Exploring Ancient Pashupati: Preliminary Results

of Archaeological Surveys and Excavations at

Bhandarkhal, Kathmandu Valley UNESCO World
Heritage Property (Nepal)
- Coningham, R.A.E., Acharya, K.P., Davis, C.E., Manuel, M.J.,
Kunwar, R.B., Hale, D., Tremblay, J., Gautam, D. and Mirnig, N.

1. Introduction 2. Archaeological Investigations

Cultural heritage is threatened by within the Kathmandu Valley and
increasing pilgrim numbers at major religious at Pashupati
sites around the world, particularly in Asia. Few excavations have been targeted at
Pashupati, a UNESCO World Heritage Site identifying earlier occupation at archaeological
in the Kathmandu Valley (Nepal), offers a sites within the Kathmandu Valley despite the very
microcosm of the impacts facing archaeology real threat to their survival from rapid urbanisation
from these global developments, providing and associated pressures. Archaeological
a case-study of heritage management for excavations have been conducted at Harigaon
sustainable pilgrimage. Located within the (Veradi 1992) and at Dumakhal (Khanal and
historic city of Deopatan (Michaels and Tandon Riccardi 2007) but both these excavations were
1994: 180), Pashupati comprises a monumental generally limited by their focus on monumental
core of temples, ashrams and funerary ghats brick architecture. Furthermore, excavations of
covering 83 hectares along the Bagmati River Patukadwon in Patan, traditionally thought to be
(Figure 1). Receiving approximately six million the site of the Kirati Palace (TPU 2003a, TPU 2003b),
pilgrims annually, Pashupati’s status as a major are yet to be fully published. More recently, the
Shaivite site in South Asia and Nepal’s premier 2015 Gorkha Earthquake has led to post-disaster
cremation destination has led to tensions and rescue archaeological assessments across
between development, religious practice the major Durbar Squares of Patan (Coningham
and the protection of cultural heritage. In et al 2016b), Hanuman Dhoka (Coningham et al.
comparison with significant conservation of its 2016c) and Bhaktapur (Coningham et al. 2016d)
eighteenth century temples, little has been done in addition to pilot excavations at collapsed
to evaluate Pashupati’s subsurface archaeology monuments. However, these investigations are
and, as pilgrim numbers rise, pressure to in their infancy and the general focus of previous
invest in the infrastructure that accompanies studies at the major monumental complexes has
development could pose a risk to heritage been towards standing architectural remains (e.g.
protection. This is of great concern as little is Basukala et al. 2014, Gutschow 1997 and 2011
known of the potential and valuable subsurface Hutt 1994, Korn 2007 and 2014) rather than sub-
heritage of this site of Outstanding Universal surface heritage and this situation is mirrored at
Value. Pashupati.

