“…order to go there onehad first to
approach the king of Utsthaladv
pa. Batak legend knows as Raja Uti who had a
boar’s head and lived on an island off the coastof Barus. All rulers, the
Singamangaraja as well as Tuanku Barus andthe king of Pagerruyung, had to offer
him presents.”


*W. J. van der Meulen

The title of Paul Wheatley!s book, The Golden Khersonese: Studiesin the historical geography of the Malay Peninsula before A.D. 1500(1961) does not leave any doubt about the authorτs confidence that theMalay Peninsula indeed represents the Golden
Peninsula which featuresin Ptolemy!s Geographike Hyphegesis(Manual for Mapping
out the Earth).1The text of his book conveys the same conviction, adding that,
thoughthe Indian Suvarnabh
ύmi or Suvarnadvΐ
pa(Gold Land or Gold Island)
in-dicated only some indistinct “eastern eldorado,” it was somehow espe-cially
connected with the Malay Peninsula, since it must have been thematrix of
hersonesos (peninsula).In spite of my admirationof the way in which
Wheatley brought together his rich source materials,for which any future
historian of Southeast Asia will be indebted tohim,I do not share his confidence
as to the location of the GoldenKhersonese.

In this paper I would like to present some of the reasonsfor my doubts.I.SuvarnadvίpaSuvarnabhumi and Suvannabhumi are the most common names connectedwith gold in old-Indian literature, both Sanskrit and Pali.The adjec-tive “gold” is also used in connection with other geographical fea-tures, but in Sanskrit works mostly in the compound Suvarnadvίpa.2Itis obvious that these names, as Wheatley rightly notes, only “featuredin early Indian folklore as an eastern eldorado where great richesmight be won.”3It must be doubted, however, whether this statementremains true when subsequently “ancient Indian folklore” is changed*This article is an extended and revised English version of a paper read in Indo-nesian at the “Seminar Sejarah Nasional II” in Yogyakarta (August, 1970).1. Ptolemy was a Greek mathematician and astronomer who worked in Alexandria in thesecond century A.D.

His Geographike must have been published between 150 and160.For this work he
depended greatly on an earlier one by Marinus of Tyre,who lived at the end of
the first and the beginning of the second century, andwhose work he tried to
rectify.Nothing is known of Marinus except what Ptolemytells us.See C. Muller
and C. T. Fisher (eds.), Claudius Ptolemaeus1″Geographia”: Selections (5 vols.;
Paris: A. Firmin Didot, 1883-1901); L. Renou,La Geographie de Ptolemee, I’Inde
(VII, 1-4) (Paris: E. Champion, 1925); J.Fischer and P. Franci de Cavalieri,
Clavdii Ptolemaei Geographiae Codex VrbinusGraecvs 82 (4 vols.; Leyde: E. J.
Brill, 1932).2. S. Lέ
vi, “Ptolemee, le Nidessa et la Brhatkatha,” Etudes As, II
(1925), pp. 1-55,431-32; R. Braddell, “Notes on Ancient’Times in Malaya I,”
Journal, MalayanBranch, Royal Asiatic Society (JMBRAS), XX (1947), pp. 161-86.3.
Wheatley, The Golden Khersonese (Kuala Lumpur: Oxford University Press, 1961),p.

Into “the ancient Indians. Poets and storytellers of course usedthese names
merely as an embellishment without caring about their exactlocation or whether
they were connected with any particular location atall. The names, however, did
not originate with these poets, but withpeople for whom they and their exact
location were of substantial im-portance, and who had, moreover, full
opportunity to know.Information about these places was not obtained from the
lonevoyage of a vesse-1 that managed to return home after being swept by astorm
to an unknown coast, whose sailors told fabulous stories that inno way could be
checked.There is an increasing conviction that thisinformation was the result of
both indirect and direct trade contactwhich began long before the Christian era
and became, at least from thebeginning of^ that era, a regular occurrence along
well-known traderoutes in which Indian shipping, though slower to start than its
Indo-nesian counterpart, played a prominent role.4A great deal of valuable
information about these voyages spread inshipping circles along the coast from
Bengal to Ceylon and from thereto Gujarat and Sindhu by way of a coastal sea
traffic which goes backto about the fourth century B.C.5This information as told
by Sindbadthe sailor for the consumption of stupefied landlubbers at home
mayhave been mixed with a liberal sprinkling of fairy tales.

Pilots,traders and their sponsors, however, whose lives and fortunes were atstake, were less given to spreading and receiving this kind of report.They sought after exact information and knew how to compare and to sifttheir sources.

(As to locations, even primitive sailors like the Poly-nesians showed an uncanny ability to estimate exact directions and dis-tances.)6Thus, though the Indian sailors were of course unable todefine locations according to modern geographical
coordinates, theywere presumably able to describe their relative positions in
terms ofdirections and normal sailing days.One might still object, however, that
these traders, who came fromdifferent parts of India and probably made their own
separate voyages,could hardly be expected to invent the same names for the
places theychanced to visit.For as far as can be judged from literary
sources(whose evidence may, however, be deceptive) they were not interested
inusing indigenous names, but invented their own, mostly in connectionwith the
principal product of the place (such as gold, camphor, coco-nuts, or barley) or
in commemoration of their own home country (Mala-4. D. G. E. Hall, A History of
Southeast Asia (3rd ed.; London: Macmillan, 1968), pp.12-24; G. Coedes, Les
£tats Hinduisms d’Tndochine et dfIndonesia (2nd ed.; Paris:E. de Boccardi,
1964), pp. 35-72; J. F. Cady, Southeast Asia. Its HistoricalDevelopment (New
York: McGraw-Hill, 1964), pp. 21-48; C. Nooteboom, “Sumatra ende zeevaart op de
Indische ocean,” Indonesie, IV (1950/51), pp. 120-27; 0. W.Wolters, Early
Indonesian Commerce. A Study of the Origins of Srί
vijaya (Ithaca:Cornell
University Press, 1967), pp. 21, 154;0. W. Wolters, The Fall of Srivij
ayainMalay History (Ithaca: Cornell University Press, 1970), p. 193; J. Innis
Miller,The Spice Trade of the Roman Empire 29 B.C. —A.D. 641 (Oxford: Oxford
UniversityPress, 1969), pp. 153-72.5. C. Maloney, “The Beginnings of
Civilization in South India,” The Journal of AsianStudies, XXIX (1969/70), pp.
603-16.6. R. C. Suggs, The Island Civilizations of Polynesia (New York: New
AmericanLibrary, 1960), especially chapter 7.

yadvipa, KalaSapura, Nagapura, Tondi). However, the weight of thisargument
depends on whether or not there were regular trade routes anda great deal of
communication at home and abroad. These voyages lastedfor many months or even
years. Traders transferred their goods fromone ship to another and together
awaited the return monsoon at meetingplaces—where they were more likely to find
lodging, safety, companion-ship and a vessel for the trip home.In these
circumstances, an under-standing of each other’s nomenclature and even a certain
uniformitywith regard to the principal aims and landmarks of their voyages,
suchas the “Gold Land11and “Gold Island,” was easily obtained. It alsoseems to
me that with regard to precisely these two features we havesome explicit
evidence to show that during the first century A.D. theywere already generally
known as two definite and distinct geographicalentities»During the reign of
Augustus, an increasing number of Greeks weretrading with the west coast of
India. They reached the east coast,probably over land, around the middle of the
first century.7One oftheir pilots, who collected his information in the second
half of thefirst century, but whose personal experience did not reach beyond
thewest coast, wrote a detailed guide for voyages around the Indian Ocean,the
Periplous.®His account of the exploration of the mouth of theGanges and beyond
is therefore probably based on information which hegathered in ports along the
northwest coast of India.9This informa-tion contained only vague indications
about the exact location of theplaces mentioned. The writer was told about a
mainland region called”Golden,” the most eastern continent toward the orient,
situatedaround, above, beyond (he peri auten} the Ganges mouth. Downwardsfrom,
opposite to or near the same river (kat’auton de ton potamon},however, and also
an extreme eastern part of the inhabited world, lyingexactly towards sunrise
(hup’auton aneohonta ton helion} , i.e., dueeast, lay an oceanic island of the
same name.Whatever the value of these indications as sailing directions, themost
interesting feature of this passage is the careful stress put onthe different
character of the two Chryses. The first is not simplycalled ohδ
ra (place,
region, country), as in the work of Ptolemy, butspecifically epeivos (mainland,
continent).In the same way, thewriter is not content to say simply nesos. This
Greek word would ordi-narily be sufficient, but it was sometimes used for a
peninsula (as,e.g., in Pelopponesos).In order to avoid any confusion he
adds”oceanic” (okeanikos} , an adjective which rules out the possibility ofa
peninsula, since it indicates exactly the opposite of cherso-nesoswhich means
literally “mainland island.” Ptolemy, who probably neverread the Periplous,
tried to avoid the same problem by explaining nesosas chersonesos, merely on the
strength of the so-called coastlines onhis map. Thus he created precisely the
confusion he wished to avoid.We will have to return to this question
presently.7. Sir Robert E. M. Wheeler, Rome beyond the Imperial Frontiers
(London: Bell,1954), pp. 141ff.; Max Gary and E. H. Warmington, The Ancient
Explorers (London:Meutheun, 1929), pp. 87ff. George Woodcock, The Greeks in
India (London: Faber,1966), pp. 136ff.; G. Juvea
-Dubreuil, “Les Ruines Romaines
de Pondichery,”Bulletin de I’Ecole fran
ais d!Extreme-Orient (BEFEQ), XL (1940),
pp. 448-73.8. Wheatley, The Golden Khersonese, pp. 129-30; R. Hennig (ed.),
Terrae Incognitae(2nd ed.; Leiden: E. J. Brill, 1944-56), pp. 335, 389-90.9. C.
Muller, Geographi Graeci Minores I (Paris: A. Firmin Didot, 1855), p. 285.

