Gavin Menzies' cartographic fiction: the case of the Chinese 'discovery' of Australia.

This article critically examines the evidence and claims made by Gavin Menzies of Ming Dynasty Chinese circumnavigation of the world in 1421: the year China discovered the World. It concludes that the evidence presented remains highly speculative and is not sufficient to justify the conclusions Menzies draws.

**********

Gavin Menzies, a retired submarine captain in the Royal Navy, launched his book 1421: The Year China Discovered the World in 2002 with a lecture in an inspired locale: hired rooms in the Royal Geographical Society's building in London. (1) He maintains that it is primarily his training and experience in seamanship, cartography and astro-navigation that has enabled him to make discoveries that have escaped 'distinguished academics in the field' (p. 9). Regrettably, the book is but one more example of what has been termed 'imaginography'. (2) The measure of credence that should be accorded to Menzies' book as a whole may perhaps be judged by an examination of his claim that the Chinese discovered Australia.

He states that he has traced voyages made between 1421 and 1423 by fleets of Chinese junks to both coasts of the Americas, to the Arctic, Antarctica and Australia. The Chinese government's sudden adoption of an isolationist policy after their departure, led to the official destruction of all records of those vessels that made it home, so there is 'virtually no evidence to show where they had sailed or what discoveries they had made' (p. 81).

Menzies' thesis is based on two major, but very dubious, assumptions. One is that the locations of a number of supposedly Chinese artefacts, plants, DNA traces, etc. worldwide are evidence that the Chinese took them there in the 1420s, and that a number of unidentified wrecked ships are Chinese ones dating from the same period. The other major assumption is that his often highly speculative identifications of specific features on certain early maps are all correct. He even bases wild hypotheses on the reported contents of maps of notoriously dubious authenticity and date that have not survived. Then, bearing in mind that Chinese junks were more or less restricted to sailing before the wind, he charts routes between his artefact and shipwreck sites, and his speculatively identified features, in accordance with prevailing wind and current patterns. However accurate his routes between his various landmarks may be, his assertion that Chinese junks sailed along them is a completely unproven hypothesis.

Works of historical conjecture such as his rely for their success upon a number of persuasive techniques. These include the use of that false logic called begging the question or circular argumentation, the all-pervasive and persuasive presentation of assertions as though they are established facts, the selective use of evidence, and frequent repetition. Menzies' work is permeated by the above techniques. He relies extensively upon secondary and even tertiary sources, mainly works of the same speculative genre as his own, and upon any unsubstantiated rumour which can conceivably be interpreted as supporting his case.

It may well be Menzies' naval training that has conditioned him to make the fundamental error of presuming that coastal outlines on early maps and charts had always been surveyed. Portolano charts were the result of the accumulated practical experience of generations of mariners; other maps and charts, especially those of unknown or little known parts of the world were either entirely hypothetical, or were largely imaginative, graphic representations of travellers' tales. Even some portolano charts had mythical information inserted on them, especially in the Atlantic Ocean. It is impossible to determine with any certainty from coastal outlines alone what a cartographer was depicting, or thought he was depicting. Only the complementary evidence of inscriptions can confirm or refute such superficial, visual identifications, and only researchers who are careful, experienced palaeographers and historical linguists can reasonably reliably read many early inscriptions, let alone correctly discern their meaning. Some will almost certainly never be demystified.

Menzies' hypothesis regarding the Chinese charting of the coastline of Australia in the 1420s involves his steering two Chinese admirals there, Hong Bao from the west, and Zhou Man from the east.

