1. Those conversant with the archaeological reports will, however, appreciate the fact that other than a few major site-areas, or individual sites, there was not a great deal of “choice” involved in Chinese field archaeology during the period 1950-1965. A glance through the data (in Chinese) in the first columns in the Table of Sites and Remains will show that the major bulk of archaeological work was concerned with the processing of chance finds which were occurring at such a rate that little time and few resources were available to indulge in the luxury of planned large-scale digs. Since Nixon's visit, however, “choice” has become an increasingly characteristic element in the reports.

2. Again consultation of our Table of Sites and Remains (last column) will show frequent reference to the fact that many burials have been found in a disturbed state—effected by grave robbers either in the past or more recently. We are, in fact, very much dependent upon reports covering numerous plundered sites or burials which, seldom being completely ransacked, often have preserved information and artifacts of considerable importance and value. In the Table it was not found possible to report every such recorded instance of plundered tombs, there being far too many to cope with (especially throughout many of the more extensively excavated site areas) in the meargre printing space available. It is for this reason (amongst others) we have repeatedly warned users of the data that the original reports must be consulted along with any research conducted upon the basis of the Table (pp. vii, xxi, et al.).

3. It is further important to note that the distribution patterns (red) presented throughout our Distribution Maps, may at all times be observed in relation to all archaeological sites recorded in publication and entered in our Table since 1923 (these are repeated printed in blue on every map). I believe it will be found that this form of presentation, which has considerable advantages, has rarely been used elsewhere. The main value of showinq the background pattern of all the source sites (blue) in relation to the superimposed patterns of sites selected for specific reasons (red) is that it demonstrates the selected patterns in terms of the archaeological situation first ot all, and then secondarily the significance of the two sets of data in respect of the geography of the area. At all times those consulting the Distribution Maps are kept visually aware of the actual scope of properly attested source materials (see Introduction, Section III, “The Distribution Maps”). The costly two colour printing (blue and red) in our Distribution Maps has not been done for aesthetic reasons! This form of presentations is strictly functional—a cautionary means of data presentation. It may be claimed, I believe, to answer in advance most of the objections raised by Trousdale–providing that users of the data make themselves properly aware of the disciplinary approaches applied, or established, in Metallurgical Remains.

4. Earlier he writes: “That the Distribution Maps bear an as yet unknown relationship to actual production and dispersal of metal artifacts should, I believe? be acknowledged by both the compilers and the users.” The lack of acknowledgement alleged here stems from too cursorily a perusal of our prelims and too hurried a reading of Keightley's review article (I assume “users” is a reference to the latter?) If users of our data provide cautionary observations such as: “The map, in short, and the reports upon which they are based, reveal…” “A study of the individual reports is needed?…” etc., no one surely can take umbrage as to the user's approach to the data. Consultation of Trousdale's own presentation of archaeological data concerning the scabbard slide (his Figs. 2, 7, and 13) and his discussion around it: “On the basis of available reports to date.…” “Though found in only three localities the broad distribution of these sites indicates…” (p. 12), etc., demonstrates no more and no less the same careful appreciation of the limitations of the data as may be noted in Keightley's use of comparable data. As to the situation in Metallurgical Remains, amongst our various comparable cautionary statements, I cite only the following: “The sites investigated (following accidental finds) and those excavated under conditions of control are distributed widely and are located through out most of the crucial areas from which archaeological information must be obtained if the resultant distribution patterns and interpretations based on them are to be regarded as reasonably reliable. At present it may be justifiably claimed that the available data is sufficient to allow the establishment of a working hypothesis firm in its support of… This hypothesis based as it is on practically all the available properly provenanced information … provides a substantial argument in favour of…” (p. xxiv). As to the manner of graphic presentation, see my preceding note.