Page 4
If Suvarnadvipa (or whatever a Gold Island was called in the lan-guage they
used) for the pilots of the first century was an island inthe strict sense of
the word and a particular one, from the end of theseventh century onward there
is no doubt at all that this island wasSumatra.10The connection is accepted by
I-tsing, by the Nalanda in-scription and by Nepalese and Arab sources as
self-evident.11This is notcontradicted by the fact that Suvarndib-Zabaj as a
political entityexceeded the limits of Sumatra, as is suggested by
Wheatley.12Srivi-jaya, Majapahit and so many other kingdoms were not confined to
theboundaries of the geographical center that gave them their names. Thesame
identification with Sumatra is suggested in several stories of
gara (stories that could have originated well before theseventh
century), mainly those about the voyages of the brahminCandrasvamiu who went in
search of his lost son and of the princessGunavat
, whose ship was wrecked on
the coast of Suvarnadvipa while onits way from Kataha to India.13Finally, in the
thirteenth and four-teenth centuries, it was the kings of Dharm§sraya (Upper
Jambi) whostyled themselves lords of Suvarnadvipa and Kanakamedin
.1^This name
for Sumatra was well founded on its rich deposits ofgold and silver.It persisted
therefore during Portuguese, Dutch andEnglish times, as we will presently see.
The Malay Peninsula on thecontrary evidently had little to offer in this
field.In older Chinesereports gold is not mentioned except in connection with
Tan-tan1 5(which might help to show that its location was really outside
thePeninsula),16and in Chinese reports of the thirteenth and fourteenthcenturies
it is referred to only as an import commodity.17The Arabwriters who glorify
Zabaj for its gold do not mention it in connectionwith the Peninsula, except for
, who tells us that there aregold mines in Pahangπwhich are seldom
exploited.”18According toUrdaneta “an enormous quantity of gold, and that of
high quality” wasimported to Malaka from Minangkabau, while only a small
quantity, andthat of poor quality, came from Thailand and Patani.1910. Wolters,
Early Indonesian Commerce, p. 267; Coedes, Les Etats Hinduises, p. 160.11. N. J.
Krom, Hindoe-Javaansche Geschiendenis (2nd ed.; The Hague: M. Nijhoff,1931), pp.
142, 247.12. Wheatley, The Golden Khersonese, p. 182. Suvarnad
b is the Arab
equivalent ofSuvarnadvipa. Zabag/Zabaj is a port in Southeast Sumatra in the
Arab texts.Its identity will be discussed below at p. 38.13. Ibid., pp.
180,182.14. Krom, Hindoe-Javaansche Geschiendenis, pp. 336, 413.15. Wheatley,
The Golden Khersonese, p. 52.16. Wolters, Early Indonesian Commerce, pp.
206-16.17. Wheatley, The Golden Khersonese, pp. 74, 87.18. Ibid., p. 228.19. P.
A. Tiele, “De Europeeers in den Maleischen Archipel,” Bijdragen tot de
Taal-,Land- en Volkenkunde (BKI), XXIV (1877), p. 348; XXVI (1879), p. 27; M. A.
P.Meilink-Roelofsz, Asian Trade and European Influence in the Indonesian
Archi-pelago 1500-1630 (The Hague: M. Nijhoff, 1962), pp. 80-81.

Page 5
Thus, if Sumatra was accepted without question as “Suvarnadvipa”since the
seventh century, we may suppose that this represented a con-tinuing tradition
from long before that time, unless we have conclusiveevidence which points to a
change. Such evidence does not exist.Whatindications we have point on the
contrary to a continuity.As to theclaim of the Malay Peninsula in particular,
there is no positive evi-dence that it was ever, before or after the seventh
century, calleddvΐ
pa (as a whole) or suvarna.The latter appellation could
hardlyhave been founded on its deposits, but neither is there evidence thatit
was used in connection with the re-export of gold or with the gen-eral idea of
rich trade opportunities. There is seemingly one solitaryexception, whose
solitariness however makes it already suspect, viz.,the use by Ptolemy of the
word cherson
sos. The question is whetherthis creation was indeed based on
original reports or on some interpre-tation of Ptolemy himself.II. The Origin of
Ptolemy!s ChersonesosWheatley states that “In this century . . . writers have,
albeitin a rather hesitant manner, usually identified the Golden Khersonesewith
the Malay Peninsula.The reasons for this seem to me unassail-able.”20The first
of these reasons is that the coastlines recon-structed on the basis of Ptolemy!s
data and starting from the Gangesdelta, a point whose location cannot be
disputed, showπa generalagreement” with the present outline of mainland
Southeast Asia, whichis “too complete to be explained by coincidence alone. The
Bay ofBengal, the Burmese deltas, the Gulf of Martaban, the Malay Peninsula,the
Gulf of Siam, the rivers of Indo-China, all are clearly recogniz-able.” Thus he
reasons that in this general agreement the featurewhich tallies with the Golden
Khersonese is undoubtedly the MalayPeninsula.21The first thing I want to call in
question is whether the lineswe see on Ptolemaean maps or their reconstructions
can simply be puton a par with “outlines” or “coastlines.” Ptolemy’s catalogue
containslists of coastal features such as ports of call [empor
a’) ,
indigenoussettlements (poleis) , estuaries, capes, etc., supposedly
arrangedaccording to the order in which these were visited or encountered bythe
seafarer.These places were charted on the maps and connected bylines, and we
pretend (as Ptolemy himself did too) that the net resultis something like a
coastline. But even if the coordinates were moreor less trustworthy, these lines
would basically only represent sailingroutes, which could (and fairly often
would) run parallel to a coast.But since the sailors of those days were not
bound to coastal shipping,what guarantee do we have that these places were
always located alongthe same coast, that the sailing route did not cross from
mainland toisland, from island to island or from one mainland to another?
Whatassures us that those who told about their voyages mentioned everycrossing
and that those who noted down the sequence of places and theirdistances paid due
attention to taxonomic consistency?Thus, in par-ticular, they could have
mentioned an island and then places (on thatisland). Ptolemy, however, or more
probably his source, would havelifted this island out of context and placed it
with the others in aspecial “list of islands.” Thus no trace of crossings would
remain.20. Wheatley, The Golden Khersonese, pp. 144-55.21. Ibid., p. 145.
It is a pity that the original sources were melted down into themould of the big
Marino-Ptolemaean world-conception and so were lost tous. Some, however, may
have escaped this fate: for example, the ac-count of one Alexander, who stated
according to Marinus of Tyre thatbetween the Golden Cherson
sos and Zabai, a
distance of twenty days,the coast faced south, whence the course lay to the
eastward of southfor a voyage of some days until Kattigara (see Ptolemy!s
GeographyBook 1, Chapter 14, Section 1, hereafter 1.14.1). We do not know
whothis Alexander was, whether he visited the region in person and whetherhis
words were reported verbatim or only approximately (especially:did he use the
term chersonesos?).It is not clear either whether fromZabai onward the course
paralleled a coast or not. Therefore, if Pto-lemy came to the conclusion that it
did not (the course is not chartedon his map), this must have been due to other
contradictory reports,rather than to the account itself.This question is even
more pressing for a region abounding inislands.It seems hardly possible that
Ptolemy received reports aboutnine islands or island groups none of which lay on
the route of theregular traffic.22Nonetheless one of these was an “island of
Fortune11(Agathou Daimonos nesos) and another was the large island
pa,”fertile and producing a great quantity of gold,” which seemed to
be,together with Suvarnadv
pa, one of the main objectives of the wholemovement
to the east, and moreover one of the first kingdoms to be men-tioned in Chinese
sources under an Hinduized name.23If Indian litera-ture may be considered as a
vague echo of what was going on, it reflectsno urge comparable to this quest for
gold, which would warrant the sup-position that traders pressed on with great
eagerness along the main-land shore in order to reach the China coast and left
those islandsapart.It is fairly certain of course that in the Later Han period
trad-ers from the West reached Tongkin and Canton by ship and that occasion-ally
Chinese set out from these harbors to the West,24ProfessorWolters has shown,
however, that whatever trade was going on passedlargely through the hands of
Fu-nan and its dependencies, who con-trolled the contacts with southeastern
China both by land and by sea.25It stands to reason that Fu-nan would try to
maintain its favorableposition as an entrepot, if need be by force. Change came
between theend of the third and the beginning of the fifth century, when the
Chi-nese imperial center was forced to migrate to the south and to lookfor new
ways of making contact with the West, and when at the same timeit became
possible to bypass Fu-nan by crossing the South China Seadirectly from
ίpa, and afterwards from Srί
vijaya, to Canton.This emphasis on the role of
Fu-nan, c.q. the Gulf of Siam, as anentrepot is essential in order to understand
the course lined out by22. E. H. Bunbury, A History of Ancient Geography among
the Greeks and Romans fromthe Earliest Ages till the Fall of the Roman Empire (2
vols.; New York: DoverPublications, 1959), p. 608.23. Ying-shih Y
, Trade and
Expansion in Han China (Berkeley: University of Cali-fornia Press, 1967), p.
177.24. Ibid., pp. 172-87.25. Wolters, Early Indonesian Commerce, pp. 31-48.