Unable to find any Chinese charts to prove his case, he attempts to prove it by searching for 'a chart of the continent that had been surveyed and drawn before the first Europeans discovered it' (my italics) (p. 151). He claims to have found what he was looking for in the manuscript Boke of Idrography made by the Franco-Scottish cartographer, Jean Rotz (1542).3 He was one of the mid-16th century cartographers of the Dieppe School who have become well known, especially over the last half-century or so, because of claims that the large landmass of Jave-la-Grande on them, immediately south of Indonesia and partially attached to it, is really a misplaced depiction of Australia, supposedly discovered by the Portuguese in the 1520s. This interpretation involves the assumption that Java was confused with Arnhem Land and Sumbawa with Cape York Peninsula (Fig 1). (4) The name Jave-la-Grande is a French translation of the Portuguese JAVA MAIOR (Greater Java) by which they meant Java, following Marco Polo's use of Java Minor to refer to Sumatra. This apparently illogical usage caused much confusion amongst cartographers, including those of the Dieppe School, and still affects many writers to this day.

[FIGURE 1 OMITTED]

Menzies bases his claim on two of Rotz's charts. The first is an eastern hemisphere one which depicts southern and south-eastern Asia, part of Indonesia and the 'continent' of Jave-la-Grande which is composed of two parts which Rotz calls The Lytil Jaua and The Londe of Jaua (Fig. 2). The second chart is a larger-scaled one of South-East Asia, Indonesia, and the northern extremity of The Londe of Jaua (Fig. 3).

[FIGURES 2-3 OMITTED]

Menzies quite correctly accepts that Rotz and the other Dieppe cartographers copied material from Portuguese charts. However, he maintains that the Portuguese had actually copied them from Chinese originals. He claims, for example, that Rotz's portrayal of South-East Asia, as far north as Hong Kong, is charted 'with such compelling accuracy' that it could not have been done by the Portuguese, because by 1542, when Rotz's charts were made, they could not possibly have acquired such an 'intimate knowledge of the coastlines of the Indian Ocean, China and Indochina' (p. 152). This completely unsubstantiated, easily refuted assumption demonstrates reprehensible ignorance of the documentary evidence of Portuguese activities in Asia in the first forty years of the 16th century. The Portuguese did not rely exclusively on their own observations, but took the obvious, commonsense steps of obtaining local charts and sailing directions, and employing local pilots. As early as 1512, Afonso de Albuquerque wrote enthusiastically to King Manuel about one Javanese world chart which actually marked in the regular routes of Chinese vessels; he stated that it was the best chart he had ever seen. The pilot Francisco Rodrigues made a tracing of part of it. (5) The rapidity with which the Portuguese added to and refined their knowledge is illustrated in Fig. 4.

[FIGURE 4 OMITTED]

However, based on the above utterly false premiss, Menzies goes on to imply that the coastal outlines of Jave-la-Grande (Lytil Jaua and The Londe of Jaua) were also copied by the Portuguese from information on charts which have not survived, but which he claims were made by Hong Bao and Zhou Man in the 1420s.

Ever since 1786, when Alexander Dalrymple first suggested that the east coast of this enigmatic landmass was that of Australia, it has been claimed that this 'continent' of Jave-la-Grande is a copy of a 1520s Portuguese map of Australia, and that the French inserted it incorrectly on a map of the world (Figs 1 and 2). (6) Yet, despite the ability of some to 'see' Jave-la-Grande as Australia, and Menzies' persistent comments on the accuracy of its depiction on the Dieppe maps, the apparent similarities between their coastal outlines are extremely superficial, for most people being restricted to the northern half of the two east coasts; there is no south coast. There is not a single surviving Portuguese map or chart which contains any coastal outlines obviously resembling those of Australia until about 1630, when details from the first Dutch charts of parts of it leaked out. There are no surviving Portuguese charts really resembling those of Jave-la-Grande's east and west coasts, either in that position, or anywhere else. Nor is there any other surviving evidence that the Portuguese reached Australia before the well-documented arrival of the Dutch in 1606. Some have stated that the absence of any evidence is due either to the Portuguese politica de sigilo (policy of secrecy), or to the destruction of records in the 1755 Lisbon earthquake. These are two decidedly overworked excuses used by proponents of improbable theories to try to explain the absence of documentary evidence that they would like to exist.

