x10cm for skechsing/roof ladder, but deodar ceder too small section and poor quality. No stone and no sand yet collected. No talboo/willow sticks for parapet.
Meetings with Mistry Sonam Wangchuk and Sonam Dorje to discuss programme, materials, and check stone and sand sites. Stone was available 1 kilometre south of the village, yamang was ½ kilometre across the river, markalak 1 hour north towards the road. Sonam Wangchuk did not initially like the skechsing proposal and favoured a steel cable tied round the outside of the building. Labour rates were agreed: rs.300 per day for the mistries, 170 for labourers, 150 for donkeys.
We broke stone and carried to site by donkey for 3 days. Sand excavation from the edge of the village started on 13 May and continued in parallel. Labour force varied throughout the construction project, averaging 5 labourers a day, the two mistries, plus Konchok Tinlas and myself, and 5 donkeys. Deodar roof beams, with Yarpa/poplar as second-best alternative was looked for in Leh.
Foundation
Work started on site with poplar pole raking shores to the NE corner. We then dug down to below the existing stone foundations and inserted new stone to underpin the existing, section by section, working along the rear N side and down the W side.
The new stone plinth was tied into the open joints and cavities of the old foundation (which generally extended only just below ground level), and brought up to the top of the old stonework, which is a little higher than internal floor level.
I reduced the width of the top of the new plinth from the proposed 50cm to about 30cm, but with a battered face to the plinth. (Local objections to children playing on it if it was wider). The top is finished with yamang sloping outwards, and bedded into the first brick course of the wall.
Removal of the added stone facing on the W elevation (built 30-40 years ago by Konchok Tinlas’s father before KT was born) revealed an extremely eroded area of foundation stone and mud brick - an entire brick depth in places, 8in/20cm. The marked belly (bulge) in the brick wall at the NW corner was in fact mostly pack animal abrasion of the lower brick courses over a period of 7 centuries.
Here we cut out badly-eroded bricks and inserted new bricks to return the wall to its original line.
Porch
The new bricks made in the village are 5in/12.5cm deep x 8in/20cm x 18in/46, but we found when taking down the parapet over the SE porch wall that the old bricks are mostly 4 to 4 1/2in deep, and that there were two sizes of brick, one 8in wide and one 11in, laid alternately. We were able to salvage about 50% of the bricks in this section of parapet intact. We used some of the salvaged bricks to reface the W wall, as in some places the new bricks would not fit. The SW corner porch wall was shored up and the new stone plinth and underpinning carried around the S face to the inside face of the porch wall.
At the SE corner the leaning wall was shored in two directions, and after removing the overturning load of the parapet above, a much deeper excavation was made to reach the base of the new buttress. The 40-year stone facing was removed to reveal the original river-stone foundations well above ground level, and much eroded. (I am surprised the building didn’t fall over 500 years ago). Again, a new stone plinth with battered faces to E and S, tied into the original foundations, was built up from an excavation 5ft6in/1.68m deep, and topped with sloping yamang/slate.
In the porch I found evidence of painting on the E wall, but I did not uncover enough to make out the design. Certainly damaged or missing in the centre, but I hope much surviving to each side. Consequently we did not demolish and rebuild the wall as originally planned.
The porch main beam was propped adjacent to the recent stone pier built to support the end of the beam which had inadequate seating left as the wall moved outwards. The stone pier was taken down, to reveal the original brickwork much eroded by rainwater through the gap in the roof above. Damaged brickwork on this corner was cut out and new bricks inserted, working from the bottom up. SW made a timber bracket from one of the redundant deodar sections, carved in a provincial style appropriate to the original capitals, and this was built into the new brickwork to support the end of the main beam.
The retaining wall around the entrance platform was completely rebuilt, to the same battered face as the plinth around the temple.
At this point we stopped work: the mistries had another 3-week job in Khaltse to complete; Konchok Tinlas had to return to Leh and we had to find roof beams. KR bought 4 massive poplar trunks in his home village, now transported to Khaltse, but these were not useable yet as unseasoned - only felled 2 months before. Deodar was not to be had anywhere, in Leh or Kargil.
Second Phase
There was a slow start to the second phase of the repair works, with the mistries returning several days late, difficulty in obtaining labour and donkeys for transport, a visit by government officials, and poor weather - heavy rain showers, and snow. Talboo (willow sticks) was more difficult to obtain than expected, but Konchok Tinlas eventually located sufficient quantities in Mulbek and Shergola.
Removal of Roof
Work began with the demolition of the parapet walls down to the lower talboo level. The brick bonding was recorded, and intact bricks and talboo set aside for reuse. The brickwork of the lower porch parapet walls was not connected at all to the main parapet, suggesting that the porch, or at least the porch parapet, could be a later addition. The brick bonding of the southeast and southwest ends of the main parapet suggested that the parapet might have extended across the roof above the entrance wall, but the evidence was not conclusive.
The upper layers of the earth roof covering were removed over the entire roof area, as much as 12in./30cm at the northern parapet. In the centre of the roof, directly above the pillar, a large flat stone 12 x 16 x 4in/ 30 x 40 x 10cm rested on the beams, and had caused one of the ceiling boards to collapse. As its most likely use in this position would be as a pillar base, this supports anecdotal evidence of a former structure on the roof.
