Place-name evidence for eastern Sumatra in western Java la Grande.
Of the several theories of the origin of Java la Grande depicted in the Dieppe maps, the most cohesive is the one expressed in the 1980s by Helen Wallis, a former Map Librarian at the British Library and a recognised authority on the subject. (1) She held that Java la Grande is probably evidence of a Portuguese discovery of Australia of which information may have been conveyed to the cartographers of Dieppe by the survivors of the Parmentier expedition which visited western Sumatra in 1529-30. (2) Recent commentators hold that this theory is unlikely to be true (Pearson 2005, 19) or, at least, unlikely to be proven without "irrefutable evidence, either material or bibliographic" (Frost 2007, 423). Nevertheless this theory, whether true, untrue, or partly true, is still valuable for its ability to guide further inquiry. To follow Wallis is to find a stretch of the western coastline of Java la Grande where many of the place-names seem to apply directly, without much need either to assume textual error or to build bridging etymologies, to the eastern coast of Sumatra. (3) The same conclusion can be reached by comparing spatial data of Java la Grande, extracted by cartometry, with that of eastern Sumatra.
Java la Grande is the land mass depicted in mid sixteenth century world maps and atlases produced by the so-called Dieppe school of cartography. In its earliest fully-developed form, in the atlas of Jean Rotz (the Boke of Idrography dated 1542) (Fig. 1.), its mainland comprises two coastlines: the western coastline is a continuation of western Java, checked only by a narrow channel that isolates Java from Java la Grande, and extends southwards until about 35[degrees]S where it ends in mid ocean; the eastern coastline is contiguous with the eastern end of Sumbawa and extends south-eastwards almost to 60[degrees]S where it ends in the same manner. In later maps these coastlines are connected to a fictitious Antarctic continent, La Terre Australle, either conjecturally, or by realistically drawn coastlines. Twenty-two works are attributed to the school of which sixteen depict Java la Grande and, of these, twelve bear place-names for coastal features and nearby islands.
[FIGURE 1 OMITTED]
In Wallis' opinion, "The answer to the enigma [of Java la Grande] may be regarded as non-proven, but with the balance of evidence in favour of a Portuguese discovery of Australia" (Wallis 1988b, 35). Her opinion was grounded on several circumstances: Portugal was the only European power that was continuously active in Southeast Asia prior to the production of the Dieppe maps that show Java la Grande (they range in date from c.1540 to 1587); the Portuguese had the capability, opportunity and willingness to undertake maritime exploration in the region; the language of the inscriptions of the Dieppe maps is a mix of Portuguese, French and Gallicised Portuguese; the maps are the products of the Dieppe school of cartography; and the Parmentier expedition, which was sponsored by Jean Ango a ship-outfitter of Dieppe and, as mentioned above, visited the west coast of Sumatra in 1529-30. In addition, she noted (1988a, 50-51) that many of the depictions of Java la Grande are adorned with illustrations of Sumatran scenes attributable to the Parmentier expedition. She was especially impressed with the lack of speculative coastlines in the work of Jean Rotz and with his letter dedicating the atlas to England's King Henry VIII where he implied that he drew only discovered coastlines. Rotz declared, "All this I have set down exactly and truly as possible, drawing as much from my own experience as from the certain experience of my friends and fellow navigators" (Wallis 1988a, 49). (4)
Wallis' research stresses the linkages between Dieppe, the Parmentier expedition and Sumatra. A map that shows all of them but has not until now been associated with Java la Grande is the map of Sumatra with which Giovanni Battista Ramusio illustrated his Italian translation of the pilot Pierre Crignon's account, written in 1539, of the expedition (Fig. 2.). The map is attributed to Giacomo Gastaldi, the Venetian cartographer, and was first published in Ramusio's Delle navigation et viaggi in 1556. That map along with others, according to Wallis (1988a, 41), "must have been obtained from Crignon". It is notable for showing the islands off Sumatra's west coast that the French encountered and named, and the ports of the west coast that they visited. Oddly, a different configuration of those islands, without the French names, appears in the depictions of Sumatra in the Dieppe maps. In the view of the present author the east coast of Ramusio's Sumatra shows links both to Dieppe and to Java la Grande. The link to Dieppe is manifest in the east coast nomenclature. The form of the names is almost identical with both the anonymous Harleian world map (c.1547) (Fig. 3.) and the regional map of Southeast Asia in Rotz's atlas (the nomenclature is more distantly related to other Dieppe maps) and, in common with the Harleian map, the line of the Equator is displaced to the north. (5) The link to Java la Grande is twofold: the eastern coastline not only looks remarkably like part of western Java la Grande, albeit rotated roughly 180[degrees], but has also a cape that seems to correspond with a prominent cape of Java la Grande and the two have similar names--whereas Ramusio's map has Campar, the Vallard atlas (1547) (Fig. 4. & Fig. 5.) has cap capar. (6) The omission of letters representing nasal sounds, m and n, is not unusual in Portuguese names in early maps so the latter name may well be read as cap campar.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
The present study focuses on a section of the coastline of Java la Grande that extends through about ten degrees of latitude from the large bay often named Baye bresille to the large inlet often named Haure (or Hame) de Sylle. The same coastline has been examined in the investigations of three authors, all of whom drew on place-names as evidence of identity. Those of Richardson (1983, 1989 & 2006) and Trickett (2007) aimed to identify the local features of Java la Grande, while that of King (2006 & 2009) aimed to identify the coastlines in general. Richardson held that this section of coastline depicts south-western Java, Trickett that it depicts the coast of Western Australia from King Sound to Shark Bay, and King that Java la Grande is a duplicate of South America, being a stage in the development of the fictitious Antarctic continent Terra Australis and being furnished with South American place-names.
[FIGURE 4 OMITTED]
These opinions are, in my view, not convincing. A cause for doubt is that the authors focused on different selections of place-names without demonstrating the effectiveness of their choices. Richardson (2006, 78) investigated names present in the Harleian world map and the maps of Desceliers (1546, 1550, 1553), Desliens (1541 [c.1561]) and Le Testu (1556) but he discounted those of the Vallard atlas (1547), with its abundance of names and distinctive west coast names, as probable inventions. (7) Trickett referred to the Vallard atlas almost exclusively, believing it to be authoritative on account of the same abundance of names and its high proportion of Portuguese-seeming names. And King selected, from many sources, names that he believed to be common to both Java la Grande and South America, but did not consider the significance of names that are present in Java la Grande but not in South America, and vice versa.
[FIGURE 5 OMITTED]
The results obtained by Richardson and Trickett are what may be expected from comparisons with the wrong coasts. The major Dieppe maps are overtly metric (i.e. drawn to a mathematical framework) and drawn to a single scale, with well-known exceptions such as Zanzibar, but Richardson deduced that Java la Grande is drawn without scale (i.e. it is non-metric) and Trickett assumed this coastline was drawn to several scales although he made no attempt to measure them. Richardson's results are all hypothetical and Wallis' criticism of them, "far-fetched and not proven" (1992, 1-13), seems justified. Trickett's results are unreliable, being for the most part obtained by the subjective method of visually matching the coastlines. In any case many of his interpretations of this coastline are based on misread, misquoted and misunderstood inscriptions. In brief, this coastline makes no sense as Java or Western Australia unless the cartographer is assumed to have been profoundly inept and unless reconstructions are applied to practically every aspect of the cartography and the text. The reverse applies to King's results. The Portuguese words and combinations of words that form his set of "common" place-names are just what could be expected in maps supposed to be of Portuguese origin, therefore they hardly indicate a dependent relationship of Java la Grande on South America.
If, as Richardson assured readers, "the place-name study techniques employed [by him] are no different from those of toponymists elsewhere" (2006, 9), then such techniques are perhaps not sufficient to solve as complex a problem as the origins of Java la Grande. To judge from previous attempts to identify Java la Grande, and its frequent description as a puzzle or mystery or enigma, investigators might do well to redefine their task as partly cryptographic. That means, among other things, paying critical attention to all of the place-name data, not just some part of it, and looking for systematic explanations. The many efforts to identify Java la Grande could have benefited from a preparatory study of the Java la Grande inscriptions, perhaps a "synoptic evaluation" such as McIntyre (1977, 117) advocated but never attempted. Even a simple kind of synopsis, a concordance of the inscriptions, has value. Without it one may not notice that the differences of nomenclature (apart from the mere number of inscriptions) between the Vallard atlas and other maps are largely confined to the west coast, or that in several maps, on short sections of the west coast, the order of names is reversed compared, again, with the Vallard atlas. (It is not yet clear what the reversals signify). A concordance also highlights something that has been mentioned often enough, that in different maps different names appear in the same locations. It has usually been interpreted either as evidence for invention or for one map or group of maps having more authority than others. However, other explanations in the same spectrum of possibilities have not been canvassed, namely that the various names could refer to different aspects of the same place or to different features in close proximity.