It has been claimed that the French must somehow have laid hands on a map of Australia, either by nefarious means in Lisbon, or perhaps by capture at sea before it reached Portugal. If Jave-la-Grande is Australia, then the French, or if you believe Menzies, his very accurate Chinese, misplaced it by some 24o of longitude and some 5o of latitude.

However, Menzies is unworried by this 'misplacement', even though he claimed, in a radio interview in Australia, that the Chinese could measure latitude and longitude to within about 30 miles. He unquestioningly accepts Rotz's Lytil Jaua and Londe of Jaua combined (Jave-la-Grande) as being Australia, and provides supposedly complementary evidence by recycling sensational material derived from the popular, but notoriously unreliable, speculative works of Kenneth McIntyre, Eric Whitehouse, Rex Gilroy and Wei Chuh-Hsien, practically all of which has been disproved or thoroughly discredited by academic research over the last twenty years or so. (7) Such hoary old chestnuts as the 'Mahogany Ship', the Bittangabee 'fort', and the Gympie pyramids play a significant role in Menzies' supplementary 'evidence'. The 'Mahogany Ship', which has not been seen since the 19th century and has never been scientifically examined, Menzies identifies as being Chinese; others have variously identified it as being Portuguese, Spanish, Dutch, and as an early, Australian-built whaler! Menzies' 'Toscanelli' map of 'Australia' (p. 416), inherited from Whitehouse, with unnamed rivers claimed to be the Murray, Darling, Cooper, Diamantina and Flinders, is actually a vast enlargement of an enigmatic feature on the famous 'Genoese' world map of 1457. There is not one iota of evidence to support Menzies' claim that 'Captain Cook ... used some of the maps of the Dieppe School' (p. 188).

Reverting to the Dieppe school's portrayal of 'Australia', it is astonishing that none of those who have accepted the identification of Jave-la-Grande as Australia before Gavin Menzies have considered the inscriptions on the Dieppe maps' Jave-la-Grande of much significance. Kenneth McIntyre considered it 'difficult for an English-speaking person (whether author or reader) to appreciate the nuances of spelling, vocabulary and usage involved in the analysis' of the inscriptions, so virtually disregarded them. (8) Failure to examine crucial evidence because it is difficult is inexcusable. So far as the place-names on Java (McIntyre's and Menzies' 'Arnhem Land') are concerned, the former specifically states that 'These north-coast names ... do not belong to Australian historical geography, and do not concern us here'. (9) This is an astonishing remark for someone with a preconceived conviction that Jave-la-Grande is Australia.

Unlike McIntyre, Menzies does pay attention to place-name inscriptions on the north coast of his 'Arnhem Land' (Rotz's Lytil Jaua), on his south-east Asian chart (Fig. 3). He states that 'A number of descriptions in medieval Portuguese were written on Rotz's more detailed chart. The names are easy to translate and all of them correspond to what is found there [i.e. in Arnhem Land] today' (pp. 188-9) (my italics). Figure 3 shows that there are actually twelve inscriptions on that coast. Menzies only mentions seven of them. Only the words canal (channel or strait), de (of), aguada (water supply source), a (the), fin for fim (end), and a misspelt version of the Portuguese surname Albim or Alvim, with d, the abridged form of de prefixed to it, are Portuguese words; there is nothing medieval about them. Eleven of the twelve inscriptions are incontrovertibly Javanese place-names, while the remaining one indicates the historically recorded site on Java's north coast where Joao Lopes de Albim obtained water in 1512.