Ptolemy.This course does not have to be “turned about” as is advo-cated, e.g.,
by Bunbury26and Hennig.27The traders did not turn northto Tongkin, but continued
from Rabana (.Raj Banam or Kurung Banam, Fu-nan) to the southeast across the
mouth of the Sinos River (South ChinaSea) and arrived, after passing Cape Notion
(Natuna?), at the AnimalGulf, viz., Datuk Bay at the northwest tip of Borneo.
Ptolemy givesthe Greek translation thevίδ
des , which tallies with Sanskrit
tiryagjaand Malay B
natang, presently still the name of a market center nearthe
mouth of .the Rajang River (in present day Sarawak), while at thecoast we still
find a Cape Jeriy
h (which may be a corruption of tir-yagja). Thus they set foot
on Yabadiou, whose capital, Argyr
polisor Silver Town, was situated on the most
western extremity of thecoast, where we still find the river and market center
of Selakau(selaka jneans silver), and whose name, Barley Island, probably
derivedfrom the Barley River or Sungai Jelai, which empties at Borneo!s
south-western corner, the first landfall on the above mentioned direct routefrom
Zabai to Kattigara. This name “Yavadv
ίpa” (next door to Javadvί
pa)of course
started to cause confusion long before our time.It did soever since Java
overtook its northern neighbor Borneo in importance.The earlier Javanese kings
even added to it by appropriating this cele-brated name in their Sanskrit
texts.Since this part of the voyage lies outside our present scope, amore
detailed discussion will have to wait. My only concern at themoment is to show
that the islands must have been part of the traderoute, which is one of the
reasons why we will have to reconsider thesignificance of the connecting lines
as merely indicating the sailingcourse. There is, however, still the important
difficulty that onPtolemy’s map no connection whatever is discernible between
the loca-tion of these islands and the location of the sailing route.We have to
note in the first place that Ptolemy, too, consideredhis lines to be coastlines,
and thus he could not very well put theseislands on top of a mainland, whatever
his sources might say. But inthe second place, the actual reason why these
islands were relegated tothe places they occupy may have been the quest for
symmetry. Ptolemycomplains that Marinus, who seems to have felt free to correct
hissources whenever they did not comply with some preconceived theory ofhis
own,28had pushed the Golden Peninsula much too far to the eastrelative to Cape
Kori, which they considered to be the southern tip ofIndia (1.14.7-8).It seems,
however, that he left the islands behind,either because this heightened the
symmetry of his composition, or be-cause there was no place for them on the
eastern continent he hadcreated.In any case, by taking our bearings from the
Gangetic delta andfrom the Golden Peninsula (whose general location–Sumatra or
the MalayPeninsula–may be considered at least approximately certain) we
canassure ourselves that no island is or in the historical period evercould have
been in the region where they are located by Ptolemy, what-ever dimension we
attribute to his degrees. Since they professedlybelong to “India beyond the
Ganges,” they will have to be moved atleast ten to fifteen degrees eastward (or
the Peninsula westward) in26. Bunbury, A History, p. 60627. Hennig, Terrae
Incognitae, p. 407.28. Bunbury, A History, p. 537.

10order to reach a tenable position in relation to the other data, un-less we
think, of course, that the sailors saw a whole series of mi-rages. We may note
that it is mainly their comparative longitudinalposition which is at fault.Thus
it would seem that there are no a priori guaranteed “coast-linesfτ
on Ptolemy!s
map of Southeast Asia. We may add that if therewere, their “general agreement”
with an existing coastline would provenothing.Such ap. argument could only have
some force if the coursewas continued for long stretches in the same direction
and under thesame conditions of wind, current, reefs, etc.The frequent and
drasticchanges of direction and conditions in the waters of Southeast
Asia,however, make this comparison as deceptive as the calculations of
Ber-thelot.^9Ptolemy’s averages, applied to such varying conditions,must,
through prolonging and dwarfing of the real distances, lead toconsiderable
distortion.The only element which may be consideredfairly reliable is the
indicated direction of the sailing course.When we come to the particulars as
summed up by Wheatley (the Bayof Bengal, the Burmese deltas, the Gulf of
Martaban, etc.), the grounddoes not become much firmer.In later times ships
bound for Suvarna-dv
pa did not bother with the Gulf of Martaban and the
overland route,but sailed south as fast and as far as they could before the
windturned against them.I-tsing, for example, made it as far as Kedah,where he
had to wait from “the first or secon^ month” until winterbefore he could go on.
arriving therefore in Sr
vijaya exactly a yearafter reaching Kedah.^° Ships that
went in the opposite direction toTamralipt
are reported to have sailed by way
of the Nicobar or AndamanIslands.It would seem, therefore, that Ptolemy’s
“Tamala” had nothingto do with Burma, but was more likely the port of Chetamala
(Sanskriteetamalaf) on Little Andaman, where ships crossing the Bay of
Bengalcould also make their landfall.From Chetamala they sailed for thenorthwest
coast of the Malay Peninsula to await the coming of thenortheast monsoon. This
accounts for the shortening of what is sup-posed to be the Andaman sea on the
Ptolemaean charts.On their way to the coast of the Peninsula they passed
Sahara.The Arabs knew a Naja Bara corresponding to Nicobar.3^The three namesmay
have been derived respectively from the Sanskrit oaura (modernChowra or Chaura),
nacaura and niscaura.The first island (maybe theChia-lan of the Chinese
mentioned by Wolters32) was apparently knownfor its piratical activity (Thieves
Island), while the rest of thegroup was considered relatively free of this
pest.We may add thatthe sounds represented by “c” and “v” in modern Sanskrit
transcriptionsare commonly transcribed in Ptolemyfs opus as sigma (e.g., Semula
=Cemula, Eragassa = Erakaccha)33and b§ta. After they crossed the29. A.
Berthelot, L’Asie Ancienne Centrale et Sud orientale d’Apres Ptolemee
(Paris:Payot, 1930), pp. 273ff.30. Wolters, Early Indonesian Commerce, pp.
227-28.31. Wheatley, The Golden Khersonese, p. 243. In an inscription of 1031
this samefeature is called Nakkavaram by the Tamils. Ibid., p. 201.32. Wolters,
Early Indonesian Commerce, pp. 189, 327.33. Renou, La Geographic, p. ix; Ph. L.
Eggermont, “The Murundas and the AncientTrade Route from Taxila to Ujjain,”
Journal of the Economic and Social Historyof the Orient, IX (1966).

11Sabarikos Kolpos or Gulf of Robbery (cauvika}, they disembarked at thetrading
post of “Besinga” or “Babysδ
nga.” This toponym may have beenderived from the
Sanskrit bahih simhala (Outer-Ceylon), a name which,for archeological and
historical reasons, could very well apply to thecoast of modern Takuapa.34Its
Malay name, Ujung Selang, conserved bythe Arabs as Urong Salah35and by the
Europeans as Junk Ceylon, seemsto denote it as a place to await the turn of the
monsoon.After the* monsoon turned, they took off from a locality called”Beroba”
or “B
δrabai” (maybe the Sanskrit virά
va} somewhere to thesouth. The important
thing is, however, that they did not continue inthe same direction, but made a
sharp turn back to the west, evidentlyaiπting at the northern tip of
Sumatra.Thus it seems that their trip around the Chryse Chersonesos wasthe trip
around Sumatra, around theMGold Island’1of the Indians andtheMoceanic island” of
the Periplous.But why did they not use theStrait of Malaka?It is unlikely that
this strait was then obstructedby a land bridge, as was supposed by many
competent Portuguese writersof the sixteenth and seventeenth century.It is
unlikely also that theancient mariners did not know about this thoroughfare. The
insistenceon the word “island11suggests that both the Strait of Malaka and
theSunda Strait were in existence and were known.It is evident, however,that
Marinus and Ptolemy were not aware of this fact, in the same wayas they remained
ignorant of the crossing to Chetamala, to the MalayPeninsula or to Borneo. For
it is difficult to accept that they simplyignored all indications of a
crossing.It must have been the organiza-tion of their sources with the islands
lifted out of the main text thatconcealed these indications.The logical
inference seems to be, there-fore, that they used mainly or even exclusively
written sources (maybeAlexanders included) and had, for this part of the world
at least,little or no oral information from eye-witnesses. There are other
in-dications that corroborate this inference, as we shall presently see.It is,
however, probable that they explained away one crossingthat was given to them,
though not in so many words. The Greek andLatin writers of the first century
knew about a “nesos” orffinsulaπconnected with gold, but never spoke about a
“chersonesos” or “penin-sula.π36Thus it is more plausible that Marinus
Ptolemy’s (Greekor Latin) sources too headed a certain paragraph with the
s nesou.” Because no crossing was mentioned, our authors hadthe
choice of positing one or explaining the word “nesos” as “cherso-n
sos,” a
meaning that it could indeed have. The first choice wouldhave marred the flow of
their “coastlines,11and would have caused dif-ficulties in charting.The second
choice was therefore obviously indi-cated. This seems at least more probable
than the supposition ofWheatley that dv
pa also meant “peninsula.”37However
loosely the word34. A. Lamb, “Miscellaneous Papers on Early Hindu and Buddhist
Settlement in North-ern Malaya and Southern Thailand,” Federation Museums
Journal, VI (1961), pp.48-55; Wheatley, The Golden Khersonese, pp. 195-97; M.
Collis, Siamese White(London: Faber and Faber, 1946), p. 238.35. Wheatley, The
Golden Khersonese, figure 41.36. H. Kern, “Java en het Gondeiland volgens de
oudste Berichten,” VerspreideGeschriften (The Hague: M. Nijhoff, 1913-36), V,
pp. 312-13; B
nbury, A History,p. 364; Wheatley, The Golden Khersonese, p.
127.37. Ibid., p. 182.