In the course of the present investigation the idea that spatial data might provide corroborative evidence was prompted by observations that the major Dieppe maps are metric--they bear scales of distance and latitude and occasionally longitude--and that if the coastline is Sumatran then its centring on the Tropic of Capricorn looks like an act done to a metric coastline. The question as to whether Java la Grande conceals metric data is one which has previously been avoided. One possible reason may be found in a quote from Richardson (1984, 1):
The late Avelino Teixeira da Mota, one of the leading authorities on Portuguese historical cartography, writing in 1973 about early portrayals of the north coast of South America stated 'certain simplistic criteria, now very much in vogue, are quite wrong when they interpret those outlines by measuring "latitudes" and "longitudes" on cartographic representations which resulted from voyages on which neither was measured.'.
Do da Mota's remarks apply to Java la Grande? If the coastline is Sumatran and drawn metrically then it ought to have a direct mathematical relationship with the actual Sumatra. It is generally agreed that early sixteenth century navigators were capable of determining latitude fairly accurately, but not longitude, therefore it would be appropriate to compare the latitudinal distances between the named features of Java la Grande and the same distances between the corresponding Sumatran landmarks that have previously been identified by place-name evidence. The results of such a comparison are described under the heading "Distribution of place-names" and presented graphically. (Figs. 7, 8 & 9.)
The purpose of the rest of this paper is to discuss the correspondences between the western coastline of Java la Grande and the eastern coast of Sumatra. The discussion will involve an examination of the inscriptions of Java la Grande, current and historical place-names of Sumatra and an examination of the geography of Sumatra. There will follow a comparison of latitudinal distances between the features of Java la Grande and the corresponding Sumatran places. A sketch of Sumatra locating places referred to is also provided. (Fig. 6.)
[FIGURE 6 OMITTED]
SIGNIFICANT PLACE-NAMES NORTH OF THE TROPIC OF CAPRICORN
The clearest indications of a Sumatran coast are the place-names at five locations on the stretch of Java la Grande's western coastline from just south of Baye bresille to the Tropic of Capricorn. For convenience the place-names are discussed in order of increasing latitude under the names given in the Vallard atlas.
Rio camor appears to correspond with the stream named Wai Sumur in far south-eastern Sumatra. The name evidently shows a use that was frequent in Portuguese spelling in the early sixteenth century of c for the initial s sound.
While the Vallard atlas has Rio camor, other Dieppe maps of Java la Grande have the names [Y.sup.S] des (islands of the) [incomplete inscription] (Harleian c.1547), [Y.sup.S] de [d.sup.o] (islands of [d.sup.o]) [obscure meaning] (Desceliers 1546) and [Y.sup.S] frimoses (beautiful islands) (Desceliers 1553), which apparently refer to a group of nearby islands. Le Testu (1556) has Rivierre /noeirre (black river). (8)
Trickett (2007, 302) stated, "Probably the nearest we can come to this [Rio camor] in Portuguese is 'Rio Comodo' or 'Commodious River'," and identified it with King Sound in Western Australia while acknowledging that King Sound is not a river but a gulf fifty kilometres wide. Neither Richardson nor King considered the name.
A Dutch map of 1598, which is certainly based on a Portuguese source, shows an island named Sumor in the Java Sea off the south-eastern point of Sumatra (Lodewijcksz 1598). Hessel Gerritzs (1620) places Laban Samora de water plaets (port Samora the watering place) on the western side of this point. (9) Another map by Johannes Jansson (1649) places Labansamora on the east of the point. An English map dated 1661, based on Dutch sources and published in Italy, has multiple instances of the name Samor written at the same place (Dudley 1661, f.89). The main Italian inscription reads, "C: Samor /per Fare aqua" (C[ape]: Samor / for taking water). Dutch maps more generally notice this coastline as a source of Vers water (fresh water).
By contrast, in the Dieppe maps' depictions of Sumatra, the south-east regularly bears the name Andalas, an early state of Sumatra and now an alternative Indonesian name for the island. (10) Acampu is present in some Dieppe maps (Desceliers 1550) and in some Portuguese maps acampu (Viegas c.1537): the name is a mutated form of Cacampom by which Tome Pires knew the river Wai Sekampung (Cortesao 1944, 158 fn.2).
The Indonesian word sumur means "well, pit or shaft" and is the normal word for a fresh water source (Stevens & Schmidgall-Tellings 2004, 968). (11) The word is the name of the stream Wai Sumur, the village at its mouth Kampung Sumur, the nearby cape Tanjung Sumur Batu and the neighbouring island group Pulau Sumur (known to the Dutch as the Zutfen or Zutphen Eilanden) (Hydrographic Office 1956). An alternative name for Tanjung Sumur Batu was Ujung Curam, a name which suggests that it could be the origin of the Cabo Cur, sometimes shown at the southern point of Sumatra in early maps. The Indonesian word curam (steep, sheer, precipitious, abrupt (of a cliff/abyss, etc.)) (Stevens & Schmidgall-Tellings 2004, 217) also suggests the possibility that the inscriptions [Y.sup.S] des and [Y.sup.S] de [d.sup.o] stem from a Portuguese word such as despenhadeiro (precipice, cliff, crag) (Taylor 1970, 221).
Rio patano south-west of Rio camor seems to be a reference to the bay Teluk Batindalem at the mouth of the river Wai Tulangbawang.
The Vallard atlas has Rio patano. Other Dieppe maps name the place Baye basse (low or shallow bay) (Harleian c.1547, Desceliers 1546), R: (river) (Desceliers 1550, Desliens 1541 [c.1561]) and R: [p.sup.o] (Desceliers 1553).
Trickett (2007, 310) misread the name as Rio Palano and stated that it "should probably be 'Rio Plano', meaning 'Smooth River' and identified it with Pender Bay in Western Australia. Neither Richardson nor King considered these names.
A seventeenth century Dutch map has the name bocht van patan (bend or bight of patan) in southeastern Sumatra (Essen 16??). It lies between places described in this map as eijlant met de hooge boemen (island with the high trees) and de groote sandbanck (the great sandbank). In other words it is situated on the bight that runs from the place later called Tree Island (near Tanjung Lumut) southwards to the mouth of Wai Tulangbawang, one of the main rivers of southern Sumatra. A smaller river, Air Mesuji, enters the sea about the middle of this coast. Neither patano nor patan appears in later maps nor do they make sense as Portuguese or Dutch words but they can be explained as Malay. The only geographical feature on this coast, besides Air Mesuji, is T[eloe]k. Batindalem (TDNI 1924), the bay on the northern side of the mouth of Wai Tulangbawang. It is probably a compound of the Malay words beting or bating (narrow shoal, sandbank at river mouth) and dalam (inside, interior) (Army Map Service 1944, 21 & 33). The derivation of patano (and patan) from beting or bating makes sense given the Portuguese rendering of the initial "b" sound of some Malay words with "p" and of "ng" with "ao" (i.e. ano). (12)
The inscriptions relating to this place seem to relate to the bay Teluk Beruga and the two short rivers that converge on the river-mouth Kuala Sondan.
The Vallard atlas has Rameriqua for a river-name between Rio patano and cap de vert. In other maps it is a bay or river named B: de gao (Harleian c.1547), R de gao (Desceliers 1546), R: de ga[?] (Desceliers 1550), R: basse (shallow river) (Desceliers 1550, 1553) and Rivierre / de corptz (corpse river) (Le Testu 1556). The maps of the Pastoret atlas (1587) have Riviere a cap de grace (shoal cape river).
Richardson (2006, 76) understood gao to mean "Java" or "Javanese" (modern Portuguese jao) which he held to be unambiguous evidence that this Java la Grande coastline depicts Java. Trickett (2007, 310) misread Rameriqua as Rio Meriqna and stated, "the closest modern Portuguese word is probably 'meritorio', giving the meaning of 'Meritorious River'", and identified it with Beagle Bay in Western Australia. Richardson noted its likeness to "R. America". King did not consider the name.
The coast north of Tanjung Lumut is relatively featureless and sparsely inhabited. Several streams break the coastline, the longest of them being Sungai Lumpur (mud river). Two river-names sometimes appear in early maps, Rio. S. Juoan (St John's River) and Rio. S. Clara (St Clara's River) (Essen 16??), but their modern identities have not be ascertained. Once again, the only geographical feature on this coast, apart from these rivers, is a bay T[eloe]k. Beroega (TDNI 1924). At Kuala Sondan two short and shallow rivers called Sungai Sondan-besar and Sungai Sondan-kecil flow into it. The word beruga (in Dutch spelling beroega) appears to be a variant of the Lampung word berugo (jungle fowl) (Junaiyah 2001, 39) and Indonesian beroga (ayam beroga = red jungle-fowl) (Stevens & Schmidgall-Tellings 2004, 127). Although the spelling of beruga and Rameriqua may seem dissimilar, their sounds bear enough resemblance to each other to support a hypothesis that the one is the source of the other. (13) Beruga also supports interpretations of gao as Portuguese galo (cock, rooster) and as the sixteenth century French jau (cock) (Huguet 1925-1973, 4/708) which has since gone out of use. (14) The name Teluk Beruga seems to have made its appearance in maps only in the twentieth century. The most probable cause is not that it is recent but that the cartographers of earlier times paid close attention only to the abovementioned Tree Island, an important navigational landmark that lay adjacent to the bay. The now-vanished island was regularly depicted and named from the early seventeenth century until the late nineteenth.