5. It is difficult to refrain from commenting here upon the inadequacy of traditional text data (i.e., Tso-chuan, Kuo-yü, Sun-tzu, etc.) for information leading to accurate “historical judgements” that might tell us more about “the type of swords” available than might the swords themselves in the archaeological context! More seriously, however, one must not lose sight of the fact that the traditional texts are strongly flavoured with Han period lore as a result of their processing by the Confucian literati (see my survey “Records of Discoveries of Bronze Vessels in Literary Sources—and Some Pertinent Remarks on Aspects of Chinese Historiography”, Journal of the Institute of Chinese Studies, Chinese University of Hong Kong VI. 2 [1973]: 455–546Google Scholar, for relevant examples of this—note especially misconceptions of the use of iron as recorded in the traditional literature). As to the numbers of swords made throughout the Ch'un-ch'iu and Chan-kuo periods, the researcher will find himself on very tenuous ground if he attempts to make “historical judgements” upon the basis of the traditional literature—the best one may hope to do is simply to make “intelligent guesses” whose validity cannot, however, always be confirmed or denied objectively. The archaeologist is, of course, well aware that each artifact he unearths represents an unknown fraction of the number that were originally made. However, as he (or the armchair archaeological-historian) work through the typological analyses of weapons listed in report after report (e.g., swords and daggers in the Yünnan Chin-ninq Shih-chai-shan ku-mu-chün fa-chüeh pao-kuo the Ssu-ch'uan ch'uan-kuan-tsang fa chüeh pao-kao, and many such others), he will eventually gain insight into such problems as the significance of the typology of artifacts, their incidence in relation to other artifacts, aspects of their function, materials and methods of manufacture, temporal and geographical inter-relationships, etc., as the programme of research proceeds extensively throughout the archaeological literature. And so will he gradually gain an understanding of the artifact in its original social/economic setting. Information on so many crucial facets of an artifact is often non-existent in the traditional literature, or when it appears it is usually inadequate or misleading (if not entirely erroneous).

6. In his Handbook to the Collection of Early Chinese Antiquities (1963) Watson actuaUy states: “Iron swords up to three feet in length first appear in tombs of the fourth century. They seem to have been specially favoured by the western and southern states of Ch'in and Ch'u, whose large armies of infantry were engaged in frequent civil war of the last two centuries preceding the Ch'in unification of 221 B.C.” (p. 60). That Watson's meaning herein tenns of “archaeological record”—is not as interpreted by Trousdale, may be demonstrated in his China Before the Han Dynasty (1961) on the very page that Trousdale has referred to in his preceding note (243): “Archaeology can yet give little support to the theory that the possession of the iron sword was the decisive factor in the success of the Ch'in conquest…” (p. 146)!

7. In this connection Keightley (p. 32) draws attention to an anomaly attending my remark (p. 60) “to the effect of wrought-iron technology ‘when applied to weapons was profound and appearantly was no mean factor in the rise of Ch'in and Ch'u and the ultimate domination of the contending States by the former’” The anomaly results from my attempting to adapt and up-date an earlier writing for the later publication. This essay on origins of bronze casting was one of several chapters prepared (cirea 1962) for a Catalogue of bronze vessels which never materialized. The point completely escaped my attention and I am glad to have it corrected. Also, there was a slip regarding MRAC Site N0.12:2-2—the Table is correct (i.e., no iron sword); the distribution dot on the maps (Map 6 and Map 6c) should be deleted. Both my coauthor and I would appreciate hearing from users of any other slips of this kind that may turn up with active use of the Table and Maps.

8. As we may expect more data to be forthcoming in the next few months, it would seem better to wait a while before attempting to make a thorough appraisal of currently available reports. However, one item of information is particularly interesting to record here, the analyses of a sword and of an arrow-head (WW 1978.5:22Google ScholarPubMed). The former is a tin/lead (? aluminum, a mistaken reading of lead? bronze: 21% tin, 2% lead (?) and the remainder of the alloy, copper. As such it matches earlier analyses of attested swords from Ch'ang-sha and Shih-chai-shan as well as most unprovenanced items (see Bronze Casting, Table Five). The presence of Aluminum (2.18% in the sword and 7.71% in the arrow-head) is very much open to question.

9. Two recently published surveys on Chan-kuo and Han iron manufacture (KKHP 1975.2 and 1978.1) confirm the practice of casting in “malleable cast iron” by a co-fusion process at the melting stage, and decarburization of iron castings to form a kind of steel. It would be worthwhile if someone well conversant vdth iron could now bring together the presently available data and prepare a survey on Han period iron casting (and wrought iron) technology. Such a study would provide a most useful basis to extrapolate and review the possible significance of the archaeological information we now have on the pre-Han iron Industry.