Menzies' transcriptions are careless. For his 'translations' of Javanese place-name inscriptions as though they were Portuguese he must have used a 'fictionary' (pp. 188-190). He somehow manages to 'translate' Canal de Sonda, Rotz's canal de onda (Sunda Strait, between Java and Sumatra) as 'narrow sea ford'; Aguada dillim, Rotz's aguada dolbim (water supply source of Albim ) as 'waterway leading to inland sea'; chumbao, Rotz's chbao (Tjeribon or Cheribon, now spelled Cirebon), as 'lead', as though it were the Portuguese word chumbo; Agarsim, Rotz's agarim (the Javanese city now spelled Gresik, preceded by a, apparently the Portuguese feminine definite article) as 'yes indeed water is here', presumably assuming that the first two syllables are a misspelling of agua (water), followed by the Portuguese sim (yes).

He utterly confuses the next two names eastwards. They actually read iru baia (Surabaya) and what is either limgrania or lungrania. He has either misread iru baia as lingrania, and translated it as 'lime trees', or has omitted it altogether. If, for whatever reason, he has omitted iru baia, then he has given two different readings for Rotz's limgrania, the second being Nungrania, which he 'translates' as 'no farmland'.

He does actually translate correctly his finjava mistranscription of Rotz's fin de jaua, at the eastern end of Rotz's Lytil Jaua, but comments that 'The Gova [sic, i.e. Gove] Peninsula, the eastern tip of Arnhem Land, is finjava, or 'the end of Java''(p. 189). His presentation of Gova (sic) as a variant reading for Jaua (Java) caps his toponymic monstrosities. The name Gove first appeared during the Second World War when the airfield on the peninsula was so named in memory of Pilot Officer William J. Gove, who was killed in 1943!10 Menzies' visual conception of Java as being Arnhem Land hardly accords with his repeated assertion of the accuracy of 15th century Chinese cartographers.

The five inscriptions on Rotz's Lytil Jaua which Menzies fails to mention or 'translate' are, from west to east: palinbam (Palimbam, one of the 165 localities wiped out in the eruption of Krakatoa in 1883), unda (Sunda, the western third of Java), Jaue (a Gallicized spelling of Java), japara (Japara), tuba (Tuban) and iru baia (Surabaya). The islands just off the north coast, which Menzies also fails to mention, are crimagaua (Pulau Karimanjawa), amadura (Madura) and pude (Pulau Sapudi).

Despite the fact that Menzies clearly implies that Rotz's Sinbaua (Sumbawa) is really Cape York Peninsula, he fails to mention the seven place-name inscriptions on it. Could he not 'translate' them? He doesn't mention either the two islands in his 'Gulf of Carpentaria', the space between Rotz's Lytil Jaua and his Sinbaua (Sumbawa), namely bacha (Bali), and anjane (Lombok's outstanding landmark, the volcano Rinjani). It was these two names that George Collingridge misread, from the spellings given on the Harleian map (see its outline superimposed in Fig. 1). Apparently believing them to be Portuguese words, he 'translated' them as a syntactically impossible Portuguese sentence supposedly meaning 'no boats go here', claiming this to be indicative of the shallow Gulf of Carpentaria !11 On his 'authority', these words have been quoted in good faith by generations of historians as being 'pure' or 'straight' Portuguese. Menzies may well have found his particular selection of Rotz's inscriptions "easy to translate", but it is to be hoped that his Englished versions will not mislead readers for as long as Collingridge's ones did--nearly eighty years.

How does Menzies deal with the inscriptions on the east and west coasts of his 'Australia'? As there is not a single one on Rotz's Londe of Jaua, Menzies provides Australian place-names wherever his imagination suggests on its two enigmatic coastlines. He admits that the other, larger-scale Dieppe charts of 'Australia' (Jave- la-Grande) such as the Harleian, and those made by Pierre Desceliers, Nicolas Desliens, and Guillaume Le Testu give 'an almost identical depiction of the continent' (p. 181), but he does not even mention the fact that, on these maps, both those coasts, and the islands off them, are identified by place-names. Why does he ignore them? Did it not occur to him that they might have had inscriptions on them that could have provided irrefutable evidence that they referred to Australia? After all, he examines some of the names on his 'Arnhem Land' in an attempt to do so!

