Plan view of the so-called "granary" or "parallel-wall structure" on Mound F at Harappa indicating areas of HARP excavations conducted in 1997 and 1999. Note that the structural remains surrounding the "granary" are, for the most part, later than the original "granary" structure. From M.S. Vats Excavations at Harappa (1940).

Overview of the "granary" area looking towards the southeast. The walls have been partly restored for conservation purposes by the Department of Archaeology and Museums, Government of Pakistan.

Excavations in 1997 at the southeast corner of the "granary" area were undertaken to recover a full sequence of pottery, architectural features, and inscribed objects. Here workers have found a seal near the base of the excavations in Trench 41NE that dates only somewhat later than the original "granary" structure.

In the upper levels of a small baked brick structure with sump pots was found. This structure, possibly a latrine, is on the northern margin of a major east-west street that ran along the southern edge of the "granary" area. It dates to late in Period 3C, perhaps 200-300 years after the original "granary" structure had been buried by later construction.

Intaglio seal (H97-3433/7617-01) with script and unicorn motif found in Trench 41NE in 1997. This seal dates to approximately 2200 BC, at the transition between Harappa Periods 3B and 3C.

J. Mark Kenoyer sitting in the right foreground takes notes during excavations at the southwest corner of the "granary" (Trench 41SW). The baked brick revetment of the "granary" platform is in the foreground, the southernmost baked brick wall of the "granary" itself in the middle ground.

A massive baked brick revetment wall [329] surrounds the solid mud-brick foundation platform [50] of the "granary" that measures approximately 51 meters north-south and 41 meters east-west. Based on analysis of the pottery and other finds from below and against the revetment wall, it is possible to date its construction to near the end of Harappa Period 3B, between 2300 and 2200 BC.

A broken steatite tablet (H97-3431/7615-01) was recovered from deposits just below the fired brick revetment of the "granary" platform and dates to Harappa Period 3B (2450-2200 BC).

A small faience ram amulet recovered from the fill associated with the southwestern edge of the "granary" platform. (Trench 41SW, H97-3434/7650-01, Length = 21.8 mm)

Excavations in 1997 at the northwest corner of the "granary" platform were undertaken to follow the outline of the structure and document its construction (Trench 41NW)).

Section through the northwestern portion of the "granary" platform directly below the baked brick "granary" walls (Trench 41NW). Similarity of composition of the mud-bricks in the northwestern, southwestern, and southeastern parts of the "granary" platform as well as nearly identical elevations of the top of the platform and of the foundation trenches in the different excavated areas indicate that the entire structure was built at one time.

Overview looking north of excavations at the southeast corner of the "granary" structure undertaken in 1999 (Trench 41C). The higher east-west running walls in the left of the image and the ruined structures in the right of the image all post-date the original "granary" structure.

After clearing the overlying silt, the original forms of the baked brick walls and hollow buttresses of the "granary" could be made out. The three high walls in the upper left of the image are part of a later rebuilding of the entire structure after ca. 2200 BC.

Clearing outside the southeast corner of the "granary" revealed the underlying mud-brick platform and the top of the baked brick revetment.

Aerial view of the exposed southeastern portion of the "granary" structure shows the nature of brick bonding and the empty sockets that would have held wooden beams and supports.

Richard Meadow photographing the southeastern portion of the "granary" structure.

Deep digging at the southeast corner of the "granary" revealed an earlier building [wall 330] constructed along the same east-west alignment. This structure could have been built as early as the beginning of Period 3B, ca. 2450 BC. The currently visible "granary" appears to have been constructed ca. 2300 BC, with a third structure built on top after 2200 BC.

Detail view of a hollow area that would originally have held a wooden beam bonded into the baked brick structure. In the background is a wall remnant from the later rebuilding of the "granary".

Hollow baked brick buttresses were later built up against the original "granary" structure on top of a shallow mud-brick platform [400] that itself overlies the mud-brick platform of the original "granary". Below these platforms is baked brick wall [330] (seen here with the scale lying on it) of the structure preceding the "granary". The precise function of the buttresses is not known, but eventually they became filled with trash and silt as the street levels outside the building grew higher.

Excavation of one buttress [99] shows how the silt and garbage from the street spilled into the hollow area from the outside of the "granary", eventually blocking it entirely.

Detail view of the fill inside the "granary" buttress. Large pottery sherds, bone, and baked brick fragments lie on the bottom with finer silts on the top.

All of the sediment and artifacts were removed from "granary" buttress [99] for analysis, revealing the method of corbelled arch construction.

After excavation, "granary" buttress [99] was reconstructed using the original bricks set into modern clay mortar.

Excavations were conducted in the narrow space running west from the buttresses and between the interior walls to determine what was inside of the "granary" structure.

In addition to documentation with still photography, J. Mark Kenoyer videoed the "granary" areas excavated and narrated details of the stratigraphy and of the brick construction.

This image looking east shows baked brick rubble and trash that had spilled through the corbelled arch of a buttress from the exterior street into a then empty channel between two "granary" walls.