Le Testu's name for the river, Riviere de corptz, looks at first sight to be an invention but the meaning of French corpse coincides with an obsolete meaning of the Indonesian word batang--"corpse, carcass or carrion" (Stevens & Schmidgall-Tellings 2004, 96). Batang (trunk, etc) is often used to qualify the lower reaches of Sumatran rivers so it is possible that the wrong translation was preserved in Le Testu's source. (15) The possibility should be noted that Le Testu's "corpse river" may actually refer to the river Wai Tulangbawang, for he left the Rio patano unnamed and the sequence of his Java la Grande inscriptions is often out of step with other maps. (16)
cap de vert
The name occurs at the westernmost point of Java la Grande, which seems to correspond to one of the easternmost points of Sumatra, either Tanjung Kait or its neighbour Tanjung Jatigombol, at the southern entrance to Bangka Strait.
In the Vallard atlas the name is cap de vert (cape of green or green cape). Other Dieppe maps describe this place as C: de grace (cape of grace) (Harleian c.1547; Desceliers 1546, 1550 &1553; Desliens 1541[c.1561]), C: des basses (cape of the shoals) (Desceliers? c.1555) and C de gre (Le Testu 1556). No doubt gre is an abbreviation of grace, which word may be understood as a corruption of Portuguese pracel (shoal) (Richardson 2006, 77-78).
Richardson did not offer an identification. Trickett (2007, 310) estimated that, "This cape is shown approximately in the position of the headland of Coulomb Point" in Western Australia and he identified it as such. King did not consider it.
In the maps of Dieppe and in Portuguese maps depicting the easternmost point of Sumatra, the nomenclature is typically Saida do canal (Exit from the channel), referring to the southern entrance to Bangka Strait, and I. de lusapara (Lucipara island) (Homem 1568). In Dutch maps two headlands in this area are typically named: Eerste hoek (the first point), the first in a series of numbered points going north through Bangka Strait; and a little to the south Lucipara punt (Lucipara point, opposite Lucipara island). However, both Dutch and English maps from the end of the eighteenth century often name the intermediate point Groen hoek or Groene hoek or punt (green point or the green point). It was, according to Findlay (1878, 192), "so called from the trees on it being of a lighter and brighter green than elsewhere". In the nineteenth century Lucipara Point was itself described as "a low green point" (Horsburgh 1852, 166) and on some maps "a Low Green Point" is placed (or misplaced) just south of it (Hydrographic Office 1942).
Eerste hoek is now called Tanjung Jati, Groene hoek is Tanjung Jatigombol, Lucipara punt is Tanjung Kait and Lucipara island is Pulau Maspari. The Dieppe inscriptions C: de grace (shoal cape) and C: des basses (cape of the shoals) describe Tanjung Jatigombol better for it is adjacent to numerous shoals and banks including the Hindostan Bank and Merapi Droogte at the entrance to the strait. On the other hand, Tanjung Kait is a much more prominent landmark and more important, in conjunction with Pulau Maspari, in navigation. It is possible therefore that the inscriptions apply to different geographical objects.
As mentioned in the introduction, Ramusio's map has Campar as the name of a cape on the east coast of Sumatra. His Campar could refer to the ancient kingdom of the same name but it probably refers to the Kampar River itself because his map of Southeast Asia of 1554 has CAIMPAR. R. (sic). The river flows into the Strait of Malacca opposite Singapore.
In the Vallard atlas the name of the corresponding cape of Java la Grande is cap capar. In the other maps of Java la Grande the cape has the names Quabesequiesce (Harleian c.1547), Quabese/quiesce (Desceliers 1546), Cap quiesce (Desceliers 1550), C: quiesce (Desliens 1561), C: de Iaua (cape of Java or Javanese) (Desceliers? c.1555), C fremoze (beautiful cape) (Le Testu 1556) Cabesse quesse (Pastoret 1587) and Cap [p.sup.o] (Desceliers 1553).
Richardson wrote nothing of cap capar but considered the alternative Quabesequiesce and its relatives (see more below). Trickett (2007, 313) stated that cap capar "is in reality a garbled abbreviation of the Portuguese word for Capricorn--'Capricornio'", and so identified it with North West Cape in Western Australia. King did not consider it.
Richardson translated the first element of the inscription Quabesequiesce as Portuguese quabeb (cubeb), the name of a spice, and the rest as Portuguese aqui esta (is here) (2006, 75-76). As cubeb grew only in western Java so he identified the coastline as Javanese. However, it is not necessary to assume a textual error quabes for quabeb. A better case can be made for understanding Quabese and Cabesse as the Portuguese word cabeca (head) thus these inscriptions say literally "head is here". In several of the Dieppe maps French cap (cape) is substituted implying that those map makers imagined the word to mean "headland". However, Portuguese cabeca can be used to mean the head of a river and of other bodies of water therefore these inscriptions can be explained, because the mouth of the Kampar river is located at the very end of the Strait of Malacca, as a shortening of the sentence [a] cabeca [do estreito de Malaca] aqui esta (the head of the Strait of Malacca is here).
In the Dieppe maps of Sumatra the same place appears under the names: ampar (Rotz 1542), Ampar (Harleian c.1547), Rio dampar (Vallard 1547), Campar (Desceliers 1546), and campar (Desceliers 1550). And the name appears in Portuguese-derived sixteenth century maps of Sumatra as bocas de Sapar (mouths of Kampar) (17) (Luis 1563), camper (Martines 1587) and in the seventeenth century as Bocas d. Capar (mouths of Kampar) (Blaeu & Hondius 1624). The absence of the initial letter of a name is a phenomenon found elsewhere in the Dieppe maps' depictions of Southeast Asia.
SIGNIFICANT PLACE-NAMES SOUTH OF THE TROPIC OF CAPRICORN
The visual resemblance of the two stretches of coastline continues southwards in Java la Grande (and northwards in Sumatra) therefore it is a reasonable conjecture that they are also identical. Two stretches of coastline are especially indicative of Sumatra.
port tamto, baie perdue and coste blanche
These names probably relate to the stretch of coastline around the mouth of the Rokan River that was regularly described by the seventeenth century Dutch as Groote Rivier Rackan (the great river Rokan).
The Vallard atlas has two inscriptions written quite close together--baie perdue (lost bay) and port tamto. (18) A third inscription coste blanche (white coast) appears in several other Dieppe maps a little to the south (Harleian c.1547, Desceliers 1546 and Desceliers 1550). A fourth is G: [p.sup.o] (Desceliers? c.1555) which, in light of baie perdue, might be expanded as Portuguese G[olfo] p[erdid]o (lost gulf).
Richardson made no comment on the instance of baie perdue on the western coastline, but regarding the one on eastern coastline he wrote, "it seems a somewhat strange name to give to any bay, especially one located on what was apparently a little-known, or hitherto unknown coast. It would appear that this is another case of the French having misunderstood a Portuguese inscription" (2006, 85). Trickett (2007, 316) believed that this name "could mean either 'Lost Bay' or 'Shipwreck Bay'", but he could not identify it. King did not consider it. Concerning port tamto, Trickett (2007, 316-317) believed that the name is "possibly a misreading of 'Porto Tamanho', meaning a 'large' or 'impressive' port" and identified it as Carnarvon Harbour in Western Australia. Neither Richardson nor King considered it. Richardson (2006, 78) wrote that several inscriptions such as coste blanche (white coast) "do not seem sufficiently specific to be helpful" and this one "could so easily just refer to stretches of white surf. Trickett (2007, 316) believed that it applies to the Western Australian coastline near Ningaloo Reef and describes the coast's "brilliant white coral sands". King did not mention it.
The identification of the port depends on an understanding of baie perdue. A large natural body of sea-water can be lost by geological or hydrological processes. The latter applies especially to the extremely broad and shallow estuary of the Rokan River. Findlay (1878, 146) wrote:
The greatest breadth of the mouth of the Reccan River (sic) is about 15 miles, decreasing about 8 or 9 miles up to 4 miles, afterwards 2 miles, and then continuing this breadth until it forms the two branches mentioned above [one called Banka and the main branch called Tanah Putie River]. It is almost dry at low water spring tides, and is rendered exceedingly dangerous by their excessive rapidity of 7 miles per hour, producing a bore on the springs, and having a rise and fall of 30 ft. (19)
In a seventeenth century Dutch map the mouth of the Rokan River is named Porto Lalang besaer (Port of Great Lalang) using an earlier name of Pulau Halang, the largest island in the estuary (Essen, 16??). It is the only place on the eastern coastline classed as a port in this comprehensive map.