94
 Although Pashupati’s later medieval offer flowers to the deities of Pashupati. A number
architecture is well documented (Michaels and of scholars attribute the formal laying out of
Tandon 1994), little is known of the presence Bhandarkhal to the rule of King Pratapamalla (r.
or nature of earlier cultural phases, with a 1641-1674 CE) and record that it was later enclosed
paucity of archaeological evidence for the site’s within a wall by King Rana Bahadur Shah (r.
beginnings and early religious activities (Mirnig 1777-1799 CE) (Michaels 2008: 6)(Hutt :1995: 180).
2013: 326). Previous excavations elsewhere in the According to Slusser, Rana Bahadur Shah donated
Kathmandu Valley have identified the possibility the entire garden to the temple at Pashupati in the
of the existence of subsurface features, such as at nineteenth century (1982: 230). However, the rich
Harigaon (Veradi 1992), although a systematic corpus of Licchavi inscriptions across Deopatan
scientific dating programme of archaeological suggests that the western hillside above the temple
sequences has never been fully realised for this complex was already densely occupied by shrines
region. Furthermore, recent excavations in the and residences by the fifth century CE and some
Natal landscape of the Buddha at Lumbini and chronicles link the area to a visit by the Mauryan
Tilaurakot have uncovered pre-brick structures Emperor, Asoka (Regmi 1965: 8). Amatya has
below standing brick remains (Coningham et also drawn attention to the Licchavi and Malla
al. 2013a, Coningham et al. 2016f, Davis et al. tradition of the establishment of gardens and tanks
2016, Strickland et al. 2016), suggesting that in order to present fruit and flowers to presiding
non-durable buildings were succeeded by more deities (2011: 10) and, as such, it is to be expected
monumental architecture. Such a scenario is likely that there would be earlier activity in the vicinity
in the Kathmandu Valley as chronicles narrate that of Bhandarkhal before Pratapamalla’s eighteenth
Pashupati was first constructed in the third century century intervention.
BCE and the presence of fourth to ninth centuries Excavations were first conducted at
CE Licchavi inscriptions (Amatya 2011, Joshi 1974, Pashupati in 1990, and resumed in 1993, by
Regmi 1983, Michaels 2008, Mirnig 2013, Tiwari Stefano Pracchia in Bhandarkhal (Di Castro 1997:
2001) suggest a deep sequence, significantly earlier 134). During these investigations, five trenches
than the present eighteenth century standing were excavated across Bhandarkhal but, as stated
remains. Such a sequence is of great importance in above, the full findings of these investigations
understanding the development of the site before, were never published. However, preliminary
and after, the emergence of Pashupati as the results suggested that occupation in the area
tutelary deity of Nepali kings and the introduction began as early as the first century CE on the
of concepts of ‘Indic’ divine kingship from the basis of ceramics comparable with those found
sixth century CE onwards (Mirnig 2013). at Harigaon (ibid: 134). One major focus of these
On account of planned improvements investigations was a depression located roughly in
to the wooded park known as Bhandarkhal, the the centre of Bhandarkhal, which was interpreted
Pashupati Area Development Trust (PADT) and as a monumental tank and attributed to the
Department of Archaeology, Government of early Licchavi period on the basis of artefactual
Nepal, invited the team to focus their primary evidence. Across Bhandarkhal, Di Castro stated
activities in that area. Bhandarkhal, meaning ‘the that three cultural horizons were identifiable.
Treasury Garden’ is a walled compound west of Based on his artefactual assemblage, he attributed
the main Pashupati Temple complex, south of the first to the first and third centuries CE, the
the Jayabageshwari Temple and west of Gausala second to between the fourth and ninth centuries
(Figure 2). Close to watershed dividing the Bagmati CE and, finally, sporadic traces of occupation
from its tributary, the Dhobi Khola, it forms rough relating to the early Malla period, which caused
rectangle measuring 300 metres along its northern the destruction of the two earlier structural phases
edge, 250 on its southern, 260 on its eastern and (ibid: 134). However, the paucity publication and
330 metres on its western. One of the parks and scientific dating leaves a lacuna of information
gardens that break up the urban environment of regarding the origins and development of this
Deopatan, these spaces are thought to have been part of Pashupati.
endowed in order for devotees and pilgrims to
95
3. Surveying Bhandarkhal as wire fences and corrugated metal shelters for
Basic topographical mapping of cattle along the eastern edge of the area, adjacent
Bhandarkhal was undertaken, allowing us to the brick wall and coach park. A chain of intense
to tie together the different elements of the dipolar magnetic anomalies which was detected
project – geophysics, augering and excavation along the southern edge of the survey area could
– into a single geographic database. This was reflect the presence of another service pipe (Figure
followed in January 2014 and January 2015 by 3 and 4).
an archaeological geomagnetic survey using a We also sampled 92 auger cores in the
Bartington Instruments Grad601-2 dual fluxgate Pashupati area in July 2014, with a number within
gradiometer. The instrument sensitivity was Bhadarkhal (Figure 5). The success of auger
nominally 0.03nT, the sample interval was 0.25 coring in South Asia has been demonstrated in
metre and the traverse interval was one metre, Anuradhapura, Sri Lanka (Coningham 1999: 49;
thus providing 1,600 sample measurements per Coningham and Gunawardhana 2013: 23), Lumbini
20 metre grid unit (3,600 sample measurements and Tilaurakot in Nepal (Coningham and Acharya