After clearing an interior area in the southeast part of the "granary", it became evident that the inner baked brick wall of the original structure had been dismantled by the ancient Harappans before they filled in the resulting void and built the overlying parallel-wall structure. Note the protruding baked bricks against which the pick-axe is resting. These bricks demonstrate the bonding that existed between the original interior and exterior walls of the building.

Looking toward the interior of the "granary" structure, the void of the original baked brick wall (vertical scale) is filled with small pieces of broken baked bricks and mud-brick rubble, while the original open space between the walls (horizontal scale) is filled with compact and relatively clean clayey sediment. There is no indication of the remains of grain or of storage containers.

After completing the excavation, the trenches were filled with sifted sediment and the "granary" structures were covered with a protective plaster made of clay and straw.

Excavations along the northeast perimeter of Mound F (Trench 41NE) identified traces of a massive mud-brick wall that protected the interior settlement (to the left) from floods and unwanted visitors. This perimeter wall of Mound F was buried under the backdirt of the excavations of the 1920s and 1930s.

Inside the city wall were domestic structures with hearths and large jars, part of structures that had encroached onto what used to be street along the inner face of the wall (Trench 41NE).

Detail of domestic structures with hearths and large jars set into the floor (Trench 41 NE).

Large jar set into the floor of one of the small rooms encroaching onto the street along the inside of the city wall of Mound F (Trench 41NE).

Excavation of the fill inside the jar (image 303) indicates it was left exposed for some time and gradually filled with silt as the room became covered with debris (Trench 41NE).

Also in a room encroaching onto the street were the upturned rim of a jar (on left) and two bases (center and right) that were used as hearths (Trench 41NE). The pottery base hearths could have been lifted and moved to different parts of the room thus serving as portable heaters as is still done today during the cold winters of the Punjab.

William Belcher has just removed charcoal for archaeobotanical analysis and radiocarbon dating from one of the hearths (image 305) and is making out a label for the sample.

Richard Meadow photographing the exposed wall and interior domestic features (Trench 41NE).

The interior of the city wall shows mud bricks made from different colors of clay used during different phases of construction.

Superimposed street levels inside the city wall indicate that this area was kept clear until the last phase of the Period 3C occupation when structures were built encroaching onto the street (Trench 41NE).

Section drawing of the cut through the Mound F city wall and adjoining interior street, facing east (Trench 41NE).

Excavations in 1999 along the western of edge of Mound F (Trench 43) revealed fallen rubble from walls that had collapsed late in the Harappan occupation of this area of the site. Drummers standing on the piles of dirt left from earlier excavations help to lighten the work.

Overview of the upper levels of the Trench 43 excavations in 1999 shows the eroded city wall in the background and the fallen baked brick walls of late Period 3C structures in the foreground.

After removing the fallen walls, the interiors of the rooms were found to be filled with surprises. Here in Room 202 were scattered clay tops, beads, and inscribed seals (Trench 43).

Detail of terracotta tops found in Room 202 (Trench 43).

Faience button seal (H99-3814/8756-01) with swastika motif found on the floor of Room 202 (Trench 43).

Glazed steatite gaming piece, faience bead, and unfinished steatite seal (H99-3860/8756-30) from Room 202 (Trench 43).

Black and white spiral faience bead (H2000-5005/9845-48 from Trench 43).

Large lumps of a bluish grey paste that may have been used to make faience objects. Faience is made by melting quartz rock and then regrinding the glassy frit to make a paste that is then fired once again. Indus faience is sometimes multicolored (see images 82, 216, 317, 319).

Unique cylindrical object made of maroon and white faience (H99-3856/8756-26 from Trench 43, length = 21.9 mm). Numerous examples of this identical type of bead have been found in other parts of Harappa, but this may have been where they were actually made.

A broken terracotta bird cage (near the scale at right) was discovered next to the remains of a broken storage jar that had been reused as a hearth (near the scale at left) (Trench 43).

Reconstructed bird cage (H99-4065/8768-7). On the other side it has a slot for a terracotta door, but no top was found. The exterior was decorated with red slip and black paint (Trench 43).

Terracotta bangle fragments decorated with red trefoils outlined in white on a green ground from late Period 3C deposits in Trench 43. This image shows both sides of the two fragments (H98-3516/8667-01 & H98-3517/8679-01).

Detail of terracotta bangle with red and white trefoil on a green background (H98-3516/8667-01 from Trench 43). Trefoil motifs are carved on the robe of the so-called "priest-king" statuette from Mohenjo-daro and are also known from contemporary sites in western Pakistan, Afghanistan, and southern Central Asia.

Black steatite wig from late Period 3C deposits in Trench 43 (H98-3521/8668-02). This small stone hairpiece, here displayed on a modern clay mannequin, may have been set on an alabaster head like similar pieces found in western Pakistan, Afghanistan, and Southern Central Asia.

Drawing of black steatite wig from Trench 43 (see image 324).

Large unicorn seal (H99-4064/8796-01) found on the floor of Room 591 in Trench 43, dating to late Period 3C. This is one of the largest seals found from any Indus site.