He is clearly familiar with nearly all the sensational, speculative matter on the hypothetical pre-Dutch discovery of Australia published in the last twenty-five years or so. However, he does not mention a single one of the articles published in scholarly journals and conference papers over roughly the same period which have minutely examined that place-name evidence. (12) Menzies' apparent unawareness of this material is further evidence of his reprehensively defective standard of research. Jave-la-Grande (Rotz's Lytil Jaua and Londe of Jaua) is not a misplaced representation of Australia at all. A forthcoming book will gather, update and expand this published evidence, but a summary of its findings is given below, and relevant references are provided in the notes.

If the coastlines of the enigmatic 'continent' are not those of Australia, what are they and why did the French place them where they did? What did they make of the place-name inscriptions? The handwriting may well not have been easy to read, and it was written by Portuguese, most of the inscriptions being Portuguese attempts to render phonetically the exotic sounding names of places which the French did not recognize, and possibly did not then even know about. There is plentiful evidence to show that many cartographers, and not only French ones, often resorted to guesswork to 'make sense' of inscriptions which they could not 'understand'.

The north coast of Menzies' 'Australia' is precisely what the inscriptions unambiguously say it is, the north coasts of Java and Sumbawa. The east and west coasts, and the small islands off them, are also what the placename inscriptions say they are, but as they stand, they are what the puzzled French made of the Portuguese originals; consequently most require demystification; there are considerable variations in spelling from map to map.

The west coast is composed of two sections. The more northerly one is a misplaced and misorientated, token, duplicate, aide memoire version of the Java coast of Sunda Strait, which appeared very early in the 16th century on Portuguese charts, and remained there on practically all their depictions of Java throughout it. (13) It was even copied, in detail, on a Dutch chart of 1598 by Willem Lodewycksz. (14) The place names on that coastline on Portuguese charts and on the Lodewycksz one clearly identify it, though the French did not quite know what to make of them.

The other section is south of the manifestly hypothetical strait or river, which is unnamed by Rotz, but is called R[iviere]: Grande on other Dieppe maps, and is claimed by Menzies (p. 188), McIntyre and others to represent the Victoria and Roper rivers in Australia's Northern Territory. It is a misorientated copy of a very early Portuguese sketch chart of part of the south-west coast of Java on a different scale from that of the main map. (15)

The east coast is copied from a very early, differently scaled, Portuguese sketch chart of the coast of Vietnam, evidently lacking any indication of latitude. It probably dates from as early as 1516, when the first official Portuguese voyage towards China, under the command of Fernao Peres de Andrade, only just reached the Gulf of Tonkin before the monsoon compelled him to turn back to Malacca. (16) The large triangular east coast projection, which has the decidedly strange name C: de fremoe (lit: Cape of beautiful) at its tip on other Dieppe maps, has presented those who believe in the Australian identification with a serious anomaly to explain away, for there is no such feature on Australia's east coast. It is the Mekong delta. The French version is the result of a misreading of the Portuguese cabo de camboja (Cape of Cambodia), now known as Mui Ca Mau. (17) Various writers have hazarded guesses as to its identity; they have suggested Cape Howe, on the south-east coast of Australia, both the southern and the north-eastern tips of Tasmania, and the East Cape on New Zealand's North Island! (18) Menzies' 'identification' is even more imaginative (pp. 169-172). One side of it is declared to be 'the southern curved part of Tasmania'. The two rivers that are shown flowing into the sea on its other side are claimed to be those existing on the minute Auckland and Campbell Islands far south of New Zealand, while both the land between them, and that connecting them to 'Tasmania' is solid ice! Menzies implies that the islands just north of his 'Tasmanian' coast are Norfolk Island (pp. 164 & 172). They too have been variously identified by others. They are actually the Con Son Islands just off the Mekong delta. (19) Menzies' 'Chinese' chart of Australia is not quite as accurate as he repeatedly makes out!