An explanation of the name port tamto might be sought in a textual error, tamto for samto, a spelling of Portuguese santo (holy). (20) But it might otherwise be derived from the name Tanahputih, literally "white land". A nineteenth century Dutch map gives the name Tonah Putin to the top of the peninsula on the right bank of the Rokan river and the name [S.sup.[on]gij] Tanah Patto to a stream on its south side (Derfelden van Hinderstein & Tindal 1872). The origin and age of this name is not clear but many other places on the eastern coast of Sumatra haveputih (white) in their names on account of white sandy beaches and Tanahputih is a fair approximation of coste blanche written in other maps in about the same place.
The relative position of Rio bono in the coastline together with an accumulation of small correspondences of names and geography tend to support its identification as Teluk Aru (Aru Bay). Ramusio's map shows a large round bay on the north side of which is a cape named Tuncan (Ujung Tamiang or Point Tamiang). The shape of the bay differs in other maps, for example Martines (1587) shows the cape Tamiam immediately north of a large funnel-shaped inlet, the only one on his east coast, like the one that appears in the Dieppe maps. In both maps the bay can be identified with Teluk Aru. The Portuguese knew the bay by the river-name R. depitaio (Silveira 1991, 29) and Rio de Picitao (Albernas c.1680), variations of the native name Sungai Besitang. The other river that empties into the bay is Sungai Salahaji.
In the maps of Java la Grande, at either the head or the middle of the inlet, the inscriptions are: Hame de Sylle (Harleian c.1547), Haure (or Hame) de sille (Desceliers 1546), hame de sille (Desceliers 1550), Haure de Sylle (Desliens 1561), Hame desille (Pastoret 1587), R: de fundo (Desceliers 1553), R: grande (Desceliers? c.1555) and B de tuches (Le Testu 1556). (21) At the mouth of the inlet are: R: de sille (Desceliers 1546), Rio bono (Vallard 1547), Cap (Harleian c.1547 and Desceliers 1550), Riviere Grande (Desliens 1541 [c.1561]), Rivierre / de Sille (Le Testu 1556) and Riviere bonne (Pastoret 1587). An island at the mouth of the inlet is named once as [y.sup.e] de [p.sup.o] (Desceliers? c.1555) and once as [Y.sup.e] de [S.sup.e] (Desliens 1541 [c.1561]) and thirdly as [Y.sup.S] debare (Desceliers 15 5 3). (22) The meanings of only a few of the inscriptions are immediately clear: R: grande (big river), Riviere bonne and Rio bono (both, good river) and Cap (cape). R: de fundo probably means "river in the background" judging by current Portuguese usage. B de tuches may be translated literally as "bay of touches" which could be a misinterpretation of de fundo under the influence of French toucher le fond (to touch the bottom).
Richardson (2006, 76) considered only the Haure de Sylla (sic) family of names. He concatenated two apparently unrelated inscriptions to make Haure de Sylla/cap and identified it with the southern Javanese port of Cilacap. Trickett (2007, 317) identified Rio bono with Shark Bay in Western Australia despite its anomalous orientation and its not being a river. King (2006, 8) considered Havre de Sylla (sic) to be the same name as R la Silla on the northwestern coast of Peru. (23)
The Haure de Sylle inscriptions in the Dieppe maps are difficult to interpret. Haure or Hame may be explicable as French havre (haven) (Richardson 2006, 76) in which case the Portuguese cognate is abra. In early maps Portuguese abra is "a small harbour between two mountains or hills, also that formed at the mouth of a river, equivalent to French havre or English haven" (Vidago 1953, 45). The definition is certainly true of Teluk Aru--the main attraction of the bay in earlier times is thought to have been its safe anchorage (McKinnon & Sinar 1981, 66). Sylle/sille could be a reflection of Silha, the name applied to Sri Lanka (formerly Ceylon) in the Travels of Sir John Mandeville--in a copy of the Travels owned by John Dee (1527-1608) is the marginal note "Silha Isle Seilan now called" (Tzanaki 2003, 275). If so, it belongs to the same group of names that was applied to depictions of Sumatra in the earliest years of the sixteenth century--Zoilon, Seylan, Seijla, Seylam, Zeila--before Sri Lanka and Sumatra were fully distinguished cartographically (Suarez 1999, 96-122). (24) Alternatively, the name is reminiscent of the word cella that is written alone and in the combinations Civitas cella and populi de cella in Fra Mauro's depiction of Sumatra in his world map of 1459, the first European map to identify Sumatra by its present name. Falchetta (2006, 216) stated that Civitas cella refers to the current Banda Aceh and populi de cella to the region of Aceh but he also noted, without drawing attention to the contradiction, that Joao de Barros mentioned that the Cellates were an ethnically Malay, not Acehnese, population of Sumatra. Teluk Aru lies at the southeastern border of Aceh.
Leaving aside the uncertainty about its most frequent name, the geography of the inlet still seems to fit that of Teluk Aru. R: de fundo could describe Sungai Besitang the entrance to which is concealed by Pulau Sembilan, the largest island in the bay. It is also a river with numerous sandbanks and therefore one that could also support Le Testu's interpretation. In the Vallard atlas the next river to the north is named Rio tercio (third river) but in the Desliens 1541 [c.1561] map it is named Riviere de [g.sup.o], possibly "River of Java or Javanese", therefore perhaps a reference to the site of the long abandoned fortification known as Kota Jawa (fort of the Javanese) on the right bank of the Deli river south of Teluk Aru (Milner, et al. 1978, 29). Immediately to the north the Vallard atlas has the inscription tarra alta (sic) (high land). The region of Teluk Aru is one of the few coastal places that has significant elevation on the east coast of Sumatra, and the mountainous interior is sometimes visible from the sea. Anderson (1826, 278) wrote, "The Delli (sic) river takes its rise at the foot of Gunong Kuali and Sukanala, (25) two lofty mountains which may be seen from the sea on a clear day" and, on approaching Ujung Tamiang from the sea, "At day-break the low land of the Sumatra shore [is] visible from the deck, and the lofty mountain peaks in rear towering above the clouds" (1826, 5). And Coste blanche or bracq (white coast) immediately to the south of the inlet in Java la Grande seems characteristic of the nearby coastline such as the "sparkling white beach" of Pulau Kompei on the north side of Teluk Aru (McKinnon & Sinar 1981, 62).
If this coastline depicts eastern Sumatra, as it seems, then the intermediate place-names ought to refer to Sumatran places too: immediately south of cap de vert the names that are applied to a river and its flanking capes ought to depict the main features of the Palembang coast; and between baie perdue and Rio bono the names ought to refer to places along Sumatra's north-east coast. The place-name evidence is less clear but not inconsistent with Sumatra and seems to be supported by a spatial comparison with Sumatran landmarks.
Place-names immediately to the south of cap de vert
Immediately south of cap de vert (also known as c: de grace and c: des basses) are two capes flanking a river or a river in a bay. The first place to the south is a cape named C: de [s.sup.t] drao (Harleian c.1547) and C: de [s.sup.t] [a.sup.o] (Desceliers 1546). Next is a river named severally Rio canicolla (Vallard 1547), Baye des [y.sup.s] (bay of islands) (Harleian c.1547), R (river) (Desceliers 1550) and R: darenes (river of sands) (Desceliers 1553). In Desliens (1541 [c.1561]) the only river on this stretch of the coastline is located just south of C: de grace so the only river-name Riviere / po presumably relates to it. Le Testu (1556) haspetitte Riviere (little river). South of the river is the second cape named cap mallo (bad cape) (Vallard 1547).
Of C: de [s.sup.t] drao and C: de [s.sup.t] [a.sup.o], Richardson (2006, 77) wrote, "It seems probable that the original Portuguese inscription was cabo [do] padrao (cape of the padrao), or possibly just padrao, misunderstood by the French as sadrao, ignoring the tilde". Indeed, there is one cartographic indication that the Portuguese erected a padrao, a memorial pillar, on the Sumatran coast, but it was north of the Equator at a place they called R. do padrao, between Air Panai and Sungai Asahan, not south of the Equator where the Java la Grande names actually appear (Silveira 1991, 29). Trickett (2007, 311) considered only cap mallo and identified it with Gantheaume Point in Western Australia. King did not consider the names.