12was used, especially in esoteric and poetic literature, it is difficultto
accept that in the second century it specifically suggested a penin-sula.To fall
back for that conclusion on word origins seems as danger-ous as calling a
present-day tenant-farmer a villain.III. The Island of FortuneThe main reason
why the ancient mariners chose the way aroundSumatra must have been the fear of
pirates.The name “Robber GulfMapplied certainly also to the northern part and
very likely to thewhole of the Strait of Malaka. This fear is reflected in
Fa-Hsien?s:MThis sea is invested with pirates, to meet whom means certain
death.”The high seas had their own dangers, but those from pirates were
worse,especially in waters where it was difficult to outsail them. On thehigh
seas the expanse of the ocean might be boundless, the ship mightbe swept at the
mercy of the winds for days and weeks, while the crewdid not know where they
were and, if the sky was overcast, had no no-tion in what direction they were
driven.They might be tossed by hugewaves without means to help themselves and if
they were swept onto anunknown coast their chances of escape were slight
indeed.38The Straitof Malaka, on the other hand, was an ideal hunting place for
sea-robbers.It was full of uncharted and treacherous mud banks, notoriousfor its
transverse and unreliable winds, sudden squalls alternatingwith doldrums.Any
immobilized ship became the defenseless prey ofhordes of small indigenous craft,
as can be inferred from later de-scriptions .39It seems plausible therefore that
the frequency of the voyagearound Sumatra stood in direct relation to the
unsafety of the Straitof Malaka.Even the Dutch ships at the beginning of the
seventeenthcentury preferred to sail buyten om (around Sumatra) as long as
thePortuguese held sway in Malaka.40Tomέ
Pires states thatπBefore thechannel of
Malaka was discovered, those of Gujarat and Bombay used totrade with Java round
the south of Sumatra. . . . It is not a hundredyears since they gave up this
route.”41This change may have occurredafter Malaka re-established some order in
the Malaka Straits, probablyafter the middle of the fourteenth century, which
had been lost sincethe decline of the older Sumatra-based maritime empire a
century ear-lier.42It is thus a fair inference that in exactly the^same way
thevoyage through the Strait held little attraction before Sr
vijaya wasfirst
able to establish a certain supremacy during the seventh centuryand at least
curb unauthorized piracy.A second reason could have been that the earlier
voyages to a con-siderable degree were animated by the quest for gold. Deposits
of this38. Fa-Hsien, Record of Buddhist Kingdoms, translated by H. A. Giles
(Cambridge:University Press, 1923), pp. 110-11.39. Wheatley, The Golden
Khersonese, pp. 89-91.40. H. Kroeskamp, De Westkust en Minangkabau (1665-1668)
(Utrecht: Drukkerij Fa.Schotanus & Jens, 1931), p. 14.41. D. F. Lach and C.
Flaumenhaft (eds.), Asia on the Eve of Europe!s Expansion(Englewood Cliffs:
Prentice-Hall, 1965), p. 20.42. Meilink-Roelofsz, Asian Trade, pp. 18, 22-23.

13metal were to be found along the west coast of Sumatra. Voyagers couldhave
obtained gold by entering the east coast rivers, but this may havebeen too
dangerous as far as the more northern portions of the islandwere concerned,
which, according to all reports, were inhabited by asavage kind of people.43The
southeastern parts were different, as wewill see, and at the west side they
could trade off the coast.In his third argument (the second will be dealt with
presently),Wheatley maintains that “the designation!golden1agrees well with
whatwe know of the early economic importance of the Peninsula.” The reasonwhy it
is not an important source of gold today, is because the worldsupply is so much
bigger than it was in ancient times.”The associa-tion of the Peninsula with the
precious metal persisted into the seven-teenth century when Eredia described the
mines of Patani and Pahang,and we find it occupying an important place in the
accounts of eveneighteenth and nineteenth century writers.”44We have noted
alreadythat in fact the Peninsula’s association with gold was a poor one andthat
this quality was what persisted.It is no coincidence that thesame Eredia was one
of the staunchest proponents of the story of theland bridge between Sumatra and
the Peninsula, evidently because hesaw this as the only way to salvage Ptolemy!s
chersonesos.45As to the”economic importance of the Peninsula” in general, it was
true only ofthe northern portion, and involved only the fact that this area
formeda barrier for east-west traffic and had to be traversed to reach therich
lands beyond.What really persisted was the association of Sumatra with
pa. The Portuguese explorers were under royal injunction to searchfor
the “Gold Island11and reported back that this could only be SouthSumatra.46They
were barely settled in Malaka when they tried to con-tact Minangkabau by way of
the Sumatra rivers. Moreover, as early as1600 the Dutch sent van Caerden from
Banten to Pariaman, Tiku, Pasamanand Aceh.In 1601 van Heemskerk went to
“Monancabo” in order to obtaingold, swords and pepper. And in 1602 they were
followed by the English43. Wolters, Early Indonesian Commerce, p. 195.44.
Wheatley, The Golden Khersonese, p. 145.45. Manoel Godinho de Eredia was born in
Malaka from a Portuguese father and a Bugi-nese mother. After being educated at
the Jesuit College in Goa, he became oneof the most energetic and scientifically
interested Portuguese explorers andadministrators. He wrote his Informa
ao da
Aurea Chersoneso in 1599, whereinhe gave his own interpretation of Ptolemy’s
Chersonesos by postulating a landbridge between Cape Rachado (Malaya) and Pulau
Rupat (Sumatra) so that thesetwo territories formed a single unit. His
contemporary, Jo§o de Barros, alreadyin 1560 gave another interpretation:f!This
Chersoneso, where our city of Malacastands, seems to have been given the epithet
of Aurea because of the great quan-tities of gold carried there from Monancabo
and Barros (Barus) . . . (that is)why ancient geographers erroneously called the
island of Samatra the Chersoneso”(Decadas Da Asia de Joao de Barros [rev. ed.;
Lisbon: Na Regia Officina Typo-grafica, 1777-78], Decada II, Livro VI, Capitulos
i). He also looked for aloophole in order to justify Ptolemy’s “peninsula.π46.
G. Schurhammer, Die Zeitgen
ssischen Quellen zur Geschichte
Portugiesisch-Asiensund seiner Nachbarlander zur Zeit des HI. Franz Xaver
(1538-1552) (Rome: Insti-tut
m Historicum S.I., 1962), no. 586, 732, 1115a,
1276, 1295, 1821, 1822.

14under James Lancaster.47They had the assurance of van Linschoten:”Sumatra,
formerly Taprobana, is also called by some historians Cherso-neso Aurea . . .
and they say that in former times it must have beencontinuous with Malaka.This
island is extremely rich in mines ofgold, silver and other ores.1’48Many
disappointments notwithstanding,49the search persisted. Raffles, while in
Bangkahulu, wrote:”The Tiga-belas Country (Pagerruyung) has always been famous
for its produce ingold; indeed to Europeans it has been known as a gold country
alone.. . .t?5° Even the Jambi war of 1902 still had much to do with rumorsabout
rich gold deposits being exploited in the foothills of the BukitBarisan.51Gold
and silver deposits are found in the whole Bukit Barisanrange from Angkola to
Bangkahulu, with smaller deposits in Aceh andLampung. Large-scale mining,
however, is difficult because the veinsare small and scattered.Nevertheless the
Rejang Lebong mine aloneproduced about 30 million guilders worth of gold and
silver between1900 and 1910.In 1924 its produce (3,870 kg.of gold and 64,800
kg.of silver) was worth 4.2 million guilders.In 1939 some six or sevenEuropean
companies were still operative in Sumatra, though with a jointoutput of only
2,400 kg. of gold and 18,200 kg. of silver.52Thus itseems improbable that the
ancient Sumatrans migrated to the MalayPeninsula in search of gold53″in view of
Malaya as a source of goldfor the ancient and medieval world.”54The “Island of
Fortune,” the island-eldorado “where great richesmight be won,” was above all
Sumatra.It would seem, therefore, thatSuvarnadvipa or the Chrysg nesos leads a
double existence in the geog-raphy of Ptolemy, first made out to be a peninsula,
and, second, leftin the shape of an island but with an alternative name, the
“AgathouDaimonos nesos” or “Bonae fortunae insula” as the Latin translationshave
it. In Hellenistic Greek the “agathos daimδ
n” was the giver ofluck, prosperity
and wealth, corresponding to the Roman Fortuna.5547. H. Koeskamp, De Westkust,
pp. 14-16.48. J. H. van Linschoten, Itinerario, Voyage ofte Schipvaert van Jan
Huygen vanLinschoten naer oost ofte Portugaels Indien: 1579-1592, edited by H.
Kern (TheHague: M. Nijhoff, 1910), I, p. 74.49. E. Hesse, Gold-bergwerke in
Sumatra, 1680-1685 (The Hague: M. Nijhoff, 1931),pp. 62ff.50. Th. St. Raffles,
Memoir of the Life and Public Services of Sir Thomas StamfordRaffles, edited by
his Widow (2nd ed.; London: John Murray, 1835), I, pp. 406-7.51. H. H. van Kol,
Driemaal dwars door Sumatra en zwerftochen door Bali (Rotterdam:W. L. § J.
Brusse, 1914), pp. 136-39.52. Th. Ligthart, P. H
vig and D. A. Rinkes, De
Indische Bodem (Weltevreden: Volks-lectuur, 1926), pp. 146-54; Indisch Verslag
1940 (Weltevreden: Landsdrukkerij,1941), p. 302.53. Wheatley, The Golden
Khersonese, p. xxxii.54. Ibid., pp. 149-50.55. F. M. Cornford, From Religion to
Philosophy (New York: Harper, 1957), pp. 37-39, 98-99.