per 30 metre grid unit). Data were downloaded on 2013; Coningham, Acharya and Manuel 2014) and
site into a laptop computer for initial processing Bangladesh (Coningham et al. 2013b), where its
and storage and subsequently transferred to a portability and ease of use allowed for rapid site
desktop computer for processing, interpretation assessment. Auger coring allows us to identify
and archiving. Geoplot v.3 software was used the depth and extent of cultural material as well
to process the geophysical data and to produce as providing the ability to record general macro-
continuous tone greyscale images of the raw stratigraphic details of the broader site. In all
(minimally processed) data. The area surveyed instances, the auger cores were undertaken using
measured a maximum of 240 metres east-west and an Eijkelkamp soil auger and were continued until
140 metres north-south with a smaller area of 30 bedrock or natural soil was reached. The cores were
by 40 metres in the north-east corner of the area. removed and analysed in 20 centimetre sections,
The main features identified by the survey were a with soil colour and consistency recorded along
number of linear positive and negative magnetic with mineral and cultural inclusions. Soil colour
anomalies, generally aligned north-west/south- was judged using Munsell soil colour charts. The
east. Typically weak and broad, it was thought that location of each auger was plotted using the total
they could reflect soil-filled features such as ditches station. The Bhandarkhal augers were sampled
or trenches where wall footings have been robbed on two east-west profiles (lines A and B) and two
out and former paths or trackways. For example, north-south profiles (lines C and D). Each auger
the existing track through the park corresponds to profile consisted of ten augers located 20 metres
a very weak, linear positive magnetic anomaly in apart from each other. Augers D1 and D2 were
the north-west of the survey area. Concentrations offset by about 40 metres to the east due to severe
of small, strong dipolar magnetic anomalies were waterlogging of the north-west corner of the site.
detected throughout the area but particularly in Natural soil, unaltered by human activity, was
the north of the survey near the existing track. reached in nearly all augers and consists of a
These almost certainly reflect ferrous and fired yellowish-brown silty sand, containing pieces
debris; deposits of fired bricks were noted on the of degraded bedrock, mica and quartz. To the
ground during survey. A scatter of further small, west, natural soil was quite shallow at 1.0 to
discrete dipolar magnetic anomalies has been 1.5 metres deep and as shallow as 0.7 metres in
detected across the area. These reflect small items auger B2. In most areas, cultural material was
of ferrous and/or fired litter, probably associated evident throughout the auger-cores and was
with the current use of the park for coach parties, predominantly brick with some pottery. Augers
cooking and recreation. The detection of sub- C8 and C9 contain evidence of a river channel, due
surface features in the north-east was hindered to the presence of a soft, sandy material containing
by the presence of a large metal water tank in the small stones and gravel, underlying a short phase
centre of the area and its associated pipe, as well of natural. This matches the geophysical survey,
which identified potential channels in this area.
96
The deepest auger was C5, taken in the centre of contained three fills (Figure 6). The primary fill
the tank in Bhandarkhal. The top one metre was was a very soft sandy material with some ceramics
a brownish grey silty sand; from one metre to two and brick fragments. This fill may have been an
metres depth was a coarse grey sand; from two early wash into the pit cut sealed by the secondary
metre to just over three metres was a greyish brown later infill. This secondary fill was silty and soft,
fine sand with evidence of iron panning, indicating also containing brickbats and ceramics. Finally, the
stagnating water. Finally, from 3.2 metres to five tertiary fill was a soft sandy silt, again containing
metres down, was a compact black cotton soil – ceramics and brickbats. Capping the tertiary fill
which may have been deliberately used to line the was an irregular brick topped surface. This discrete
base of the tank as the soil is very good at retaining cluster of broken bricks, some placed on edge,
water, rather than allowing it to drain away. appeared to form a small area of paving (Figure 7).
This is possibly a deliberate attempt at creating a
4. Excavations in Bhandarkhal temporary platform forming a more solid footing
Due to the potential of encountering for individuals utilising the tank edge for either
structures and archaeological stratigraphies ritual or utilitarian purposes. The bricks sat within
dating to the Licchavi period in Bhandarkhal, we a silty clay bonding material that contained smaller
decided to locate two trenches at the monumental fragments of brick bat. Whilst the purpose of the
tank during our first season of excavations in pit is unclear, during our investigations we noted
2014. Trench 1 was located on the eastern edge of several depressions in the base of the cleared tank.
the tank and Trench 2 on its northern edge. Both Perhaps dug out for material to be used elsewhere,
trenches were excavated in order to define the over time it is assumed that these pits filled with
tank edge and the developmental sequence of the sediment from washes during periods of rains,
tank. In 2016, Trench 3 was located in the south such as monsoon.
of Bhandarkhal in order to investigate features Whilst we did not identify formal brick
identified in geophysical survey of the area and facings for the tank as is seen in the later medieval
