This volume is the third in this hugely important series that aims to produce a complete Corpus of Indus Seals and Inscriptions so far known [CISI]. CISI 1 (1987) and CISI 2 (1991) published respectively seals from collections in India and Pakistan. The project is led by Asko Parpola who, in a brief preface (pp. vi-xv) traces its history back to 1964, and goes on to discuss the contents of the present volume (CISI 3.1) namely the “42 objects in 14 museums, in 11 cities and 8 countries” listed on pp. vii-viii and “447 [previously published] objects from Harappa . . . [that] cannot be found in any museum; the figure is 266 from Mohenjo-daro” (the original field registers for this latter site are being published by Michael Jansen and colleagues of the Technical University of Aachen). Parpola discusses the sorry history of the Harappan finds before and after partition, many of which are known to have been stolen (pp. viii-x). He explains how photographs were traced and copied, and how pieces were identified. New excavations planned at Mohenjo-daro “did not materialise”, but Richard Meadow and Mark Kenoyer, the directors of the Harappa Archaeological Research Project (HARP), have published all the relevant material from their new excavations in the present volume, together with an introductory essay (see below, pp. xliv-lviii). The contents of Volume 4 are summarised (pp. xiv-xv) and will include concordances and “an updated computer edition of all readable texts in the Indus script”. There is a general bibliography on p. xvi, and detailed bibliographies follow their respective essays.

The first essay, by Ute Franke (pp. xvii-xliii), is entitled “From the Oxus to the Indus: Two compartmented seals from Mohenjo-daro (Pakistan)”. These seals are illustrated on pp. xxviii Fig. 3:1, xxxii Plan 2, xxxiii Fig. 5, and xxxv Fig. 6, their background, where they were found, comparative material and chronology (see p. xxv, Table 1) are discussed in a well-illustrated paper. Franke does not mention the useful publication which came out of a 1988 conference at the British Museum, The Indian Ocean in Antiquity (London, 1994) edited by Julian Reade, in which I tabulated the occurrences of similar motifs in Mesopotamia, the Gulf, Margiana, the Indus and other locations (D. Collon, “Mesopotamia and the Indus”, pp. 209–229). A reason for all these contacts may well have been the tin trade which is documented in the archives from the merchant colony at Kültepe in central Turkey, and on which an interesting paper has appeared in connection with the tin found on the Uluburun ship that sank off the south coast of Turkey c. 1300 bce. J. Cierny, T. Stöllner and G. Weisgerber have desmonstrated that there were early opencast tin mines in Central Asia (“Zinn in und aus Mittelasien” pp. 431–448 in a work edited by U. Yalçın, C. Pulak and R. Slotta, Das Schiff von Uluburun - Welthandel vor 3000 Jahren, (Bochum, 2005).

The second essay (pp. xliv-lviii), “Inscribed objects from Harappan excavations 1986–2007)” was contributed, as stated above, by the excavators Kenoyer and Meadow. The excavation programme is summarised, including a chronological table (p. xlv) and plan, and a discussion of seal carving and the changes in technology and style over time, followed by a bibliography. The discussion of the inscribed material is very general, with limited reference to actual pieces; two uninscribed “button” seals are illustrated by drawings on p. liii.

Finally, Asko Parpola's short essay (pp. lviii-lix) sheds “New light on ‘Major Clark’, owner of the first published Indus seal”, published in 1875 by Alexander Cunningham (not “1975” as in line 4 of the essay). Cunningham's publication of this seal (H-1048, presented by Clark to the British Museum 1892), “eventually led [Sir John Marshall] to initiate the excavations at Harappa”. It should be mentioned that it was also through Alexander Cunningham that the Oxus Treasure was later acquired by A.W. Franks for the British Museum.

The remainder of the volume consists of illustrations of objects from Mohenjo-daro on pp. 1–136, with numerous photographs of each object in different lighting and at different scales, concentrating on the inscribed face and its impression, but rarely showing the undecorated face, probably because they were not part of the original photographic coverage of missing seals (M-1711B and F on p. 25 are among the exceptions). There are seals (M-1660 to M-1997), impressions on pottery (M-1998 to M-2011), “bas-reliefs” on “tablets” that are circular, square, rectangular, twisted strips, strips with tapering ends, strips with rounded ends and cubes (M-2012 to M-2038); incised tablets that are disc-shaped and rectangular (M-2039 to M-2058); graffiti on pottery, stoneware bangles, ivory sticks, copper objects including axes and gold objects (M-2059 to M-2132), and Addenda and Corrigenda.

This is followed by similar early evidence from Harappa on pp. 137–206: seals (H-2590; H-1020 to H-1080), impressions on pottery (H-1081 to H-1099), rectangular tablets and unusual shapes including fish, discs, prisms with anthropomorphic designs, and cylinders in bas-relief (H-1100 to H-1157), numerous fragments of incised rectangular tablets, some with one or two rounded ends, some with their sides indented, some animal-shaped, discs, bars etc (H-1158 to H-1357); and graffiti on pottery including potter's bats and chuck molds, on stone and on miscellaneous objects including an animal figurine, an ivory stick and copper axes (H-1360 to H-1520). It is interesting that the range of inscribed objects varies considerably from one site to the other.