15Since “suvarna” means both gold and wealth or fortune, and since twosimilar
names are not known from Sanskrit literature, it seems reason-able to suppose
that both names are Greek translations of the sameSuvarnadvί
pa, the first one
with a peninsula-correction by Marinus orPtolemy.It is possible, of course,
though less likely, that these two menmade up the alternative name “Island of
Fortune.” If they had severalsources mentioning aMGold Island/1one describing a
place-to-placevoyage starting from the Ganges delta, and another perhaps based
on adirect crossing of the Bay of Bengal and enumerating the islands thatwere to
be found en route, adding to its “Gold Island11the explanation”or the island of
wealth,” they could have regarded this as an alterna-tive and,more appropriate
name, in order to distinguish it from thegenuineMGold Island11described in their
main source, which accordingto their calculations occupied a totally different
place and besideswas evidently a peninsula.Explanations were also attached to
thenames of labadiou, the Satyroi and Maniolai islands (7.2.30-31) andothers,
and the explanation given here would have been an enormous stepforward compared
with those given by Pomponius Mela!s and Pliny’ssources, which said respectively
that the name meant that the islandwas paved with gold and that it consisted
entirely of gold mines.The only way to get any confirmation for the
identification of theIsland of Fortune with Sumatra, would be to attempt to
identify otheritems of Ptolemy’s island constellation in its vicinity and to
seewhether their respective locations produce a picture that
sufficientlyharmonizes with the distribution of islands on a modern map.
Othertoponymical data are of course of no avail, because, as we argued,
theislands became dislocated from the rest in Ptolemy’s account. Ptolemygives
for this Fortuna Island only one set of coordinates without anyexplanation as to
what part of the island is indicated. We may supposetherefore (not a new
supposition incidentally) that it refers to thefirst approach to the island,
viz., its northern tip.If this is correct, we will have to find a chain of
islands to thesouth and somewhat to the west of this point (the exact Ptolemaean
co-ordinates are 145 east and 4.15 north against 142 east and 2 south),called
the ten Maniolai Islands. The story told about these islandswas that they were
magnetic and clutched at ships that were constructedwith iron nails (7.2.31).
This was a precious piece of information andsome scholars have looked as far
afield as Mt. Tambora in East Indone-sia in search of this magnetism.It seems
more likely, however, thatthe magnetism was contained in the name and not in the
rocks. VonHumboldt56already suspected a connection between the name Maniolai
andthe Sanskrit mani, which means both pearl and magnet. The originaltoponym was
thus probably manim
, a chain of magnets, meant as a punon the Niassic toponym
la” (my rendering of its pronunciationby students from Nias), which is
still the name of a part of SouthNias57and was in ancient times apparently given
by extension to thewhole chain of islands west of Sumatra of which Nias forms a
part.56. W. von Humboldt, Ueber die Kawi-sprache auf der Insel Java (3 vols.;
Berlin:Druckerei der k
niglichen Akademie der Wissenschaften, 1836-39), p.
31.57. W. L. Steinhart, Niassche teksten (Batavia: Koninklijk Bataviaasch
Genootschapvan K
nsten en Wetenschappen, 1937), p. 7.

16To the east and farther south (at 152 east and 5.20 south) and duenorth of the
Sindai islands (152.20 east and 8,20 south) we have tolocate the Baroussai
Islands.According to these coordinates theBangka-Belitung group would be the
most likely candidate.Is thereany connection between the name Baroussai and a
toponym on one of theseislands?I considered Batu Rusa on Bangka, but rejected
this possibil-ity because there are probably very few indigenous names (and
onlyprominent ones such as Malayu) at the base of Ptolemyfs nomenclature.All the
names we* have met seemed to be derivates from Sanskrit wordsand I think that
this is the general rule.Before applying this theoryto Bangka, however, I will
try to give it a wider foundation.IV. Ptolemy’s NomenclatureSome accounts like
that of Gary and Warmington58create the im-pression that the Greeks discovered
the trade winds across the Gulf ofBengal and sailed to the East in considerable
numbers. For theseauthors, the Greeks might just have well have been the only
peoplepresent. Archeological evidence of their presence in Southeast
Asia,however, is actually very scarce and the Chinese reported this
presenceevidently as a curiosity.They evidently did not build up a trade oftheir
own with their own ships, entrepots, nomenclature, etc. It isfar more likely
that they joined as traders in the existing traffic andmixed with the Indian
(and probably partly Malay) crews, with tradersand other passengers, trying to
communicate in whatever language wascommonly spoken, for interpreters who knew
Greek must have been veryscarce. Afterwards they told their story, if the good
n gave themhis protection and they returned safely home.If Ptolemy!s
Geography was based on their reports, it may strikeus as curious that a
considerable number of the names they rememberedas being used in this maritime
commerce were Sanskrit.Yet it seemsundeniable that many names are just
έs59already had a gen-eral feeling that “la nomenclature gέ
pour l!Inde transgang
έ-tique, est dέja remplie de toponymes δ
sanskrite.” Egger-mont feels sure “that Pliny as well as Ptolemy availed
themselves ofexcerpts from an extensive Greek text that showed Sanskrit words
accom-panied by their Greek translations,”60a remark that is true in rela-tion
to Ptolemy’s basic text for both India and “India beyond theGanges” (Southeast
Asia).For the region beyond the Ganges we have three instances where theGreek
translation is placed alongside the original Sanskrit toponym.

Precisely because “labadious” was said to mean “Krithes Nβsos” or Bar-ley Island (7.2.29) scholars were able to recognize this toponym as acorruption of the Sanskrit yavadvΐpa (or maybe the hybrid yavadΐvu}.In the same way it seems evident that Nangalogai, translated as “landof the naked” (7.2.18), and Aginatai, said to mean “naked people,” goback respectively to the Sanskrit nagna-loka and nagnata (naked mendi-cant).

In addition, the story about the Maniolai contains the Greektranslation of the word manimalδ. Eggermont (I.e.) also gives “Kakobai”58. Gary and Warmington, The Ancient Explorers, pp. 104-5.59. Coedes, Les Etats, p. 43, note 2.60. ggermont, “TheMurundas,” p.277.

17(7.2.19) as a derivation of the Sanskrit kakubh (summit) in connectionwith
Pliny’s translation “Capitalia.” We may presume therefore thatthe nine other
Greek translations in Ptolemy’s text, given without theoriginal, go back on
Sanskrit words. This is fairly certain in thecase of Chrysβ Chersonδ
sos (7.2.5
ff.) and probably also for “NotionAkron” and “Th
riodgs Kolpos11(7.3.2) as we
showed above.With regard to those appellations that are not translations, wecan
point to a few, besides cetamala, caura and bahih-simhala, that areobviously
also Sanskrit derivatives.Thus the “Maiandros” mountain(7.2.8), that took the
name of a Greek river, is not a translation, buta corruption of Mahendra, a
Sanskrit name for various ranges. Anothermountain range is called “S
(7.2.8) because it forms the bor-der (sema} with Thinai or Cina (Semacaina).
“Aganagara” (7.2.7) ispure Sanskrit for “Mountain City.” The “Dorias” river
(7.2.11) wasobviously “connected with a gate {dvarya}, probably the same gate
thatwas afterwards guarded by Dvaravat
(the Portress or Gate-City), viz.,the
Three Pagoda Pass.The “Bassanarai” (7.2.19) were the “aromatic-raisins-people”
sanara), who were, according to Eggermont, calledBasatai by the Periplous and
Nareae by Pliny.”Kirradia” (7.2.16) wasobviously inhabited by notorious robbers
αtα).61An inland townin the Chrysg Chersonβ
sos is aptly called
“Kongkonagara” (7.2.25), thatis Kanakanagara or Gold Town.These few examples do
not yet prove that all Ptolemy’s names arederived from (meaningful) Sanskrit
appellations. They prove, however,the existence of a phenomenon which would seem
to raise several ques-tions.