The earliest Portuguese depiction of the coast of Palembang, drawn by Francisco Rodrigues as a result of the exploratory voyage of the Portuguese to the Moluccas in 1512, bears the river-names as tres bocas depallambam and cacogia. The first is "the three mouths of Palembang". The channels that flow north from Palembang, the main one of which is Air Musi, form islands in the flood plain and they could be the inspiration for the name Baye des [y.sup.s]. Certainly, many early maps show numerous islands north of Palembang, and Wolters (1979, 46) wrote, "The scene near Palembang in earlier times probably resembled a large lake with numerous and ever growing islands ... The memory of the islands is preserved today in the expression 'Thousand Islands' (Pulau Seribu) to refer to the southern bank opposite Palembang city". In the name Riviere po the abbreviation could be interpreted as Palembang but po is frequently found in the Dieppe maps and could have several sources including the Portuguese primeiro (first) and perhaps in this case poderoso (powerful) (26) although Le Testu's petitte Riviere suggests that his source bore the same abbreviation and that he interpreted it as Portuguese pequeno (little). R: darenes is a fitting reference to the sedimentation at the mouths of the channels. Rio canicolla is enigmatic. (27)
Rodrigues' cacogia now has the form Kuala Sugihan. It gives indirect evidence of a possible identification of the Java la Grande names C: de [s.sup.t] drao and C: de [s.sup.t] [a.sup.o], perhaps Tanjung Selokan or Tanjung Limaubungkuk. The Dutch usually called the cape immediately to the east of the river Derde hoek (the third point) but sometimes Tandjoeng Kesoegian after the river. Nowadays the cape is called Tanjung Selokan using the word selokan meaning "gutter, drain, ditch, chute, trench" (Stevens & Schmidgall-Tellings 2004, 900). The meaning coincides with two Portuguese words sangradouro (drainage ditch; trench, canal; spillway; bleeder valve) (Taylor 1970, 568) and desaguadouro (drainage ditch; spillway) (Taylor 1970, 208) and their forms suggest they could have been the source of the Java la Grande abbreviations. The first word has its origin in the traditional site, the inner elbow, of surgical blood-letting, and an account of a natural phenomenon suggests why such a word might have been used in a place-name: Peter Floris, (28) sailing westwards across the mouths of the Palembang in June 1612, reported, "Heere wee were forced to come to an ancker by reason of the greate streame which came against us, having very read [red] water on the coast of Sumatra, a league or more from the shore att 8 or 9 faddem [fathom] ..." (Moreland 1934, 30). His narrative indicates the ship's position as west of Tanjung Limaubungkuk, named Vierde Hoek (the fourth point) by the Dutch, on the Sumatra shore and south of the hill Gunung Menumbing on Pulau Bangka. If these indications are true then cap mallo is probably Ujung Batakarang, a headland often named in early maps, north of Palembang.
The coastline further towards the Tropic has few inscriptions in any map. The Vallard atlas has the isolated word Gouffre (gulf), then a river named S. fransois (St Francis?) and finally G: ancra (possibly, gulf + small bay). The word fransois may be suspected of being a misspelling of fremose the French loan-word from Portuguese formoso (beautiful) because Le Testu applied C fremoze to the place where the Vallard atlas has cap capar and on his map this cape is one step away from the river S. fransois of the Vallard atlas. It may be supposed from its location that S fransois refers to the first great Sumatran river south of the Equator, Sungai Indragiri (or Inderagiri), therefore the origin of the fransois-fremoze inscriptions could be the translation of Indonesian indah (beautiful, pretty, charming, picturesque) (Stevens & Schmidgall-Tellings 2004, 385) on the false premiss that this word is the first element of the river-name. In the Dieppe depictions of Sumatra, Indragiri is variously written amdragu[?] (Rotz 1542), Andraga (Harleian c.1547), R: andra[...] (Vallard 1547), andracin (Desceliers 1546), adraguimi (Desceliers 1550), Andragam (Desceliers 1553), andragu (Desceliers? c.1555), and andra /guim (Desliens 1541 [c.1561]). (29) All the major maps show in the middle of this stretch of coastline a line of islets or rocks (usually three) named just once as Roches (Desceliers 1550). They have the right number and location for thinking that they might depict the three islands of the Alangtiga group which lie in the middle of the shipping lane. Perhaps the inscription basses (shoals) (Desceliers 1553) in the same place also refers to them.
Other names south of the Tropic
On the Vallard atlas coastline south of cap capar are Rio grant (big river) and the widely spaced words Goulfo (gulf) and amatora. If Rio grant is not a reference to the Kampar then it could indicate the Siak, another important navigable river which is otherwise unnoticed. Goulfo and amatora perhaps comprise a single name. The latter word could be an instance of the name Sumatra where an initial S, C or C has been omitted (compare the above-mentioned ampar, Ampar and Rio damper for Campar (Kampar) and acampu for Cacampom (Sekampung)). (30) Almost the same spelling appears on a very early Portuguese map of the Indian Ocean dated 1509: a muyto homrrada & muyto Rico ilha de samatoro (the very worthy and very rich island of Sumatra) (Uhden 1939, 10).
On the coastline of the Vallard atlas going southward from port tamto to Rio bono the inscriptions begin to refer to high country. It is consistent with northeastern Sumatra but the first inscription Serra alla is problematic. A Portuguese serra (mountain range) is not expected at this point unless it refers to the mountains of the interior; if it does, then alla should be an error for alta (high). The inscription is more likely an error for Terra d'Aru (land of Aru) which usually appears in northeastern Sumatra in early maps. The next name is cap negro (black cape). Richardson (2006, 77) reported that Portuguese negro (black) is sometimes mistaken for Malay negeri (village) in later maps of the East Indies, so it is possible that, if it is not a literal description, then it might be a reference to a coastal village. In eighteenth century Dutch maps the districts of northeastern Sumatra are generally marked by a sequence of river and negeri names from the Asahan river north to the Deli river: [R.sup.r] Assahang, De Negerey assahang, [R.sup.r] Catoebara, De Negerey Catoebara, Songi Badageen, Negery badageen, Negerij delij, [R.sup.r] Delie. (Anonymous c.1742). Beyond cap negro the names indicate high places and the rivers are mostly numbered. They are: Rio amgra (river + small bay), Rio secondo (second river), cap hault (high cape), Rio tercio (third river) and the already-mentioned tarra alta (sic) (high land). They seem consistent not only with the geography of northeastern Sumatra but also with the often poor European geographical knowledge of it that lasted into the eighteenth century. Note that Ramusio's map of Sumatra gives only three place-names in the seven hundred kilometres between Campar (Kampar) and Tuncan (Tamiang), namely Liaca (Siak), (31) terra d'aru (land of Aru) and Entree des basses (Entry to the shoals).
DISTRIBUTION OF PLACE-NAMES
A comparison of the latitudinal distances between the features of Java la Grande and the corresponding Sumatran places seems to corroborate the place-name identifications (Fig. 7.). The distances obtained for comparison are, for Sumatra, the distances above or below the Equator calculated from the latitudes supplied, in most cases, by Garren, et al (1982). (32) For Java la Grande, they are the distances above and below the Tropic of Capricorn calculated from the pixel coordinates in a digital image of the relevant chart of the Vallard atlas using the latitude scale drawn in that chart. (33) (The Vallard atlas was used because of the ready availability of a high resolution digital image). The main assumptions are that the length of a degree of latitude is sixty nautical miles (as it should be if the latitudes were established by astronomical observation, irrespective of the navigator's or cartographer's estimation of the circumference of the Earth) and that capes that reach further south than adjacent river mouths are exaggerations that may be discounted. In this type of comparison, if the two coastlines are exactly the same and drawn to the same scale, the lines of correlation between places should all be parallel; but if one coastline is drawn to a smaller scale, then the lines should resemble the ribs of a fan that converge on a point of origin on the side of that coastline. On the other hand, if one coastline is different or defective then some crossed lines or irregularly fanned lines, or fanned lines with points of origin on both coastlines, may be expected.
There are three fundamentally positive points of comparison with Sumatra. Firstly, the perpendicular distance between the latitudes of Tanjung Tua, the south-eastern point of Sumatra, and Teluk Aru is practically the same as the distance between those of the corresponding (unnamed) northern point of this part of the Java la Grande coastline and the northern side of the mouth of Rio bono--603 and 601 [+ or -] 3 nautical miles respectively. (34) Secondly, the lines of correlation are remarkably parallel south of the Tropic; and, north of the Tropic, they are noticeably parallel at points which might be taken to be especially significant to navigation. Only one pair of lines intersects. Thirdly, cap capar is clearly a depiction of the vast "headland" between Sungai Kampar and Sungai Indragiri.
The single instance of intersection--an unnamed river and Rio grant--could be taken as a sign of an incorrect identification but it is probably caused by an original cartographic error. The intersection occurs at a point which, the lines of correlation suggest, is the junction between two separately drawn coastlines. (35) Two other views of the latitudes (Fig. 8. & Fig. 9.), suggest that the cartographer failed to reconcile the coastlines and in doing so failed to locate Sungai Kampar and Sungai Siak accurately. The remaining irregularities are the depiction of Rio bono, which appears to be an in situ enlargement of the kind used before the invention of insets, and the far northern correlation with Rio camor. The latter has the most irregular correlation while being among the most securely identified by name. The cause could perhaps be the wrong trend of the coastline (northeast-southwest instead of north-south) or the cartographer's need to compress this section in order to maintain the overall north-south dimension of the coastline.