also to gain an occupational sequence for the site squares of Patan, Bhaktapur and Hanuman
to link to the evidence from the tank. Dhoka, we did identify a firm, fairly sterile
material that formed a stepped edge that rose to
4.1 Trench 1 and Trench 2 the east (Figure 8). Whilst we await the results of
Trench 1 was located close to the centre geoarchaeological analysis, we suggest that ponds
point of the eastern edge of the tank depression at and tanks with earthen edges may have preceded
Bhandarkhal. The trench originally measured eight the brick-lined monuments that are synonymous
metres east to west and one metre wide. This was with the Durbar Squares of the Kathmandu Valley.
later extended in the eastern four metres to a two It is postulated that earthen cut tanks may have
metre wide slot, with a further one metre extension been a feature of Licchavi urban planning and
to the north running for three metres from one were then later developed into monumental brick
metre to the east of the trench. At the base of the structures or, as at Bhandarkhal and Deopatan, fell
sequence was a sterile, soft, sandy deposit that may into disuse or never gained the patronage for such
have been the sandy tank infill or the natural soil. a development with investment focused on the
Rising to the east, two almost complete vessels were major monuments of Pashupati. The stepped edge
found nestled on the surface of this deposit. These was overlain by a soft silty sand, which contained
vessels were relatively intact and may have been ceramics and small fragments of brickbats, and
dropped into the waters of the tank from its edge, this may be an infill and wash relating to the
either by accident or as part of a ritual deposition. abandonment of this phase of tank.
Above the sand was a soft silty sand containing The next phase of development in Trench
brickbats and ceramics and this may represent the 1 relates to the construction of a large brick wall
primary tank fill once the feature went out of use running north-south across the trench and to the
and was no longer cleaned out. Cutting through east side of the trench. This wall sat in a foundation
this material and into the sterile sandy base of the cut, which contained large fragments of pot
tank was a large pit, with a rounded base, which within its fill. Badly robbed, the wall survived to
97
seven courses in the south and two in the north. All bulk undiagnostic brick, ceramics
Initially thought to be a tank edge, the wall may and tile from Trenches 1 and 2 were counted and
relate to later occupation at the edge of the tank. weighed from all contexts excavated. Ceramic
It was associated with a clay surface, which may wares included Red Ware, Coarse ware, Black
represent an occupation horizon. Overlying this Ware and Red Slipped Ware and we are currently
wall, and running east-west was a narrow brick developing diagnostic ceramic typologies with
wall, measuring 0.40 metres wide, which may reference to those at Harigaon (Verardi 1992) and
represent a buttress or railing on the edge of the Dumakhal (Khanal and Riccardi 2007). We are
tank (Figure 9). This balustrade was damaged and continuing to develop a full artefactual typology
appeared to split towards the western slope of the and catalogue that will be linked to these previous
tank but was also robbed out with a robber pit of excavations and also to our programme of
loose rubble material including charcoal, brick scientific dating. In lieu of this, several special finds
and stone cut through it to the east. provided evidence of the Licchavi occupation of
The latest phases of the Trench related to Bhandarkhal including a decorated Red Slipped
several phases of ephemeral edging of brick and ware body sherd with a stamped circular motif
tile fragments. Represented by linear features (Figure 13), very similar to the Pashupati coins
running north-south, it is suggested that these with raised dots forming a circle that frames
alignments may have been temporary tank a depiction of a bull in the centre. Similarly, a
edgings formed from rubble. The clearing of a further decorated body sherd with a stamped
similar alignment on the northern edge of the tank design depicted a trident and the head of a bull,
led to the opening of Trench 2 (Figure 10). Trench as well as four and a half characters of Licchavi
2 was originally a two by one metre trench opened script (Figure 14). We also uncovered a Pashupati
over east-west running alignment of tile and brick, coin with a depiction of the Yogi (Figure 15). In
of which the upper surface was visible through addition to these finds, we recovered several
the topsoil. Seemingly constructed from debris, cylindrical terracotta objects that tapered at one
there is again the potential that this alignment end. Verardi suggested that similar cones found
may represent an ephemeral edging to the tank, at Harigaon, one of which was associated with a
much like those identified in the upper surfaces tank, may have been related to offering materials
of Trench 1. It may also relate to the consolidation (Verardi 1992: 168). Such a hypothesis requires
of the tank edge in the later phases of occupation, further investigation and further analysis of the
perhaps relating to earthquake damage or a phase entire artefactual assemblage relating to context
relating to squatter-like occupation of the site. The and scientific dates.
eastern half of this alignment was removed and
a slot trench measuring six metres north to south 4.2 Trench 3
and one metre east to west was excavated towards After a hiatus of activity due to the
the centre of the tank down the slope of this 2015 Gorkha Earthquake, the team returned to
monumental feature. This was later extended by a Bhandarkahl in February-March 2016 to undertake
further one metre to the north, resulting in a seven further excavations. Trench 3 was located in