From page 207 the material from the recent excavations is illustrated by period, continuing the sequence of numbers from H-1521 to H-2583. Period 1 (Early Harappan: Ravi Phase) is represented only by graffiti on pottery (H-1526 is finely painted). In Period 2 (Early Harappan: Kot Diji Phase, p. 211 ff.) there are seals with geometric designs, seal impressions on tags, and many pottery sherds with graffiti. Mature Harappan seals appear in Period 3 (Harappan Phase, p. 231 ff.) and there are impressions on tags and on a circular tablet; bas-relief tablets made of faience, terracotta, steatite and even cast copper, are numerous: rectangular, round-ended, discs, miscellaneous shapes including fish, prisms, bars, and twisted strips bear a variety of impressions both of figural designs and of inscriptions and incised designs, and there are more than 300 incised sherds and terracottas. Period 4 (Late Harappan: Late Harappan Phase), by contrast, has only one seal with a geometric design, and two sherds with graffiti, and share page 360 with Period 4/5 (Late Harappan or Cemetery H Phase) with two sherds with slashed rims. The Harappan catalogue concludes, as did the Mohenjo-daro catalogue, with Addenda and Corrigenda in numerical order (pp. 361–363), and three fragmentary seals and one stamped bulla are illustrated in colour on page 365.

Pages 413 to 443 at the end of the volume provide explanatory information, a list of abbreviations, and the data relating to the objects, arranged by page number. Unfortunately these data are minimal and I had, for instance, hoped for at least some information on the strange gold objects M-2128 to M-2129 resembling small fountain-pens with incised signs, and described only as “Gold Objects” on pp. 130–131. In the entry on p. 422 there was a reference to the publication of M-2129 in a Japanese exhibition catalogue, and fortunately (having acted as courier for the British Museum at the end of the exhibition) I had the catalogue, and it informed me that the objects were necklace pendants!

On pages 366 to 412 there are colour photographs, hugely enlarged, of some of the more spectacular seals from both sites, including some of the new Harappan finds (any number above H-1520). The seals generally show various animals; there are also rectangular tablets, mostly impressed with an inscription, but also with figurative scenes in relief. Pages 391, 394 and 395 show copper plates with incised or etched designs: one with an Indian elephant and inscription, and two showing archers with small bows, wearing horned headdresses. It has been suggested that the latter indicate contact with the Akkadians of southern Mesopotamia from c. 2330–2200 bce. Indeed, R.M. Boehmer (“Das Auftreten des Wasserbüffel in Mesopotamien in historisches Zeit und seine sumerische Bezeichnung”, Zeitschrift für Assyriologie 64 (1974), pp. 1–19) has argued that water-buffalos may have been imported into Mesopotamia at this period, together with the way of depicting them with their horns seen from above, as on p. 374 (see Collon in Reade, op. cit., p. 215, Figs 11a and 11c).

Seal M-353 at the top of p. 411 is an interesting piece because although the square shape is that of an Indus seal, the white stone, style of carving and design would suggest a Mesopotamian origin, perhaps in the Diyala region north-east of Baghdad, precisely where a contact with the Indus Valley has been demonstrated by the discovery of a cylinder seal with clearly Indus decoration (see H. Frankfort, Stratified Cylinder Seals from the Diyala Region (OIP LXXII, Chicago 1955, nos 76 and 642).

Several seals from Period 3 of the new Harappa excavations are said to be rectangular tablets, but the photographs clearly show a domed surface, suggesting that they are either cylinder seals or half cylinders. I was particularly interested in H-1970 (p. 269) and H-1971 (pp. 270 and 396) which have a longitudinal domed surface bearing a contest scene of Mesopotamian type with a hero between two animals - in this case non-Mesopotamian tigers! However, the base of both objects is clearly flat and was impressed with a seemingly unrelated scene: a matador attacking a bull from the front and about to spear it. Both subjects occur at Mohenjo-daro, and M-306 and M-492 are conveniently also illustrated on p. 396 for comparison.

Stamp seal H-166 with its deckled edges (p. 412), is paralleled by a seal from Gonur Tepe in Central Asia and, less closely, by seals from Mari in Syria, Sharh-i Sokhta in Iran and the Murgab in Central Asia (Collon in Reade, op. cit., pp. 218–9, Fig. 21a-b and 27a-c). Another interesting feature of Indus seals are the drilled centre-dot circles on M-2093 and M-2094 (p. 122) and on numerous examples from the new excavations at Harappa: from Period 2 (p. 211), and from Period 3 on pp. 262, 266, 280, 281 and 295. This feature also decorated the backs of the so-called “Dilmun” seals from the Persian Gulf (H. Crawford: Early Dilmun Seals from Saar, Ludlow 2001, p. 17, Figs 10–11) and features in many of the seal designs at Saar and Failaka - particularly the eyes of animals (e.g. Crawford, op. cit., pp. 70–73; P. Kjaerum: Failaka/Dilmun 1:1. The Stamp and Cylinder Seals, Aarhus 1983, p. 15 nos. 1–3 and passim). We know there was trade between the Gulf and the Indus in the late third millennium bce and early second, but where the centre-dot circle originated is not clear. This feature was only used much later in Mesopotamia, in the Late Bronze Age glyptic of the Mitannian period in the north. Interestingly enough, H-1703 on p. 237 looks like a copy of a Dilmun seal without the decoration, but the quadripartite division on its base also exists on a seal from Failaka where the design is cruder (Kjaerum, op. cit., p. 133, no. 324).

This third volume of the CISI is therefore full of potential for those examining contacts between the Indus and the Gulf, and beyond. The use of a similar and distinctive technique and motif - namely the centre-dot circle - also has important implications for those working on relative chronologies between the two areas. All those who put the volume together, and particularly Asko Parpola, the editor, are to be congratulated and thanked.