Then all the earth over the southern bay was removed (a further 6in./15cm) to expose the ceiling boards. Sheets of multi-layered paper-like bark from the Stakpa tree (birch), which still grows in the high mountain areas, were found laid over the boards, together with shukpa chippings. The boards in the southern bay were not shukpa, according to the mistry, and were severely damaged by wet rot, even though in most cases this was not apparent from below. Only ten boards out of fifty were reusable. The ceiling boards were removed, and then the six beams carefully taken out. All the beam ends were severely decayed, and in two cases had completely disintegrated.
Main Beam
The main beam, which was deflected more than 4in./10cm at its centre, was raised 2-3cm with a 45 ton screw jack, propped, and the central pillar and capital removed. The base of the pillar, 10cm below floor level, was not affected by rot, and the padstone and ground appeared quite firm. The top of the pillar and capital were trimmed to correct the movement from the vertical which had taken place, and then replaced and realigned, with an additional stone at the base. The eastern beam bracket could be pushed back into the wall, returning it to the horizontal position and providing a greater bearing area for the end of the beam. I believe that no further deflection of the main beam will take place, without the load of the upper roof structure removed 40 years ago, and now with a much reduced thickness of earth above and with the eastern wall stabilized.
The six new yarpa beams, 6 x 4in/15 x 10cm in section, were carefully fitted into enlarged sockets in the top of the entrance wall without affecting any areas of surviving wall paintings.
Ladder Framework
The skechsing ladder framework to tie the wall-tops together was fabricated on the ground and then assembled at roof level. Each member had to be made from two pieces glued and screwed together, because 17ft/5m lengths were not available in the market, and could not have been carried to Kanji by donkey (There were no yaks available at the time). All joints between members were pinned with hardwood pegs of Kikar from Kashmir. Two additional east-west ties were introduced to the skechsing, one above the main beam, and one at half-span of the new secondary beams. These were pegged through to the beams below.
The ceiling was boarded with the few sound original boards, assorted shukpa boards collected by the Mistry, sound sections of the original beams, the new yarpa boards brought from Leh, and a door from Konchok Tinlas’s house. The boards were then covered with the salvaged stakpa bark and shukpa chips, and a layer of bosi grass from the mountain (yagzes grass would not have grown until August). 2in/5cm of earth was then laid, level with the top of the skechsing, and then a 1in/3cm layer of markalak paste. The markalak is cut out of a cliff face one hour´s walk north of the village and carried to the building site by donkey. The solid lumps of markalak clay have to be soaked in water for two days until they dissolve into a sticky paste.
Parapet
The rebuilding of the parapet walls commenced with the laying of the lower course of talboo, using both new and old sticks, to support the projecting brick string course. The original bonding pattern of the brickwork was followed as closely as possible, but was limited by the different sizes of new and old bricks. Only complete courses of new or old bricks could be laid because of the different heights (4in/10cm and 5in/13cm); there were no 11in/28cm wide new bricks, while the majority of old bricks were 11in. wide, with few 8in/20cm wide old bricks.
A slate/yamang drip was introduced to protect the top of the lower string course from rain.
The upper string course at the top of the parapet was protected with yamang as originally, finished with a rounded coping of sand and river-sand mortar.
The overall height of the building has been increased by the introduction of the skechsing 4in/10cm, the additional height of five courses of new bricks, and the new yamang course above the lower string course.
Roof
Following the completion of the parapet walls, earth was laid to falls over the markalak layer. To keep the thickness of earth to a minimum by shortening the drainage runs, the main roof was drained from the centre of the east side, between the two buttresses, and the porch roof also from the east. This follows the drainage pattern of the Graz proposal, but without the introduction of the parapet wall across the roof.
The two water spouts are traditional open-topped wooden channels carved from 6in./15cm diameter yamang, 6ft./1,85m long (not plastic or G.I. pipe).
A second layer of 1in./3cm Markalak was laid over the earth of the main roof, finished to the water spouts and the parapet walls, and topped with a protective layer of dry sand.
Walls
On the external walls, layers of loose paint were peeled off to a sound base (sometimes the original plaster of the wall), and the white lime wash carefully salvaged for reuse. The nearest source of the mineral is 5-6 hours away over the mountain to the northeast of the village.
No further traces of a red base coat were found on any of the walls, so perhaps the red I had earlier found in a small area of the west wall was rain run-off from the red parapet above.
Mud plastering was carried out to areas of new or exposed brickwork: on the porch wall exteriors, the base of the west wall, and the parapets.
Third Phase
Shrinkage cracking of the mud plaster on the external walls necessitated pointing of the cracks with markalak, and so some new markalak had to be dug and brought to site by donkey. One area on the west porch wall was replastered with a mix of gritty black river sand, which is less susceptible to shrinkage.
Paint
The tsak clay for the red parapet paint was excavated from a cave 2 ½ hours (and two river fordings) south of Kanji on the road to Rangdum Monastery, Zanskar. The orange clay is then burnt to turn it to a red colour, before being dissolved in water. The salvaged flakes of kartse whitewash were dissolved in water, and applied to the walls, from the new stone plinth up to the lower talboo course, with a rag on a long stick.
The painting of the red parapet, and final clearance of debris from around the site, was to take place in the next two days. Due to car breakdown delays and the additional plasterwork, yamang slate for paving the entrance platform and west side of the temple was not collected, and this work will take place next year.
An aluminium extension ladder should be purchased to facilitate roof maintenance and snow clearing.