metre long slot trench. Several other features were the southern half of the walled enclosure of
identified that bore a similarity to those in Trench Bhandarkhal and was laid out on a north-south
1. A narrow brick wall, constructed of at least seven alignment at a distance of one hundred metres west
courses of brick, ran north-south into the tank of the main gate, perpendicular to the main east-
(Figure 11). This also may represent a balustrade/ west path. As stated above, geophysical survey
railing running from the tank edge and was sealed had identified several parallel east-west features
by several deposits and a brick paving, which in running across this area and the excavations
turn was found underneath the ephemeral tank were targeted at characterising and dating these
edges. The base of the tank had a sterile sandy fill, subsurface features. The trench was initially one
like that in Trench 1 and several phases of infill metre wide and 30 metres long, running north-
were identified, which contained brickbat rubble south between 60 and 90 metres from the southern
as well as ceramic sherds (Figure 12). edge of the main central east-west path. Only the
98
20 metre portion between 60 and 80 metres was were bulk-counted and weighed. Much of the
excavated with the area between 80 and 90 metres artefactual material related to modern finds in the
subject to surface cleaning. Once archaeological upper levels, particularly topsoil, with substantial
features had been recognised within the slot, numbers of objects of glass, plastic, metal and
the main portion of the trench was extended to even fabric. From secure archaeological deposits,
horizontally expose structures, resulting in an we identified several artefact types comparable to
extension of 15 metres north-south and five metres those from Trenches 1 and 2. As in our previous
east to west from the original one metre wide slot excavations in the vicinity of the tank at the centre
at a distance of between 65 and 80 metres from of Bhandarkhal, we also recovered a number of
the main east-west path. Additionally, a further cylindrical terracotta objects that tapered at one
one metre east-west by two metres north-south end. As previously noted, Verardi suggested that
extension was excavated to the east of the original similar object from Harigaon may have represented
one metre slot on account of the presence of a large offerings (Verardi 1992: 168), although this is still
architectural block under the west-facing section unconfirmed. A single heavily corroded copper
wall at a distance of between 70 and 72 metres to alloy coin was recovered, although whether it is
the south of the main path (Figure 16). a Pashupati coin will have to await conservation.
The excavations here successfully exposed We also recovered a number of small oil lamps,
the presence of at least three major phases of spouted vessels, stamped sherds, slag, stone
occupation in the vicinity of the trench. The oldest mortar and pestles and moulded and carved
was encountered at the base of a deep one metre terracotta objects. In addition, a large quantity of
wide section cut into the south-facing section of undiagnostic ceramic sherds was recovered as
the extended trench. Excavated in order to fully well as 966 fragments of roof tile, weighing 32.231
investigate the stratigraphy of the locality, virgin kilograms. These artefactual findings suggest a
soil was identified at a depth of 1.83 metres similar corpus of artefacts across Bhandarkahl and
below the surface and the stratigraphy confirmed perhaps the markers of Licchavi material culture,
the presence of earlier human activity in the though as noted above, a full artefactual typology
locality of the trench. Primarily represented by and catalogue is being developed to link with
the ceramic-rich fills of a pit cut (Figure 17), the previous excavations and also to our programme
recovery of fragments of brick and tile indicate the of scientific dating before any concrete hypotheses
nearby presence of early structures. These early can be made.
levels were then sealed below the next phase of
occupation, represented by the construction of a 5. Conclusion
major east-west oriented brick paved platform, Our preliminary excavations at
measuring at least six metres in width (Figure 18). Bhandarkhal have succeeded in highlighting the
Although badly robbed, its southern edge was presence of earlier archaeological sequences at
defined by what appeared to be foundations of Pashupati, some of which were characterised
stone cobbles and its western end by brick wall. by non-durable remains. Whilst waiting for the
A well-constructed stone drain with a stone block results of our scientific dating, it is not possible to
bottom ran from the structure in a south-westerly yet categorically state the dating of these phases
direction (Figure 19). The complex had been badly but our investigations have begun to characterise
robbed but the distinct scatter of tile, brick and the nature of this activity. In Trenches 1 and 3,
stone architecture suggested that the structure we found clear evidence of cut pit features. Both
had been monumental in nature. Following the areas highlight that activity at Bhandarkhal
deposition of silty sands over the robbed remains and, by extension within the Pashupati area,
of the platform, the final occupation comprised may not always relate to durable monumental
the digging of two large post pits after which the architecture. Whilst the pit cut in the tank may
site was abandoned (Figure 20). be a much later feature, the pit cut in Trench 3
During the excavations at Trench 3, over suggests that these features may have formed an
100 special finds were catalogued and undiagnostic architectural characteristic of earlier occupation at
brick, ceramics and tile from the excavation trench the site. In addition, rather than the conforming