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
The correlations seem to support the possibility, mentioned earlier, that Rio tercio (also named Riviere de [g.sup.o]) is Sungai Deli. And they seem to support new and revised conjectures, notably relating to Salmedia and port tamto. If the correct alignment is taken to be that of the northern side of the mouth of Rio bono then it suggests an identification of Salmedia with Tanjung Jambuair. A short distance west of this cape is the site, at Lhokseumawe, of the ancient state of Samatra so it may be wondered whether Salmedia is a textual error for Samatra. Secondly, the location of the port tamto inscription of the coastline of the Vallard atlas corresponds, not with Tanahputih, but with Pulau Alang-besar (Pulau Halang), the island in the estuary of the Rokan River which, as mentioned above, was known as a port to the seventeenth century Dutch.
In the introduction attention was drawn to the prominence of Sumatra in Helen Wallis' theory of Java la Grande, to the consideration of different selections of Java la Grande inscriptions by recent investigators together with their incompatible identifications of its western coastline, and to the absence from the literature of interest in spatial data. Ramusio's map was noted as the immediate trigger of this investigation because it afforded two clues, in the coincidence of the shape of its eastern coastline and of the name of a prominent cape, that eastern Sumatra might be the origin of part of western Java la Grande.
The investigation was begun on the premiss that inscriptions on the major Dieppe maps present authentic, although varied, views of an actual coastline. If that were true then their links to actual places should be predictable, direct and systematic. The most convincing sign that this coastline of Java la Grande depicts eastern Sumatra is the presence of several native names, all in the Vallard atlas (Rio camor = Wai Sumur, Rio patano = Teluk Batindalem, Rameriqua = Teluk Beruga, cap capar = Sungai Kampar). Of these names Rameriqua appears in other maps in Portuguese and/or French translation (gao may represent Portuguese galo and French jau). The presence of so many native names in the Vallard atlas leads to a suspicion that its S.fransois and Le Testu's C fremoze are mistranslations of the first element of the river-name Indragiri and that other inscriptions may prove to be native names or translations (e.g. the unexplained Rio canicolla). The meaning of the alternative name Ujung Curam of the headland near Sumur suggests that the abbreviated names [Y.sup.S] des and [Y.sup.S] de [d.sup.o] present another instance of translation, in this case into Portuguese despenhadeiro. In the non-Vallard maps, the inscriptions that are not translations appear to be concise descriptions of the coastal geography, notably cap de vert (green cape = Tanjung Kait or Tanjung Jatigombol) and baie perdue (lost bay = Sungai Rokan). There is room for suspicion that some maps show textual error both in the form of names (e.g. Serra alla =? Terra d'Aru; S. fransois =? fremoze) and in their locations (notably in Le Testu's map) but it is not so great that doubt may be cast on the identification of the coastline. The metric comparison appears to corroborate the identifications produced by the place-name investigation, and the results provide an insight into the construction of the Java la Grande coastline.
A discussion of the identity of the adjacent coastlines and islands requires more space than this paper can afford. There is no reason to believe that the islands are not Sumatran but a discussion of them must necessarily include two most interesting and contentious names, namely Baye bresille and [Y.sup.e] de neige. It seems possible that the coastline that extends beyond Rio bono up to the southern end of the coastline as depicted in the Harleian map and in Jean Rotz's atlas is also Sumatran. On the other hand, the southern extensions of the coastline, the short one of the Vallard atlas and the much longer ones of the other maps, show no obvious signs of belonging to Sumatra.
The finding of Sumatra in Java la Grande and of the variety of names of its landmarks hints that the source of information available to the Dieppe cartographers was richer than hitherto imagined. It renews the old questions of the origin of the information and the means of its transmission to Dieppe; it provides a new basis for trying to understand the relationships both of the various maps of the Dieppe school to each other and to other depictions of Sumatra; and it raises new questions as to how and why this coastline came to be embedded in Java la Grande.
Richardson's and Trickett's identifications of this coastline with, respectively, Java and Australia can hardly be maintained and it now seems less feasible that Java la Grande could have been furnished with South American names as proposed by King. Helen Wallis' theory seems to be partly right in presuming a Portuguese discovery, in as much as the first European ships to enter the Strait of Malacca were Portuguese, but wrong in that this particular coastline can hardly be Australian. She may have read a little too much into the words of Jean Rotz. They seem to suggest that his coastlines are real and in the right locations although they actually say that he had done his best with the most reliable information available. It remains to be seen whether her theory will be vindicated in some of the other coastlines of Java la Grande.
ALBERNAS II, Joao Teixeira (c.1680), "Demonstracao da Costa que corre das Ilhas de Pullo Botum, ate Malaca e parte da Ilha de Samatra" [map], in Mota, Avelino Teixeira da (1977), Cartas Portuguesas antigas na coleccao de Groote Schuur, Junta de Investigacoes Cientificas do Ultramar, Centro de Estudos de Cartografia Antiga, Lisbon, v.105, Fig. 16.
ANDERSON, J. (1826), Mission to the east coast of Sumatra in MDCCCXXIII, under direction of the Government of Prince of Wales Island; including historical and descriptive sketches, an account of the commerce, population, and the manners and customs of the inhabitants, and a visit to Batta cannibal states in the interior, William Blackwood, Edinburgh & T. Cadell, London.
ANONYMOUS (c.1742), "[Chart of the Strait of Malacca]", in de Roever, Arend & Brommer, Bea (2008), Grote atlas van de Verenigde Oost-Indische Compagnie / Comprehensive atlas of the Dutch East India Company, v.3, "Indische archipel en Oceanie / Malay Archipelago and Oceania", Asia Maior/Atlas Maior, Voorburg, sh.80.
ARMY MAP SERVICE (1944), Glossary of terms appearing on maps of the Netherlands East Indies, Corps of Engineers, Washington, DC. prov. ed.
BLAEU, Willem Jansz. & Hondius, Jodocus (1624), "[World map]" [map], in Schilder, Gunter (1977), The world map of1624 by Willem Jansz. Blaeu & Jodocus Hondius, Nico Israel, Amsterdam.
CORTESAO, Armando (ed.) (1944), The Suma Oriental of Tome Pires: an account of the East from the Red Sea to Japan, written in Malacca and India in 1512-1515; and, The Book of Francisco Rodrigues, rutter of a voyage in the Red Sea, nautical rules, almanack and maps written and drawn in the East before 1515, Hakluyt Society, London.
CORTESAO, Armando & MOTA, Avelino Teixeira da (1960), Portugaliae monumenta cartographica, Lisbon, 6v.
DERFELDEN VAN HINDERSTEIN, Gijsbert Franco von & Tndal, G.A. (1872), Kaart van Sumatra [map]. Scale: 1:2,000,000. Gebroeders van Cleef, The Hague. Royal Tropical Institute (KIT), Amsterdam, no. 05338, acc. No.542682. Available online at http://testbiblio.kit.nl/scripts/mwimain.dll/ 144/catalog_direct/biblioweb_long_smartsite/accession_number+ 542682?sessionsearch#
DESCELIERS, Pierre (1546), "[Mappemonde]" [map], in Coote, Charles Henry (1898), Autotype facsimiles of three mappemondes: with an introduction, including a short notice on Desceliers'later mappemonde of 1553, n.p., [Aberdeen].
--(1550), "[Mappemonde]" [map], in Coote, Charles Henry (1898), Autotype facsimiles of three mappemondes: with an introduction, including a short notice on Desceliers' later mappemonde of 1553, n.p.,[Aberdeen].
--(1553), "[Mappemonde]" [map], in Die Weltkarte des Pierre Desceliers von 1553: Im Auftrage der Geographischen Gesellschaft in Wien, herausgegeben von Eugen Oberhummer, Vienna, 1924.
--(attr.) (c.1555), [Chart of South East Asia and Java la Grande] [map], in Atlas attributed to Pierre Desceliers, f.11v-12. MS M.506 Morgan Library, NY.
DESLIENS, Nicholas (1541[c.1561]), "[Part of Mappemonde]" [map], in Hantzsch, Viktor & Schmidt, Ludwig (1903), Kartographische Denkmaler zur Entdeckungsgeschichte von Amerika, Asien, Australien und Afrika aus dem Besitz der Koniglichen offentlichen Bibliotekzu Dresden, Karl W. Hiersemann, Leipzig.
DUDLEY, Robert (1661), "Carta particolare del'Isole di Jaua maggre e minore con la parte Austrle del Isole di Sumatra e Burneo" [map], in: Arcano del mare di D. Ruberto Dudleo Duca de Nortumbria e Conte de Warvich, Florence, ed.2, v.II. , f.89.