Having revealed the real identity and composition of the Dieppe cartographers' enigmatic apparent continent, it remains to explain why they placed its three differently scaled constituent elements where they did. The basic world map, mainly of Portuguese origin, would probably have depicted Java and Sumbawa with their south coasts left blank, as they usually were on 16th century charts. In view of the name Jave-la-Grande, it seems probable that the Portuguese original sketch chart of the western half of Java's south coast had on it the one inscription they would have almost undoubtedly recognized, Jaua maior, thus explaining why they attached it to the south-western extremity of Java. It also seems probable that they misunderstood the name, thinking that it applied to a landmass larger than what we know as Java, for Rotz calls Java itself Lytil Jaua, and at least one other Dieppe map calls it PETITE JAVE. In view of the widespread belief in the existence of a large unknown southern continent, the attachment of the other unidentified coast, actually Vietnam, to the south-east coast of Sumbawa would have seemed logical. Some of the Dieppe maps did actually join the east and west coasts of Jave-la-Grande, by a manifestly hypothetical coastline, to the north coast of Tierra del Fuego which had been discovered, not by the Chinese, as Menzies claims, but by Magellan in 1520. It is strange that none of the Dieppe cartographers seem to have been aware of the route home of Magellan's surviving ship, the Vitoria, for its course south of Java is blocked by Jave-la-Grande. McIntyre suggests that the R: Grande was inserted to 'let the Vitoria through', which contradicts his, and Menzies' contention (p. 188) that it represents the Victoria and Roper rivers; the former was certainly not named after it! (20)

Jonathan Swift, commenting on the literal acceptance of illustrations included on maps by many cartographers, commented:

   So geographers, in Afric-maps,
   With savage-pictures fill their gaps;
   And o'er uninhabitable downs
   Place elephants for want of towns. (21)

Menzies, apparently unacquainted with this famous verse, appears to believe that he may have found evidence that the Chinese brought horses to Australia (p. 183). It consists of the depiction of four or five of them on one of the Vallard atlas's maps of 'Australia'. There are also, however, camels, houses on stilts and an obviously oriental procession! It is well known that the Vallard illustration derives from the Parmentier brothers' expedition to Sumatra in 1529.22 Menzies even seems to believe that one of Eric Whitehouse's fictitious Australias could confirm a story that Chinese cavalry rode across Australia from the site of the elusive 'Mahogany Ship' near Warrnambool to where Cooktown now stands (p. 154)! Would any academic researcher present such uninformed, uncorroborated speculation as evidence?

Menzies assumes that the enigmatic landmass south of Indonesia is Australia, and repeatedly says that it is. He never proves it, but time and again uses Australian place-names to identify unnamed features, on unidentified coasts on Rotz's hemisphere chart, as if he had confirmed it. One cannot condone either Menzies' incredible 'translations' of Javanese place-names, or his omission of the other Dieppe maps' place-name evidence which torpedoes his thesis, at least so far as Australia is concerned. A variously-scaled, ignorantly-assembled, misplaced, composite chart of Java, Sumbawa and Vietnam is hardly a reliable one for Menzies, or his Chinese, or anyone else, to steer by when sailing in Australian waters! Those who seek to rewrite history on the evidence of early maps and charts should not so readily ignore the research of relevant academics, historians of cartography, historical linguists, toponymists and palaeographers. They are aware of hidden 'rocks' and 'reefs' which radar, sonar and astro-navigation can never detect.