99
to architecture seen in the medieval tanks of Furthermore, we are undertaking visitor
the Durbar Squares of the Kathmandu Valley, and business surveys to understand the social
where brick lined tanks are present, the tank at and economic impacts of Pashupati in the present.
Bhandarkhal had an earthen edge. However, its We will also map and record epigraphic evidence
size suggests that it was still a major undertaking across Pashupati to understand changing
and although not of durable materials, was a major patterns of patronage, providing a historical
feature within Pashupati. Furthermore, the finds context to our archaeological investigations.
of brick and tile and large fragments of worked This will also provide additional measurements
stone in Trench 3, also suggest that monumental and photographic records of inscribed and non-
structures were present within this area of the inscribed objects dating to the Licchavi period.
Pashupati Complex. By extrapolating from the Rubbings have been taken of newly recorded
exposed monument in Trench 3 to the results of inscribed objects, and further analysis will be
the geophysical survey, it would seem plausible undertaken on the inscriptions, and compiled
that much of the now wooded enclosure lies above into a database to aid future management and
a series of monumental platforms running east to protection. The combination of these approaches
west with the large tank at its centre. and methodologies will not only shed light on
In addition to our excavations, further the early occupation and the social and economic
archaeological investigations are being undertaken character of the site and its development in the past,
at Bhandarkhal and across Pashupti. The project it will also lead to the devising of an Archaeological
is multi-disciplinary and over the course of our Risk Maps for Pashupati, protecting cultural
investigations we have, and will continue to, heritage from future development by guiding
integrate topographical mapping, geophysical management of its subsurface archaeology, whilst
survey, auger coring, geoarchaeology and also providing information on the current social
epigraphic analysis with our systematic excavations and economic impacts of pilgrimage, tourism and
and artefactual analysis. Geophysical survey has site use to facilitate sustainable development of
enabled the identification of areas of archaeological the site. For instance, the project's Archaeological
interest and areas that require protection from Risk Map for Bhandarkhal will now be used by the
development. Combined with auger coring, which Pashupati Area Development Trust to design the
has facilitated the identification of the depth and rehabilitation of this area of the site to ensure that
spread of cultural material, this has allowed for amenities are provided to pilgrims and residents
rapid site assessment that have revealed areas of whilst protecting the vulnerable subsurface
significant subsurface heritage as well as providing heritage.
the ability to record general macro-stratigraphic Finally, in relation to the aftermath of the
details of the broader site. These surveys have 2015 earthquakes, we should also recognise that
guided our excavations and, as stated above, we plans to swiftly reconstruct temples and buildings
are undertaking geoarchaeological analysis of the will necessitate research interventions and
excavated stratigraphies to date the origins and excavations to establish evaluate the strength and
subsequent phases of development and to assess the phasing of foundations. Detailed recording and
relationship between these to local environmental scientific analysis of these interventions is critical as,
conditions. We will process Optically Stimulated discussed above, almost all published architectural
Luminescence (OSL) samples taken from soil studies of the Kathmandu Valley are focused on
stratigraphy to provide absolute scientific dates standing remains and do not contain plans or
for key contexts. Thin section micromorphology sections of foundations (Coningham et al. 2016a).
samples have also been taken to answer key We argue that all post-earthquake interventions
questions relating to site formation processes should be accompanied by archaeological
including the assessment of the change from rural excavations in order to evaluate the sub-surface
to urban forms of land use as well as providing stability of foundations as well as evaluate and
insights into the emergence and development of sequence evidence for prior structures at that
tank-based water management at Bhandarkhal and place. As a result, it is recommended that post-
the Kathmandu Valley. disaster rescue excavations are undertaken within
100
the Pashupati UNESCO World Heritage Site prior Salgaha Watta: Volume 1 The Site. Society for
to the rebuilding of collapsed structures as well as South Asian Studies Monograph Series.
accompanying the demolition of badly damaged Oxford: Archaeopress.
structures in accordance with the Department of Coningham, R.A.E., Acharya, K.P, Strickland,
Archaeology’s newly agreed draft Conservation K., Davis, C.E., Manuel, M.J., Simpson, I.,
Guidelines for Post 2015 Earthquake Rehabilitation: Gilliland, K., Tremblay, J., Kinnaird, T. and
Conservation Guidelines (CGPERCG 2015). Sanderson, D. 2013a. The Earliest Buddhist
Shrine: Excavating the Birthplace of the
6. Acknowledgements Buddha, Lumbini (Nepal) Antiquity 87(338):
We would like to acknowledge the support 1104-1123.
of the following individuals and institutions: Coningham, R.A.E. and Acharya, K.P. 2013.
Dr Govind Tandon, Mr Ramesh Kumar Identifying, Evaluating and Interpreting the
Upreti, Mr Taranath Subedi, Mr Bharat Marasini, Physical Signature of Lumbini for Presentation,
Mr Sitaram Risel, Mr Rewati Raman Adhikari, Mr Management and Long-term Protection: Report
Tej Bahadur, Mr Krishna, Mrs Kanchi Tamang and of Season Three of Activity Two. UNESCO
other staff at the Pashuapti Area Development Report.
Trust; Mr Bhesh Dahal, Ms Manju Singh Bhandari, Coningham, R.A.E and Gunawardhana, P. 2013.
Mr Bishnu Prasad Pathak and Mr Mohan Singh Anuradhapura: Volume 3. The Hinterland. BAR
Lama, Rojendra Sthapit and Shyamsundar International Series. Oxford: Archaeopress
Rajbansi of the Department of Archaeology, Coningham, R.A.E., Hale, D., Manuel, M. J., Davis,