Excavations in Trench 43 revealed collections of broken pottery that date to the final phases of the Harappan occupation of Period 3C, ca. 1900 BC. Note the antler piece (image 328) to the right of the broken dish.

Part of an antler, including the burr and the brow tine, shed from a large deer, probably sambhar (Cervus unicolor) or swamp deer (Cervus duvauceli) (H99-3828/8763-01 from Trench 43). This specimen may have been used as a pick or possibly a punch for flint knapping.

One of the falling walls in Trench 43 crushed a vessel that belongs to Harappa Period 4 (ca. 1900-1800/1700 BC). In this part of the mound, there appear to be transitional levels between the Harappan and Late Harappan periods.

This Period 4 globular pot (H99/8763-503) was found crushed beneath a fallen wall in Trench 43 (see image 329). The rough textured exterior was covered with a thick layer of blackened clay that indicates it was a type of cooking pot and not for water storage as earlier scholars have argued.

Period 4 globular pot (H99/8763-503) from Trench 43 after reconstruction. (See images 329 & 330).

Period 4 (Late Harappan transitional) kiln with hollow lower fire box and arched floor with holes for allowing heat to enter the upper firing chamber. This type of kiln was introduced at Harappa ca. 1900 BC and allowed the potters to reach higher temperatures more efficiently (Trench 43).

Looking into the hollow fire box of the Period 4 kiln with arched floor (see image 332). The column of soil in the center was left for support of the floor (Trench 43).

Collection of reconstructed pottery from the late levels of Trench 43. These shapes include the final Harappan forms (late Period 3C) and transitional Late Harappan period shapes (Period 4).

Continued excavations in 2000 focused on the northern area of Trench 43. Here too were found fallen walls and well preserved living floors. Note the Period 3C circular platform in the background at a considerably lower level.

Overview of Trench 43 in 2000 looking north, showing the HARP-exposed circular platform in the foreground and the "granary" area in the background. Note the wall voids to the west, south, and east of the circular platform (see also image 356).

Pakistani excavators working in Trench 43. The man on the right has just discovered a spherical agate weight (image 338) while cleaning the section for photography.

Spherical agate weight with one side flattened (H2000-4496/9880-01). This weight does not conform to the standard Harappan weight system and may indicate the use of a separate weight system for international trade with Central Asia or other regions.

Excavations of a large late Period 3C pit in Trench 43 resulted in the discovery of many broken figurines of humans and animals.

Collection of human and animal figurines from a large refuse dump in a pit in Trench 43.

Female figurine (H2000-4997/9811-02) from Trench 43.

Female figurine (H2000-4993/9845-07) from Trench 43.

Male figurine (H2000-4975/9878-07) from Trench 43.

Male figurine (H2000-4976/10,024-01) from Trench 43.

Male figurine (H2000-4974/9845-01) from Trench 43.

Randall Law and Shamoon excavating a red stoneware bangle in Period 3C levels just below the surface in Trench 43.

Red stoneware bangle (H2000-4490/9843-01) with no inscription. The lack of inscription may indicate that this may have been a place where the bangles were stored prior to inscribing them for distribution.

While clearing a floor area in Trench 43, J. Mark Kenoyer discovers a broken unicorn seal.

Unicorn seal (H2000-4500/10007-01) found on the floor of Room 603 in Trench 43 (image 348), before conservation.

Unicorn seal (H2000-4500/10007-01) after conservation. Note the deeply chiseled engraving of the script similar to that found particularly on Period 3C rectangular seals.

Copper tablet (H2000-4498/9889-01) with raised script found in Trench 43.

Terracotta bead with molded script (H2000-4494/9881-01) found in Trench 43.

Circular platforms in the southwestern part of Mound F excavated by M.S. Vats in the 1920s and 1930s, as conserved by the Department of Archaeology and Museums, Government of Pakistan.

Plan of Vat's excavations showing circular platforms. In some cases remnants of the baked brick walls that probably surrounded each platform can be seen on the plan, although earlier and later walls are also shown. From M.S. Vats (1940) Excavations at Harappa.

The circular platform excavated by Wheeler in 1946 (left) and the one excavated by HARP in 1998 (right). Both of these platforms were found inside small square rooms that originally had baked brick walls, subsequently removed by brick robbers (Trench 43).

Detail view of the HARP-excavated platform in Trench 43 with Wheeler's platform to the east (toward the top of the image). Note the mud-brick wall foundations that surround each platform to the east, south, and west (the north walls remain unexposed). Traces of baked brick thresholds can be seen on the right (south). See also image 336.

A large concentration of straw impressions was found in one part of the floor next to the platform, but there is no evidence of chaff from processing grain as was suggested by earlier excavators (Trench 43).

J. Mark Kenoyer excavating and sampling the sediments associated with the HARP-excavated platform, which was partly robbed of baked bricks during the Harappan period itself (Trench 43). Pottery found under the platforms permits them to be assigned to Harappa Period 3C, probably toward the middle of that period (ca. 2100-2000 BC).

Greenish clay layers were found in a deep depression in the center of the HARP-excavated platform. One theory that is being investigated is that the platform and the central pit were used for production of indigo dye (Trench 43).