Menzies' apparent ignorance of such hazards 'down under' does not inspire confidence in the reliability of the conjectural 'evidence' he puts forward about other parts of the world. Doubtless others will comment in detail on his treatment of areas within their own particular areas of expertise. One may wonder, for example, about the validity of his claims that Chinese vessels passed through the Straits of Magellan on the 'evidence' of the Piri Reis map of 1512, or that they circumnavigated Greenland, primarily on the 'evidence' of the controversial Vinland map. One may question his adjustments of the scale and orientation of enigmatic islands in the western Atlantic (Caribbean?) on Zuane Pizzigano's map of 1424 to make them more closely conform to his hypothetical identifications (pp. 252). As for the inscriptions on those islands, they have defeated interpretation by all experts, but not by Menzies. It will be no surprise, however, to learn that his 'translations' of them, after dipping into dictionaries in the British Library (pp. 363-367), are every bit as fanciful as his Javanese 'translations'. He identifies one of those islands, named satanazes, as Guadeloupe, by rendering two names he finds on it, con and ymana, as 'a volcano erupts there' (p. 247), apparently assuming that those two words are cognate with English 'cone' and 'emanates', and then unjustifiably stretching those meanings. (23) He goes on to build on that 'translation' by examining volcanic eruption dates, finally claiming that from this information he 'had first-hand evidence that a cartographer had been in the Caribbean no later than 1424, sixty-eight years before Columbus' (p. 247).

'Imaginography' and uninformed, wildly speculative 'translations' of toponyms are not conducive to a credible rewriting of history.

ENDNOTES AND REFERENCES

(1.) Menzies, Gavin, 1421: The Year China Discovered the World, London, Transworld Publishers, 2002. For the American market it was published as 1421: The Year China Discovered America, New York, William Morrow, 2003. The texts appear to be identical. A TV documentary is promised.

(2.) Richardson, W.A.R., '"Imaginography': Sensational Pseudo-Discoveries', The Skeptic, vol. 19, no. 1 (Autumn 1999), pp. 26-30; reprinted in IMCoS Journal, 79 (Winter 1999), pp. 5-14; for reasons which will become apparent, the present article inevitably covers some of the material dealt with in 'Imaginography'. For another case of 'imaginography', see Richardson, W.A.R., 'South America on Maps before Columbus? Martellus's 'Dragon's Tail' Peninsula', Imago Mundi, 55 (2003), pp. 25-37.

(3.) Rotz, Jean, Boke of Idrography (1542), British Library, Royal MS 20.E.9; see fols. 30ro and 9vo.

(4.) The literature on the Dieppe maps is very extensive, but see especially Wallis, Helen (ed.), The Maps and Texts of the Boke of Idrography Presented by Jean Rotz to Henry VIII, Oxford, 1981, which Menzies lists in his bibliography but never cites, even though it reproduces all of Rotz's charts along with a very scholarly study. It includes a list of all the Dieppe maps on p. 39. See also Coote, C. H. (ed.), Autotype Facsimiles of Three Mappemondes ..., Aberdeen, 1898, which reproduces in detail the anonymous Harleian or Dauphin map (British Library, ADD. MS 5413) and Pierre Desceliers' maps of 1546 and 1550.

(5.) Menzies seems unaware of two of the major works covering the subject: Cortesao, Armando and Teixeira da Mota, Avelino (eds), Portugaliae Monumenta Cartographica, Lisbon, 1960, 6 vols, which reproduces over 600 early Portuguese charts, including many of South-East Asia and Indonesia, and Cortesao, Armando, The Suma Oriental of Tome Pires and the Book of Francisco Rodrigues, London, Hakluyt Soc., 1944, (2nd series, vols 89 and 90), especially pp. lxxviii-lxxix; all the charts by Francisco Rodrigues made c. 1512 are reproduced in it (as well as in vol. I of PMC); one set of local sailing directions, Camynho da Chyna appears on pp. 301-2.

(6.) Dalrymple, Alexander, Memoir Concerning the Chagos and Adjacent Islands, London, 1786, p. 4.

(7.) (i) McIntyre, K.G., The Secret Discovery of Australia: Portuguese Ventures 250 Years before Captain Cook, Medindie, South Australia, and London, 1977; (ii) Whitehouse, Eric, B., The Northern Approaches: Australia on Old Maps, 820-1770, Mooroka, Queensland, 1994; (iii) Gilroy, Rex, Pyramids in the Pacific, Gympie, Queensland, 2000; (iv) Chuh-Hsien, Wei, The Chinese Discovery of Australia, Hong Kong, 1961.