Is it possible that in the first and second centuries Sanskritnames were already in regular use among traders, sailors and advan-turers, and that some of these names persisted until later times, evenup to the present? Who were the authors of these names and how did thenames get accepted by the common local traders so as to allow theirGreek colleagues to present them as “the” names they used to hear?Were there so many literati among the people who went east that theycould dominate the traders’ usage?

For it is of course impossible toexplain the whole phenomenon without inferring a dominant influenceexerted by people who knew Sanskrit.It is indeed likely that it did not take long for Buddhist mis-sionaries and scions of the higher castes to interest themselves andto take part in the trade voyages to the East.Interest in these voy-ages is in any case already evident in Indian literature of the
firstcenturies A.D. It may have stimulated various kinds of enterprisingclerics,
adventurers, and fugitives from justice or paternal authority(we have already
seen Candrasvamin frantically looking for his vanishedson— see p. 4), to whom
this “colonial” world offered a wide range ofopportunities and the prospect of
fabulous riches.When the king of”Yeh-tiao” (Yap-div) sent an envoy to China in
131 A.D., he certainlydid not use this name at the instigation of his native
sorcerer. Un-fortunately, we cannot be certain this Chinese phonogram is
derivedfrom the Sanskrit Yavadvipa.It could have originated from somePrakrit
dialect the sailors used, and was transmitted by them to theChinese. We can only
note that it would be a rather unique case of61. Ibid., pp. 280-81, 287.

18Prakrit influence in Southeast Asia.62Nevertheless, it seems very unlikely
that such an array of San-skrit names as are suggested by the Geography could
have been in gen-eral use in Southeast Asia at that early time. Do we have,
therefore,to refer the text of Ptolemy to a subsequent age? The arguments
thatare put forward in support of the conventional date (shortly after 150A.D.)
of Ptolemy’s opus as far as India beyond the Ganges is concerned,e.g., by
Jack-Hinton63and Wolters,61* seem fairly conclusive. Thefact, moreover, that
Ptolemy’s text is ignorant of a direct crossingof the Chinese Sea from south to
north and leads us only as far asOc-eo, points to the same conclusion.What other
kind of explanation,then, can be proposed?Earlier we pointed out that certain
confused ideas of Marinus andPtolemy suggested that they did not have oral
reports by eyewitnesses,but relied mainly, if not exclusively, on written
texts.(In any case,the catalogue of names could not, as it is, have been
reported frommemory.) These written texts would have contained not only
“Sanskritwords accompanied by their Greek translations,”65but also
untranslatedSanskrit appellations. One possibility is that they were
actuallytranslations of an original Sanskrit text, wherein names were
sometimestranslated or explained.Supposing that such Sanskrit texts once
didexist, they could easily have found their way to the Mediterraneanthrough the
mediation of Greek traders or of Indian traders and lite-rati, of whom a good
many were settled in Syria and Egypt at the time.66These literati would have
been able to provide Greek translations andtransliterations. Though the names in
these manuscripts would havereferred to existing places, their use (in Sanskrit
form) must havebeen still limited to a small circle of literati, except maybe in
thecase of names denoting a few much-frequented centers. A number of theothers
may have spread and obtained recognition and official status inthe course of the
Hinduization of Southeast Asia centuries later (inthe same or variant forms).
Others again fell into oblivion with thedecline of the centers that they
indicated.It is in any case veryunlikely that the names used by Ptolemy could
have been picked up byGreek traders from their Indian colleagues.These Indian
traders mighthave recognized some of the Sanskrit names used in Ptolemy’s
itinerary,viz., those of the general meeting-places mentioned above where
resi-dent clerics might have resided.The authors of such Sanskrit
manuscripts–lists of names to whichwere added the distances between the places
they indicated–could havebeen those clerics who joined in the voyages and
composed a diary or62. Sanskrit is the religious-literary language of Brahmins
and partly of Buddhists.Prakrit is a collective noun for the diverse popular
dialects of India. K.Bhattacharya, “Recherches sur le vocabulaire des
inscriptions sanskrites deCambodge,” BEFEQ, LII (1964), p. 6.63. C. Jack-Hinton,
“Marco Polo in Southeast Asia,” Journal of Southeast AsianHistory, V, No. 2
(1964), pp. 69-70.64. Wolters, Early Indonesian Commerce, pp. 57, 277.65.
Eggermont, “The Murundas.”66. Woodcock, Greeks in India, pp. 156-58.

19itinerary in Sanskrit, either during their voyage or after they settledin some
kampung Keling or returned home. Fa-Hsien!s diary of his vis-its to the holy
places in India may be cited as an example, for want ofan Indian counterpart
that escaped destruction.On the other hand, theauthors could have been Indian
colleagues of Ptolemy, who stayed homeand collated reports from travelers,
translating them, names and all,into Sanskrit, if need be with the help of
interpreters.In that caseit would have been necessary, however, that the
manuscripts were knownto Brahmins of Buddhist monks, who took them along when
they wentabroad and settled in diverse centers of Southeast Asia. Thus the useof
these names could have been promoted centuries after they were re-corded in
Ptolemy’s Geography.There is, however, still another possibility, namely that
thebasic manuscripts were the work of Greek travelers, who composed a com-plete
or partial Periplous of the far-eastern seas.In that case itwould have to be
postulated that they made the voyage in the companyof literati who supplied them
with a Sanskrit version of the names orthat they met with these people in
certain centers. This also wouldimply, however, that some literate tradition of
geography was alreadyin the process of emerging. But whatever the origin of the
more orless extensive texts that traveled to the West in the course of thefirst
and second centuries (though Marinus or Ptolemy probably usedone rather complete
principal source), the only trace that remains ofthem is the Geography itself,
which seems inexplicable without somesuch substratum.If our supposition is true,
we may wonder what are the chances ofreconstructing the Ptolemaean odyssey on
the base of the names that arementioned. Regarding those appellations, that at
the time or later be-came regular names which were used also by other sources or
are evennow extant, the difficulty is not more or less than elsewhere. Forthose
few appellations that became real names, i.e., were universallyused and recorded
in sources of diverse provenance (Southeast Asian,Chinese, Arab and others), the
chance that they can be recognized andlocated is in any case there, especially
if the names, albeit in cor-rupted form, are still used in loco.But what about
the others? Maybethe great majority? Must any search for them a priori end in
failure?This seems to me by no means certain, since their choice was by nomeans
arbitrary.Some of the examples whose original form and meaningcan be suggested
were apparently connected with the products of theplace (gold, copper, silk),
with the condition or occupation of theirinhabitants (naked, robbers,
raisin-gatherers) or topographical posi-tion (at a boundary, on a hill, coming
from a pass). Such indicationscan become relevant as elements of a series: i.e.,
if a certain plausi-ble location seems corroborated because neighboring elements
fit inwith this location, the more so if other sources also point in the
samedirection.Other examples, however, seemed to have a direct connection
withthe sound or the meaning of an indigenous name.”With the sound11doesnot mean
that the Ptolemaean toponym is simply a more or less success-ful rendition of
the indigenous name, such as can be found in Chinesetexts or on Portuguese or
Dutch charts. It means that an attempt wasmade to link the sound of the
indigenous name to a Sanskrit word bymaking it the subject of a wordplay,
sometimes accompanied by an im-plausible yarn. An extensive example of this kind
of game is the

20(Tamil) Tanjore inscription of 1030.6 7Another example might be thename Kedah
whose sound is “translated1 1into the Sanskrit Kataha andTamil Kadaram.Among the
Ptolemaean toponyms, we met with such a word-play on the Miassic name Maniemδ
accompanied by a yarn which onlymade sense in the framework of this
wordplay.Still another example,made more difficult to decipher because Ptolemy
gives only the Greektranslation of the name, is the Satyroi Islands (7.2.30) and
Cape(7.3.2).The denizens of the islands are supposed to grow tails,Pto-lemy
reports, and- are said to have the appearance of satyrs.Theirappearance has
nothing to do with that of apes, except that both havetails and both live in the
woods.Now satyrs were imagined by theGreeks as having the feet, ears and tail of
horses or jackasses. Fur-thermore they “played the ass,Mliked tomfoolery,
tricks, drunken boutsand laziness*6 8The Sanskrit words for ass and jackass are
khara andkhar
.In India, too, these animals are thought of as “playing theass”
(kharaya, khavay
ta} and of performing mad (matta] tricks. Bio-logical
abnormalities or folkloristic performances that conform to someof the above
“satyric” qualities are easier to find than magnetic rocks,Nevertheless, we have
to remember that we are not dealing with a sea-man^ yarn, but rather with one
composed by the “literati.” The storythen may well be simply an elaborate
wordplay on the name Karimata(khar
-matta}, an island off the southwest corner
of Borneo.We have also met, however, with instances where the Sanskrit
ap-pellation seems connected, not with the sound, but with the meaning ofthe
indigenous name.Such instances were, e.g., Yava/Jelai, Argyre/(R
h and Raj(Banam)/Kurung(Banam).The practice of
translating indigenous names into San-skrit certainly existed in Hinduized
Southeast Asia.The practice isknown from Fu-nan.6 9The Raktamrttika (Red Land)
inscription of Kedahmay refer to an indigenous Tanah’Merah or Tanah Abang.7 0The
samepractice is found in Java, not only in the time of Majapahit, but alsomuch
earlier.7 1We shall thus have to reckon with the possibility ofits application
by Ptolemyfs sources.It may appear that only Malay names (or at least those that
ad-mitted of a Malay explanation, which of course says nothing about thereal
etymological origin) and a few of Khmer origin (such as kurung-*)were
translated.This could mean that Malay was already starting tobecome the “lingua
franca” for Southeast Asian commerce, so that to67. See, e.g., Wheatley, The
Golden Khersonese, pp. 199-201. The Tanjore inscrip-tion proclaims the victories
of the Cola-king Rajendra in various parts of Su-matra and Malaya. Each name is
accompanied by a wordplay based on its sound.So Talaittakk
lamis said to be
praised by scholars, because in Tamil “talai”means science andMtakkorffscholar;
Malayur has a strong mountain for its ram-part, because “malai” means mountain
in Tamil and “ur” a stronghold, etc. Theseapplied meanings based on similarity
in sound with Tamil words have of coursenothing to do with the real meaning of
the originals.68. Oxford Classical Dictionary (1961), p. 797.69. Coedes, Les
Etats, pp. 74-75.70. Wheatley, The Golden Khersonese, p. 33.71. W. J. van der
Meulen, “Keradjaan!Ho-lingf, Tjarita Parahyangan dan RahyangSandjaja,” Basis, XV
(1965-66), pp. 41,166.