Government of Nepal; Mr Bharat Mani Subedi; C.E. and Villis, R.G. 2013b. Archaeological Risk
Raju Rimal; Mr Duncan Hale and Ms Patricia Assessment Evaluations at the UNESCO World
Voke of Durham University; Prof Ian Simpson of Heritage Sites of Paharpur, Mahasthangarh
the University; Ms Nabha Basnyat-Thapa of the and Bagerhat. Report for the Department of
UNESCO Kathmandu Office; Mr Ranjan Dulal, Ms Archaeology (Government of Bangladesh)
Sunita Bhadel, Ms Shanti Sherma, Mr Shivasharan on behalf of the South Asia Tourism
Rajbhandhari, Mr Jagat Bahadur Katuwal, Ms Infrastructure Development Project
Sunita Hamal, Mrs. Shakuntala Acharya, Mr Coningham, R.A.E., Acharya, K.P. and Manuel,
Urban Coningham, Mr Benjamin Parker, Mr Jay M.J. 2014. Strengthening Conservation and
Thapa and Mr Kulnath K.C. Management of Lumbini, the Birthplace of the
Generous financial assistance for the Lord Buddha, World Heritage Property (Phase
archaeological investigations was provided by the II): Report of the first (Jan-Feb 2014) season of
Pashupati Area Development Trust (Government field activities. UNESCO Report.
of Nepal), The British Academy Stein-Arnold Coningham, R.A.E., Acharya, K.P., Davis, C.E.,
Fund and Durham University. Kunwar, R.B., Tremblay, J.C., Scmidt, A.,
Simpson, I. and LaFortune-Bernard, A.
7. References 2016a. Post-Disaster Rescue Archaeological
Amatya, S. 2007. Monument Conservation in Nepal: Investigations, Evaluations and
My Experience with the World Heritage Sites Interpretations in the Kathmandu Valley
of the Kathmandu Valley. Kathmandu: Vajra World Heritage Property: Recommendations
Publications. from a UNESCO mission in 2015. Ancient
Amatya, S. 2011. Archaeological and Cultural Nepal : 72-92.
Heritages of Kathmandu Valley. Kathmandu: Coningham, R.A.E., Acharya, K.P., Davis, C.E.,
Ratna Pustak Bhandar. Kunwar, R.B., Tremblay, J.C., Simpson, I.A.
Basukala, B, Gutschow, N. and Kayastha, K. 2014. and Schmidt, A. 2016b. Preliminary Results of
Towers of Stone: Sikhara Temples in Bhaktapur, Post-Disaster Archaeological Investigations
Vatsala and Siddhilaksmi. Kathmandu: Himal at the Char Narayan Temple and within
Books. Patan Durbar Square, Kathmandu Valley
Coningham, R.A.E. 1999. Anuradhapura: The British UNESCO World Heritage Property (Nepal).
- Sri Lankan Excavations at Anuradhapura Ancient Nepal : 52-71.
101
Coningham, R.A.E., Acharya, K.P., Davis, C.E., Joshi, H. 1974. Ancient Inscriptions of Nepal.
Kunwar, R.B., Simpson, I.A., Schmidt, A. and Kathmandu: Royal Nepal Academy.
Tremblay, J.C. 2016c. Preliminary Results of Khanal, M.P. and Riccardi, T. 2007. Archaeological
Post-Disaster Archaeological Investigations Excavation in the Kathmandu Valley: A Report on
at the Kasthamandap and within Hanuman the 1984-85 and 1988-89 Seasons at Dumakhal.
Dhoka, Kathmandu Valley UNESCO World Harvard: Harvard Oriental Series.
Heritage Property (Nepal). Ancient Nepal : Korn, W. 2007 (second Edition). The Traditional
28-51. Architecture of the Kathmandu Valley.
Coningham, R.A.E., Acharya, K.P., Davis, C.E., Kathmandu: Ratna Pustak Bhandar
Kunwar, R.B., Tremblay, J.C., Schmidt, A. Korn, W. 2014. The Traditional Newar Architecture
and Simpson, I. 2016d. Preliminary Results of of the Kathmandu Valley: the Shikaras.
Post-Disaster Archaeological Investigations Kathmandu: Ratna Pustak Bhandar.
at the Vatsala Temple and within Bhaktapur Michaels, A. 2008. Śiva in Trouble: Festivals
Durbar Square, Kathmandu Valley UNESCO and Rituals at the Pashupatinatha Temple of
World Heritage Property (Nepal). Ancient Deopatan. Oxford: OUP
Nepal : 3-27. Michaels, A. and Tandon, G. 1994. Deopatan and
Coningham, R.A.E., Acharya, K.P., Davis, C.E., the Temple of Pashupati. In Hutt, M. (ed.)
Strickland, K.M., McDonnell, G., Tremblay, Nepal: A Guide to the Art and Architecture
J., Manuel, M.J. and Bidari, B. 2016f. Defining of the Kathmandu Valley: 180- 195. Stirling:
the Chronological and Cultural Sequence of Kiscadale Publications.
Mound V Tilaurakot (Nepal): a report on Mirnig, N. 2013. Favoured by the Venerable
pilot excavations conducted in 1999. Ancient Lord Pashupati. Tracing the Rise of a New
Nepal 190: 18-29. Tutelary Deity in Epigraphic Expressions of
CGPERCG. 2015. Conservation Guidelines for Post Power in Early Medieval Nepal. Indo-Iranian
2015 Earthquake Rehabilitation: Conservation Journal 54(3/4): 325-347.
Guidelines 2015 (draft). Kathmandu: Regmi, D. 1965. Ancient Nepal. New Delhi: Rupa
Department of Archaeology. and Co.
Davis, C.E., Coningham, R.A.E., Acharya, K.P., Regmi, D. 1983. Inscriptions of Ancient Nepal. New
Simpson, I.A., Tremblay, J., Kunwar, R.B., Delhi: Abhinav Publications.
Manuel, M.J., Krishna Bahadur, K.C. and Strickland, K.M., Coningham, R.A.E., Acharya,
Bidari, B. 2016. Re-investigating Tilaurakot’s K.P., Schmidt, A., Simpson, I.A., Kunwar,
Ancient Fortifications: a preliminary report R.B., Tremblay, J., Manuel, M.J., Davis, C.E.,
of excavations through the northern rampart Krishna Bahadur, K.C. and Bidari, B. 2016.
(Nepal). Ancient Nepal 190: 30-46. Ancient Lumminigame: A Preliminary
Di Castro, A.A. 1997. Appendix: Exhibition on the Report on Recent Archaeological
Works carried out by the Italian Team from Investigations at Lumbini’s Village Mound
1984-1995. In Veradi, G., Di Castro. A.A. (Nepal). Ancient Nepal 190: 1-17.
and Garbini, R. (ed.) Nepalese and Italian Slusser, M.S. 1982. Nepal Mandala: a Cultural Study
Contributions to the History and Archaeology of of the Kathmandu Valley. Princeton: Princeton
Nepal: 133-145. Rome: IsIAO. University Press.
Gutschow, N. 1997. The Nepalese Caitya: 1500 Tiwari, S.R. 2001. The Ancient Settlements of the
years of Buddhist Votive Architecture in the Kathmandu Valley. Kathmandu: Centre
Kathmandu Valley. Stuttgart. for Nepal and Asian Studies, Tribhuvan
Gutschow, N. 2011. Architecture of the Newars: a University
History of Building Typologies and Details in TPU 2003a. News from Nepal. The Past Uncovered.
Nepal. Chicago: Serindia Publications. February 2003: 2
Hutt, M. (ed.). 1994. Nepal: A guide to the Art and TPU 2003b. Asia Urbs Project. The Past Uncovered.
Architecture of the Kathmandu Valley. Stirling: October 2003: 4
Kiscadale Publications. Veradi, G. 1992. Excavations at Harigaon Kathmandu.
Rome: IsMEO
102
Figures