ESSEN, Joan van (16??), Insula Sumatra [map]. National Library of Australia, MAP RM 1899. Available online at http://nla.gov.au/nla.map-rm 1899
FALCHETTA, P. (2006), Fra Mauro's World Map with a Commentary and Translations of the Inscriptions, Brepols, Turnhout.
FINDLAY, A.G. (1878), A directory for the navigation of the Indian Archipelago, China, and Japan, from the straits of Malacca and Sunda, and the passages east of Java. To Canton, Shanghai, the Yellow Sea, and Japan, with descriptions of the winds, monsoons, and currents, and general instructions for the various channels, harbours, etc., R.H. Laurie, London.
FROST, A. (2007), "Jave la Grande", in D. Buisseret, The Oxford Companion to World Exploration, Oxford University Press, Oxford, v.1, pp.422-423.
GARREN, W.R., PETERSON, B.D. & PAGE, C.R. (1982), Gazetteer of Indonesian names approved by the United States Board on Geographic Names, Defense Mapping Agency, Washington, DC.
GERRITZS, Hessel (1620), "[Chart of Sumatra]" [map], in de Roever, Arend & Brommer, Bea (2008), Grote atlas van de Verenigde Oost-Indische Compagnie / Comprehensive atlas of the Dutch East India Company, v.3, "Indische archipel en Oceanie / Malay Archipelago and Oceania", Asia Maior/Atlas Maior, Voorburg, sh.38.
[HARLEIAN MAPPEMONDE] (c.1547) [map], in Coote, Charles Henry (1898), Autotype facsimiles of three mappemondes: with an introduction, including a short notice on Desceliers'later mappemonde of1553, n.p., [Aberdeen].
HOMEM, Diogo (1568), "[India, South East Asia and China]" [map], plate XV in Hantzsch, Viktor & Schmidt, Ludwig (1903), Kartographische Denkmaler zur Entdeckungsgeschichte von Amerika, Asien, Australien und Afrika aus dem Besitz der Koniglichen offentlichen Bibliotek zu Dresden, Karl W. Hiersemann, Leipzig.
HORSBURGH, J. (1852), The India directory, or directions for sailing to or from the East Indies, China, Australia, and the interjacent ports of Africa and South America, v.2, ed.6, William H. Allen & Co: London.
HUGUET, E. (1925-1973), Dictionnaire de la langue francaise du seizieme siecle, 7 vols., Paris.
HYDROGRAPHIC OFFICE (Great Britain) (1942), StraatBanka [map], #3471, 1:250 000. Corrected to 5 December 1958.
--(1956), Selat Sunda and approaches [map], #2056, 1:200,000. Corrected to 1962.
JANSSON, Johannes (1649), Indiae orientalis nova descriptio [map], Amsterdam. National Library of Australia, MAP RM262. Available online at http://nla.gov.au/nla.map-rm262
JUNAIYAH, H.M. (2001), Kamus Bahasa Lampung-Indonesia, Balai Pustaka, Jakarta
KING, R.J. (2006), "The Jagiellonian Globe, Utopia and Australia", 400 years of Mapping Australia, Darwin conference, 23 to 25 August 2006: author papers & biographies, CD-ROM, Mapping Sciences Institute Australia, [Darwin].
--(2009), "The Jagiellonian Globe: a key to the puzzle of Jave la Grande", The Globe, 62:1-50.
KRAMER, A.L.N. & KOEN, W. (1993), Tuttle's concise Indonesian dictionary English-Indonesian IndonesianEnglish, Charles E. Tuttle Co., Rutland (Vt) & Tokyo.
LE TESTU, Guillaume (1556), La Cosmographie Universelle de Guillaume Le Testu, CD-ROM reproduction of original manuscript atlas held by Bibliotheque Nationale in Paris, produced in 2008 by Department des publics et de la valorisation, Service Historique de la Defence.
LODEWIJCKSZ, Willem M.A. (1598), "Nova tabula insularum Javae, Sumatrae, Borneonis et aliarum Malaccam ..." [map], Cornelis Claesz, Amsterdam, reissued 1617 by Claes Jansz Visscher, in Nebenzahl, K. (2004), Mapping the Silk Road and Beyond: 2000 years of exploring the East, Phaidon, London & New York, p. 87.
LUIS, Lazaro (1563), "[Chart of South East Asia]" [map], in Albuquerque, Luis de & Santos, Maria Catarina Henriques dos (1990), Atlas de Lazaro Luis 1563. Codice de Academia das Ciencias de Lisboa, Lisbon.
MARTINES, Joan (1587/1973), Atlas de Joan Martines, Ministerio de Educacion y Ciencia, Madrid.
MCINTYRE, K.G. (1977), The secret discovery of Australia: Portuguese ventures 250 years before Captain Cook, Souvenir Press, Medindie (SA).
--(1982), The secret discovery of Australia: Portuguese ventures 250 years before Captain Cook. Pan Books, Sydney.
MCKINNON, E.E. & SINAR, T.L. (1981), "A note on Pulau Kompei in Aru Bay, Northeastern Sumatra", Indonesia, 32(Oct):49-73.
MILNER, A.C., MCKINNON, E.E. & SINAR, T.L. (1978), "A note on Aru and Kota Cina", Indonesia, 26(Oct):1-42.
MORELAND, W.H. (1934), Peter Floris: his voyage to the East Indies in the Globe 1611-1615, The contemporary translation of his journal, Hakluyt Society, London.
[PASTORET ATLAS] (1587), "[Java la Grande]" [map] & "[Section of the East Indies]" [map], British Library Egerton MS 1513; both images reproduced as figs 2.17 (p.70) & 2.18 (p.71) in Wallis, Helen (1988a), "Java la Grande: the enigma of the Dieppe maps", in Glyndyr Williams & Alan Frost (eds), Terra Australis to Australia, Oxford University Press in assoc. with the Australian Academy of the Humanities, Melbourne.
PEARSON, M. (2005), Great Southern Land: the maritime exploration of Terra Australis, Dept. of Environment and Heritage, Canberra. Also available online at http://www.environment.gov.au/heritage/publications/ about/great-southern-land.html
RAMUSIO, Giovanni Battista (1554), 'Terza Tavola" [map], in Ramusio, G. B. (1606), Delle navigation et viaggi, v.1, Venice. National Library of Australia, RBq 910.4 R184.
--(1556), "Sumatra" [map], in Ramusio, G. B. (1606), Delle navigation et viaggi, v.3, Venice. National Library of Australia, RBq 910.4 R184.
RICHARDSON, W.A.R. (1983), "Is Jave-la-Grande Australia? The linguistic evidence concerning the west coast", The Globe, 19:9-46.
--(1984), "Java-la-Grande: a place name chart of its East Coast", The Great Circle, 6(1, April):1-23.
--(1989), The Portuguese discovery of Australia: fact or fiction?, Occasional Lecture Series No.3, National Library of Australia, Canberra.
--(2006), Was Australia charted before 1606? The Jave la Grande inscriptions, National Library of Australia, Canberra.
RODRIGUES, Francisco (1524-30), "[Ceylon, the Nicobar Islands and Strait Malacca]" [map], in Wallis, Helen & Sijmons, A.H. (1985), Atlas de Santarem: facsimile of the final edition, 1849, pl.68, no.15.
--(1524-30), "[NE Coast of Sumatra, Linga and Bangka islands & the NW coast of Java]" [map], in Wallis, Helen & Sijmons, A.H. (1985), Atlas de Santarem: facsimile of the final edition, 1849, pl.68, no.16.
ROTZ, Jean (1542), [South East Asia] [map]; and double-hemisphere world map [map],reproduced in Wallis, H. (ed.) (1981), The Maps and text of the Boke of idrography presented by Jean Rotz to Henry VIII: now in the British Library, n.p., Oxford.
SILVEIRA, L. (1991), Livro das plantas das fortalezas, cidades e povoacoes do Estado da India Oriental com as descricoes do maritimo dos reinos e provincias onde estao situadas e outros portos principais daquelas partes, Ministerio do Planeamento e da Administracao de Territorio, Secretaria de Estado da Ciencia e Tecnologia, Instituto de Investigacao Cientifica Tropical, Lisbon.
STEVENS, A.M. & SCHMIDGALL-TELLINGS, A.E. (2004), A comprehensive Indonesian-English dictionary, Ohio University Press, Athens (Oh.).
SUAREZ, T. (1999), Early mapping of Southeast Asia, Periplus, Singapore.
TAYLOR, J.L. (1970), Portuguese English Dictionary. Stamford University Press, Rio de Janeiro.
TDNI (Topographische Dienst in Nederlands Indie) (1924), "Overzichtskaart van Sumatra 1:250.000", blad 26, in Diessen, J.R. van et al. (2004), Grote Atlas van Nederlands Oost-Indie, Asia Maior/Atlas Maior, p.195.
TRICKETT, P. (2007), Beyond Capricorn: how Portuguese adventurers secretly discovered and mapped Australia and New Zealand 250 years before Captain Cook, East Street Publications, Bowden (SA).