(8.) McIntyre, op. cit., p. 108.

(9.) ibid., p. 113.

(10.) Appleton, Richard and Barbara, The Cambridge Dictionary of Australian Places, Cambridge, 1992, p. 121.

(11.) Collingridge, George, The Discovery of Australia, Sydney, 1895, p. 168.

(12.) Richardson, W.A.R., (i) The Portuguese Discovery of Australia: Fact or Fiction?, Canberra, National Library of Australia, 1989; (ii) 'Is Jave-la-Grande Australia? The linguistic evidence concerning the west coast', The Globe, 19 (1983), pp. 9-46; (iii) 'Jave-la-Grande: a place-name chart of its east coast', The Great Circle, vol. 6, no. 1 (April 1984), pp. 1-23; (iv) 'Jave-la-Grande: a case study of place-name corruption', 12th ICA Conference (Perth, WA), August 1984, Technical Papers, vol. 2, pp. 221-248; (v) 'Jave-la-Grande is not Australia', The Globe, 37 (1992), pp. I/13-I/16; (vi) 'A critique of Spanish and Portuguese claims to have discovered Australia', Investigator (Journal of the Geelong Historical Society), 120 (1995), pp. 83-107, and 121 (1995), pp. 131-147.

(13.) See maps including Java reproduced in PMC, especially in vols I and II.

(14.) The relevant part of the Lodewycksz chart of Sumatra and Java is reproduced in Richardson, W.A.R., 'The origin of place-names on maps', The Map Collector, 55 (Summer 1991), pp.18- 23; see p. 19.

(15.) Richardson (ii) passim, but also idem (i) pp. 16-19, and (iv) pp. 231-234.

(16.) See an account of the voyage in Joao de Barros, Decada Terceira da Asia, Lisbon, 1563. For reference to the 1777 edition see Richardson (iv), note 65.

(17.) Richardson (iii), Fig. 7, reproduces a text containing the words Reyno decamboja (sic) (Kingdom of Cambodia) in the handwriting of Tome Pires (c. 1512); it shows how easily, in a similar hand, decamboja could have been misread as fremoe.

(18.) McIntyre (op. cit.), FitzGerald, Lawrence, Java La Grande: The Portuguese Discovery of Australia, Hobart, 1984, and Herve , Roger, Decouverte fortuite de l.Australie et de la Nouvelle Zelande . . ., Paris, 1982. Between 1955 and 1982 Herve changed his mind from the southern tip of Tasmania identification to the New Zealand alternative.

(19.) Richardson (iii) passim, but also idem, (i) pp. 20-26 and (iv) pp. 234-241.

(20.) McIntyre (op. cit.), p. 197.

(21.) Swift, Jonathan, On Poetry, (1733).

(22.) The 'Vallard Atlas' is in the Huntington Library, San Marino, California (MS. HM. 29). For details, see Wallis (op. cit.), pp. 59-60, where the map is reproduced.

(23.) Irrespective of who may have helped Menzies with this 'translation', it is sheer speculation. The words con and ymana, as they stand, do not mean 'volcano' and 'erupts' in either Spanish, Portuguese or Italian, and there is no word present to mean 'there'. The basis of the 'translation' must be an assumption that con is a misspelling of Port. cone or Sp. & It. cono (cone), and that ymana is a variant spelling of Port., Sp., & It. emana (emanates). The only words that could justify such a translation are cognates of English 'volcano erupts', Sp. volcan erumpe, Port. vulcao faz erupcao, or It. vulcano erutta, perhaps with some minor early spelling variations.

W. A. R. (Bill) Richardson, Associate Professor (Spanish and Portuguese), School of Humanities, The Flinders University of South Australia.