Figure 1: Main monument zone of Pashupati looking across the Bagmati River towards the Pashupati Temple.

103
Figure 2: General view across Bhandarkhal.

Figure 3: Trench locations, geophysical results and topographic map of Bhandarkhal.

104
Figure 4: Geophysical interpretation of Bhandarkhal.

Figure 5: Location of auger core transects across Pashupati.

105
Figure 6: Pit cut with three fills in Trench 1.

106
Figure 7: Brickbat platform within Trench 1 below later north-south wall.

Figure 8: North-facing section of Trench 1 showing stepped tank edge and later infills.

107
Figure 9: North-south wall with intact ceramic vessels within fill of the wall cut and east-west brick balustrade in Trench 1.

Figure 10: Ephemeral east-west alignment constructed from brick and tile fragments in Trench 2.

108
Figure 11: North-south buttress wall alignment in Trench 2

Figure 12: South and east-facing sections of Trench 2.

109
Figure 13: Stamped Red Slipped ware sherd of pottery Figure 14: Stamped red ware body sherd depicting a bull
depicting a bull within a circular motif. and a trident as well as five characters of Licchavi script
within a circle.

Figure 15: Pashupati coin depicting the Yogi.

Figure 16: Trench 3 looking south.

110
Figure 17: Early pit cut in west-facing section of Trench 3

Figure 18: View of east-west oriented monument with paved floor and stone-lined drain in Trench 3 looking west.

111
Figure 19: Detail of stone-lined drain looking west.

Figure 20: Post pit belonging to the final phases of occupation in Trench 3.
112