TZANAKI, R. (2003) Mandeville's medieval audience: a study of the reception of the book of Sir John Mandeville (1371-1550), Ashgate, Aldershot.
UHDEN, R. (1939), "The oldest Portuguese original chart of the Indian ocean, A.D. 1509", Imago mundi, 3:7-11.
[VALLARDATLAS] (1547). folio 2, La Jave [map]; Terra Java [map], ff.2&3. Huntington Library, San Marino, Ca., HM29. Available online at http://dpg.lib.berkeley.edu/webdb/ dsheh/heh_brf?Description=&CallNumber=HM+29
VAN DER MEULEN, W.J. (1974), "Suvarnadvipa and the Chryse Chersonesos", Indonesia, 18(Oct.):1-40.
VIDAGO, J. (1953), "Glossary of Portuguese words used as components for topographical features and landmarks in early Portuguese cartography", Imago mundi, 10:45-49.
VIEGAS, Gaspar (c.1537), "[Map of Southeast Asia]" [map], chart 15 in both the Atlas of 26 charts, held by Biblioteca Riccardiana, Florence, and the Atlas of 24 charts held by the Archivio di Stato, Florence. Both are reproduced in Cortesao Armando & Mota, Avelino Teixeira da (1960), Portugaliae monumenta cartographica, Lisbon, v.I, pl.52.
WALLIS, H. (ed.) (1981), The Maps and text of the Boke of idrography presented by Jean Rotz to Henry VIII: now in the British Library, n.p., Oxford.
--(1988a), "Java la Grande: the enigma of the Dieppe maps", in Glyndyr Williams & Alan Frost (eds), Terra Australis to Australia, Oxford University Press in assoc. with the Australian Academy of the Humanities, Melbourne, pp.39-81.
--(1988b), "Did the Portuguese discover Australia?", History Today, 38(March):31-35.
--(1992), "Java-La-Grande and Australia?", The Globe, 37:I-12 - I-13.
WHEATLEY, P. (1954), "A curious feature on early maps of Malaya", Imago Mundi, 11(1):67-72.
WOLTERS, O.W. (1979), "A note on Sungsang village at the estuary of the Musi river in southeastern Sumatra: A reconsideration of the historical geography of the Palembang region", Indonesia, 27(Apr):33-50.
(1) Wallis was the editor of the facsimile edition published in 1981 of the atlas of Jean Rotz.
(2) According to Wallis (1988a, 40) the Channel port of Dieppe was "the great centre of French maritime activity" in the early sixteenth century.
(3) I have found two instances of Sumatra being mentioned as a possible source of the coastlines of Java la Grande but both are obvious mistakes. Wallis (1988b, 34) reported that Andrew Sharp had proposed western Java and Sumatra but he actually proposed western Java and Sumba. Suarez (1999, 158) described Java la Grande as "a large austral land comprising Sumatra and Java" but the discussion and the explicit references following these words relate to Sumbawa and Java.
(4) The original French text is, "Et ce au plus certain et vray quil ma este possible de faire, tant par mon experience propre que par la certaine experience de mes amys et compagnons navigateurs" (Wallis 1988a, 79 fn.40).
(5) The location of the Equator is correct in Ramusio's earlier map of Southeast Asia (published in 1554) and in other Dieppe maps such as those of Rotz, Desceliers and Desliens.
(6) An image of the Vallard atlas can be seen at http://dpg.lib.berkeley.edu/webdb/dsheh/heh_brf? Description=&CallNumbei^HM+29
(7) Not all of the Java la Grande inscriptions in the atlas of Le Testu are considered in the present study.
(8) Le Testu's inscription is not discussed further here. It seems to be an evolution of the neige names which, wherever they appear, apply not to a river but to an offshore island.
(9) Laban is a contraction of pelabuhan (anchorage, port) (Kramer & Koen 1993, 20 & 209).
(10) [adallas?] (Rotz 1542), andrelas (Vallard 1547), Adallos (Harleian c.1547, Desceliers 1546, Desceliers 1550, Desliens 1541 [c.1561]).
(11) Malay words are described in this paper as Indonesian where their definitions are supplied by a dictionary of the Indonesian language.
(12) Malay "b" is often transcribed in early, Portuguese-sourced, maps of the East Indies as p, v or u. The sound "ng" is often transcribed as ao which, if expanded by someone unaware of its origin, could produce "ano". Therefore the word could have evolved as follows: beting/bating > patao > patano and finally, if the "typical" Portuguese ending "o" were dropped, > patan.
(13) The sound of "m" may be described as a nasalised "b" and "qu" as an unvoiced "g".
(14) The word jau (cock) is also attested with the spellings gal and jal (Huguet 1925-1973, 4/252).
(15) In its ending Rameriqua resembles Mandarika and Mandalican, names that substitute for Jambi in, respectively, Ribero's world map of 1529 and Ramusio's map of Sumatra, and madalica in both Rotz (1542) and the Harleian world map (c.1547). Van der Meulen (1974, 37) proposed its origin in a Sanskrit term ahan-dharika (pillar of the day) and associated it with the Malay words hari (day) and batang ("staff, pillar, treetrunk,etc."), the same words that form the river-name Batang Hari that runs through Jambi. Rameriqua could perhaps have been another such river-name but this argument is less convincing than a derivation from beruga.
(16) The likeness of batang to beting and bating (the first element of Batindalem) is another potential source of confusion.
(17) An obvious error. Sometimes even the Portuguese were unsure of the phonetic value of "c".
(18) Le Testu (1556) has the inscription Riviere de perrles (river of pearls) on the western coastline where it could be a reinterpretation of baie perdue.
(19) Findlay and other authors of such pilot guides mostly follow the descriptions of Lieut. Rose following his 1822 survey.
(20) Compare the variations of the name of Porto Santo in the Madeira archipelago given in the La Valliere atlas (another cartographic product of Dieppe): porto saint, porto samto, porte samte, porte samto (Cortesao & da Mota 1960, 5/134).
(21) Both of the instances where the word is transcribed here as 'haure' follow Richardson's readings.
(22) McIntyre (1982, 60) published a freehand copy of the name of this island from Desceliers 1553 which he must have read as Is Sabare.
(23) The source of the Peruvian name is the range of hills now named Cerro Silla de Paita. The Spanish word silla (seat) refers to the distinctive saddle-like shape of its summit.
(24) McIntyre (1977, 143) declared, without explanation, that the word Sylle is a former name of Ceylon (Sri Lanka) although he identified the Java la Grande place to which it applies with Cape Leeuwin in Western Australia.
(25) Probably Dolak Sibayak (2172m) and Dolak Sinabung (2417m).
(26) Rio poderoso was a late fifteenth and early sixteenth century Portuguese name for the River Congo.
(27) The word canicolla is rich in ambivalent letters. The letter "c" could stand for "k" or "s" and what is read here as "n" could also be "u" or "v".
(28) Peter Floris, a Dutchman, was sailing in the English ship The Globe.
(29) A well-documented Rio formoso (beautiful river) on the southwestern Malay Peninsula is shown in European maps, including those of Dieppe where the name is spelled in various French forms based on fremose, from the early sixteenth century to the end of the nineteenth. Wheatley (1954, 71) identified it with Batu Pahat.
(30) Ampar could be a genuine form of the name rather than a misspelling. Compare Batu Ampar, the current name of the administrative district that straddles the lower reaches of Sungai Kampar.
(31) Liaca (for Siaca) is a spelling produced by a misreading of the long manuscript s.
(32) Teluk Batindalem, Teluk Beruga and Tanjung Jatigombol are not mentioned there. Their latitudes were estimated from various charts.
(33) Distance calculations were based on an average of ten degrees of latitude, giving 1.64 pixels to one modern nautical mile. This chart of the Vallard atlas, described by the Huntington Library as folio 3 "Terra Java (west coast of Australia?)", has a map scale of about 1:13 500 000 calculated from the latitude scale. Its very small scale may make it seem unlikely that the two distances, actual and mapped, could nearly coincide, nevertheless the calculation can be repeated and each repetition produces a similar result. Note that Rotz's regional chart of Southeast Asia, which depicts northern Java la Grande, has a scale of 1:9 000 000 according to Wallis (1981, 36b)- its scale is the largest of all depictions of Java la Grande.
(34) The degree of precision is equal to about twice the width of the line delineating the coast.
(35) Just such a separation of the northern and southern coastlines of eastern Sumatra appears in the charts of Francisco Rodrigues (1524-30).
Andrew Eliason 
 Andrew Eliason is a Petherick Reader at the National Library of Australia. He was born and raised near Shepparton, Victoria, was educated at the Australian National University and the University of Canberra, and was an officer of the Commonwealth Department of Transport (one of its many titles) and the Australian Maritime Safety Authority from 1981 until 2006. He is currently an independent researcher with interests in, among other things, maritime history, Indonesian history and the early mapping of Southeast Asia. Contact: email@example.com