21some degree the maxim of van Linschoten:”Whoever in India cannotspeak Malay is
not allowed to be of the party1172was already operative.It is in any case
reasonable to suppose that the Indians recruitedMalays (or half-castes) as
guides and interpreters, and even as membersof their crews.73Besides, the kind
of people who composed these itin-eraries might be expected to have had the
intellectual curiosity to in-quire about languages and the names during their
long voyages and stops.In view of the nature of Ptolemy’s toponyms, it seems
essential toadhere to the serial pattern laid down in the text. No
identification,whether of the name or especially of its location, can be
accepted out-side the context of a sufficiently continuous and connected series
ofplaces of which the names and positions fit the pattern of names andcourse
directions in the text (distances are of not much use and areeven apt to
confuse) as well as the geographical and historical pecu-liarities of a
particular region. There is no certainty that a toponymever existed in loco in
that form.If every toponym had been trans-lated into Greek, we would have to
search continuously for a plausibleconnection of its meaning with some locality
it could indicate, and nosimilarity of sound would help us.(Fortunately only a
few names weretranslated.)But the Sanskrit names were also superimposed (maybe
astranslations) and are only partly better–in that there is a chancethat they
later became accepted local names. When and where this wasthe case, however,
cannot be determined a priori or by means of othersources, though the latter
are, of course, indispensable for confirma-tory evidence.Thus single
identifications, based on similarity ofsound with extant names or those found in
other sources, or on a simi-larity of geographical configurations, have proved
to be as deceptiveand disputable as the suppositions on which they were
based.The limited number of names from India Cisgangetica that can
beidentified74does not provide us with much guidance about the way inwhich the
transliteration of Sanskrit names into Greek script was ef-fected.Centuries of
copying mutilated nearly all names, especiallythe more intricate ones. But there
may also not have been much uni-formity from the start. Thus a Greek “zSta” may
stand for a SanskritMJM(Ozoamis/Ujjayinί) or “S” (Zadaros/Satadrϋ
; in the
Periplous,Muziris/MuSiri). The Sanskrit “C” is commonly transliterated as
mula; Erakaccha/Eragassa; Ahicchattra/Adisadra; in
Arrianus,75Candragupta/Sandrakottos or Sandrakyptosin Pliny, Pracya/Prasii)
andonce as tau-iota (Castana/Tiastanes). The Greek omikron (or omega) maystand
for “0” (Goaris/GodSvari), “A11(Prapi
tai/Paripatra), “A” (Modura/Mathura;
Dosara/Dastrna; Kognandaua/Kakan
va), “AV” orMVA” (Aroades/Airavati;
Ostha/Vatsaj, “0” (Soraisangga/S
ύrasena) and “U” (Ozoamis/Ujjayinί
). Thus there
is more possibility than regularity.The reasonwhy some of the names were
translated into Greek is also far from evi-dent.It may be that Ptolemy’s main
source gave no translations atall, but that he found some in other
manuscripts.72. Van Linschoten, Itinerario, p. 74.73. F. von Richthofen,
ber algemeine Siedlungs- und Verkehrsgeographie(Berlin: D. Reimer,
1908), p. 25.74. Renou, La Geographie, p. ix; Eggermont, “The Murundas,11pp.
257-96.75. A Hellenistic scholar (c. 95-175 A.D.) who wrote a history of
Alexander theGreat’s exploits.

22It will be evident from all that has been said so far that I donot share
Wheatley’s confidence in his second argument, where, withmuch assurance, he
locates two of Ptolemy’s toponyms on the MalayPeninsula.76This is done on the
base offfthe combined testimony ofreferences in early Chinese, Indian and Arab
accounts.” The subse-quent text is, however, less explicit. Moreover, this
combined testi-mony, if it existed, would be of no avail as long asMThe whole
patternof settlement as set forth in the Geography is strangely anomalous inthe
historical geography of Malaya.”77In his Early Indonesian Commerce, Wolters
identifies the fiveBaroussai Islands with the place Barus and with Varusaka (the
regionof Barus) on the basis of the resemblance in sound and a general opin-ion
about the respective location of the diverse parts of Ptolemyfsgeographical
structure.78That Barus-Varusaka is a port or region andthe Baroussai are
described as a group of islands is not so important.Any little known coastal
region was liable to be viewed as broken upinto islands, as in the case of the
Sindai and Saba(dί
bai) islands.

Similarity of sound may be an indication, the more so because in thisinstance the name Barus does not have an obvious meaning.Thereforethe unknown authors of the Sanskrit source might have looked for a San-skrit sound-synonym as in the case of the Maniolai and Karimata Islands,The position of Ptolemy’s Baroussai Islands relative to Fortuna Islandand the Maniolai, however, does not fit with Barus. For one thing, itis too far south, and the Barus-position would entail the necessity ofdrowning some interesting islands that are supposed to be to the westand south of it in the Indian Ocean.

The Sanskrit word nearest insound to Ptolemyfs Baroussai is paruςa
(once substituted by L
έvi forVarusaka on the basis of its Chinese
transcription79). This word means”stiff,” “rough,” “knotty,” and is the exact
equivalent of the Malaybangka or bangkar.Though we already noted a few examples
of plausibletranslations of indigenous names, we may still consider this
similarityof meaning a curious coincidence, and leave it at that until the
phe-nomenon in conjunction with other data becomes too much of a coinci-dence
.V. The Voyage along the West CoastAccounts of and directions for navigation
along the west coast ofSumatra only became plentiful in the West from the
seventeenth centuryonward.From the beginning of that century Dutch, English and
Frenchships visited up and down the coast.The Dutch came from Banten andBatavia,
or made a stopover on their voyage from the Cape of Good Hopeand sometimes from
India.80From before that time we have only frag-mentary evidence.Pate Unus
(Sabrang Lor of Demak) led his fleetaround Sumatra (1512) in order to spring a
surprise attack on the76. Wheatley, The Golden Khersonese, p. 145.77. Ibid., p.
159.78. Wolters, Early Indonesian Commerce, pp. 186, 326.79. Ibid., p. 326.80.
Koeskamp, De Westkust, pp. 14-16, 26-27, 132; E. B. Kielstra, “Onze kennis
vanSumatrafs Westkust omstreeks de helft de 18e eeuw,” BKI, XXXVI (1887), pp.
499-560; BKI, III (1855), pp. 106-141. See also Geographical Handbook Series
NavalIntelligence Division, Netherlands East Indies (2 vols.; Washington: Office
ofNaval Intelligence, 1944), pp. 110-11; and Encyclopedie van Nederlandsch
Indie(2nd ed.; 8 vols.; The Hague: Nijhoff, 1917-39), pp. 172-73.

23Portuguese in Malaka.81A Chinese map in the Wu-pei Chih>probablybased on the
voyages of ChSng Ho early in the fifteenth century, listsa good many places
along the coast.82We noted already the statementof Tomδ Pires that people from
Gujarat used to trade around Sumatrauntil the end of the fourteenth century.I
ventured the opinion above that, apart from special objectives,the choice of
this route was generally influenced by the degree of se-curity which existed in
the Straits of Malaka.Foremost among thespecial objectives were the search for
gold, camphor, and later pepper.The second commodity is especially known as
kapur Barus and appears(we are going back in time from the sixteenth century) in
ninth andseventh century Chinese literature as respectively ku-pu-p’o-lu
andp’o-lu perfume (compare the Acehnese pronunciation kaph
δ Bavδίh}.Barus
itself did not produce camphor.It was merely a reasonably shel-tered roadstead
at the mouth of the Batu Garigis or Camphor River,pro-tected by a protruding
spit of rock. But it had something that alongthis coast was truly exceptional
and important, namely rather easy ac-cess to the interior, here the Kalasen
region. Only in this mountain-ous and heavily-wooded part of Sumatra, and
confined to a comparativelysmall area, were to be found the very old giant trees
(Dryobalanopsaromatica Gaevtn.} that produced an excellent kind of
camphor,83MarcoPolo estimated that it was worth its weight in gold or silver,
and evenin the nineteenth century it still fetched about a hundred times
theprice of ordinary camphor in the Chinese market.84Its main export route was
by way of Barus, but some of it reachedthe coast further south through the
hinterland of Tapanuli Bay andthrough the Angkola Mountains. To the north there
was also a longerand more difficult route to the Simpang region and Singkel.In
thesixteenth century a virtual monopoly was established by Aceh, whichexported
the camphor by way of its own markets. The Dutch, who didnot like monopolies
except their own, occupied Barus. After it tran-spired that some of the camphor
still reached Aceh by way of Singkel,they stationed their fleet north of this
place, which did finally stopall traffic.85From at least the ninth century the
Arabs traded by way of Barus,though they had their own trade center further
inland at Pansur, a namethat replaced Barus in the Arab accounts. The oldest
Arab compiler,Ibn Khurd
δdhbih (844-48 A.D.), still mentions “Balus” as a source
of81. Pate Unus (or Yunus) is the name the Portuguese gave to Pangeran Sabrang
Lor”who crossed the sea to the north,” elder son of the first prince of
Demak,Raden Patah. H. J. de Graaf, Geschiedenis van Indonesie (TheHague: van
Hoeve,1949), p. 82.82. Ma Huan, Ying-yai Sheng-lan (TheOverall Survey of the
Oceanfs Shores), trans-lated and with an introduction, notes, and appendices by
J. V. G. Mills(Cam-bridge: Hackluyt Society, 1970), pp. 281-82.83. E. G.J. Mohr,
De bodem der tropem (Amsterdam: De Bussy, 1937), II/3(Sumatra),p. 390.84. Col.
Sir Henry Yule (trans, and ed.), The Book of Ser Marco Polo, 3rd ed. re-vised by
Henri Cordier (2 vols.; London: J. Murray, 1903), II, pp. 299-306.85. Kroeskamp,
De Westkust, pp. 149-50.

24excellent camphor.The next oldest, however, the author of The Rela-tion about
China and India (851 A.D.), tells us:MIn this sea (ofHarkand) . . . (lies) among
others an island called Lambri, whereinare many kings. . . . There is found much
gold, and a certain place,Fantsur, that produces a mass of first-class
camphor.11A neighboringisland is called ‘JNiyan” (Nias).8 6Ibn Safid (thirteenth
century) men-tions LSmuri, Fancϋr and JSwa as three separate ports of call.8
7Other foreigners, such as the Chinese and Marco Polo, followedthis usage, while
the indigenous traders held to the name Barus, as welearn from the
άgarakrtάgama, the Acehnese and Tomβ Pires, who ex-plained this to his
readers.That different peoples traded at differ-ent places up the same river and
even fortified their respective fac-tories, was .still common in the nineteenth
century.For example, inNatal the Acehnese had their fortification, called Kuta
Malaka, nearthe mouth of the river, while the Malays had theirs at Lingabaya,
oneday’s journey inland.8 8This must have been the situation at Barus aslong as
the port was not monopolized by one single group.To illustratethat the situation
at Barus was similar to that at Natal in later times,the Tamils probably
occupied the site of Lobu Tuwo, as may be inferredfrom the inscription of 1088
erected by a Tamil trading corporation.8 9Wolters points out that in the seventh
and eighth centuries Chi-nese sources located Barus somewhere on the north coast
of Sumatra.9 0This mistaken identity can easily be explained if, during the
ascen-dency of Srivijaya and the reopened navigation through the Straits
ofMalaka, some still little-known place had seized the opportunity tomonopolize
the camphor trade and to centralize the export of the famousproduct in its own
harbor (as was later done by Aceh).This hypothesisseems to be corroborated by
the^name Lang-p
τo-lu-ssu used by the HsinT’ang for the northern half of
Srivijaya.The first syllable probablyrepresents (as is suggested by Wolters) the
word Lam (village) asfound, for example, in Lam-uri^ the form of the name used
in the N
άga-rakrtάgama and the Sejarah Melayu and is thus most likely the
rightform.The second part certainly represents Barus.”Lam11may indeed be an
abbreviation of the subsequently famousLam-uri (Lambri, Lan-wu-li, etc.), a
place very aptly located by TeukuIskandar9 1slightly east of Kutaraja, about 18
kilometers inland at aspot now called Lamr
δh.Thus Pires was probably right in
saying thatit “stretched inland,” though he did not mean “as far as Fansur,”9
286. J. Sauvaget, Ahbar as-S
ίn wa’1-Hind: Relation de la Chine et de I’Inde
redigeeen 851 (Paris: Belles Lettres, 1948), p. 14.87. G. Ferrand (ed. and
trans), Relations de voyages et textes geographiques arabes(Paris: E. Leroux,
1913-14), II, p. 343.88. Kielstra, “Onze kennis,” p. 516.89. K. A. Nilakanta
Sastri,MA Tamil Merchant Guild in Sumatra,” Tijdschrift voorIndische Taal-,
Land- en Volkenkunde (TBG), LXXII (1932), p. 326.90. Wolters, Early Indonesian
Commerce, pp. 179-96.91. Teuku Iskandar, “De Hikajat Atjeh,” Verhandelingen van
het Koninklijk Instituutvoor de Taal-, Land- en Volkenkunde (VKI), XXVI (1958),
pp. 28-29.92. Wolters, Early Indonesian Commerce, p. 193.

25which is quite impossible since the latter for all practical purposescould
only be reached by the sea road. Thus the name “Lang-pfo-lu-ssu”seems to point
to a composite unity. It is possible that after “Lam11became a town its name
acquired an extension.Cowan explains the nameas Lam-puri, which is, however,
rejected by Lombard.93We may notethat uvί in Acehnese means “a
magically-protected boundary,”91* whichcould indicate that it was surrounded by
ramparts.It seems more prob-able, however, that the Chinese thought the name
Lang-wu-li-p!o-lu-ssuwas too long for a barbarian kingdom and shortened it, as
was theircustom.In the time of I-tsing the name Lam had not yet come to thefore
at all (as occurred after the ninth century) and he did not wantor was not given
the opportunity to visit the place himself. Thus hecontented himself by simply
indicating it as P’o-lu-shih.Again according to Wolters,95the name Barus was
used in the fifthcentury for a region in North Sumatra, while in the sixth
century itbecame connected with camphor, and still later with a port. Apart
fromthe scantiness of the evidence, which may well testify to the neces-sarily
confused ideas of authors who never went near the places theydescribed, it seems
exceptional that a place inherits the name of a(vanished) region, though the
reverse is very common (Borneo from Bru-nei, Sumatra from Samudra, etc.). That
the hinterland of Barus consti-tuted a “cannibal” area tells us nothing new. The
population livedbehind its mountain barriers and in forests far from the coast,
appear-ing only occasionally to barter small quantities of the precious cam-phor
for Indian fabrics (especially the more expensive ones), irontools, gold and
salt, as in the time of the VOC.96They even may havepracticed silent barter, as
was still done by some Batak tribes in thenineteenth century.97The king that is
mentioned can hardly have beenanything but: either the head of the Lamurian or
Acehnese colony atthe coast, who styled himself king, or a proto-Singamangaraja
in theinterior, who had some charismatic influence amongst the Ginting tribesof
Kalasen. While the place itself was certainly not attractive, itsproduce was,
and that was evidently what counted.We left our Ptolemaean sailors after they
took off from the coastof the Malay Peninsula at a place called “B
βrabai,” which
may have beenRa Way, at the southern tip of Pulau Bukit, or Kerabi, a little to
thesoutheast, and headed back west. After passing the northern extremityof
Sumatra, they made a sharp turn southward. At a certain point, how-ever,
probably to the south of Simeuleue, the “hog-island” of the En-glish sailors,
the course becomes southeast.Evidently they started tofeel their way coastwards
through the Banyak Islands and one of thepassages in the barrier reef. Their aim
was a place known for itstrade opportunities and called “Tak
δla” by Ptolemy. It
is the firststation on the voyage around the Golden Chersonesos.93. D. Lombard,
Le Sultanat d’Atjeh au temps d’Iskandar Muda 1607-1656 (Paris: Ecolefran
d1Extreme-Orient, 1967), p. 31.94. Hoesein Djajadiningrat and G.W.J. Drewes,
έhsch-Nederlandsch Woordenboek (2vols. Batavia: Landsdrukkerij, 1934).95.
Wolters, Early Indonesian Commerce, pp. 184-85.96. Kroeskamp, De Westkust, pp.
149-50.97. L. de Scheemaker, “Aanteekeningen gehouden op eene reis naar de
marktplaats(pedagangan) der Lima Laras,” TBG, XVII (1869), pp. 412-30.

Page 26

Page 27
LITTLE ANDAMANChetemalaKeyCH ChrysoanaZ Zabat/JambiT ThipinobαstαiS SdbαnαsPT
αtrαsαSA Sαmαrαnd&L LeistaiPerimulαbay of PAttάbαs,RiverC. UαleoukolonPA
αlαndαs RiverJ C. JeriyehBThe real route of the voyage.Ptolemy’s route.He based
his cal-culations on uniform averages ofspeed for each day. Therefore hisroute
credits the voyagers withabout three times their real speedand distance on the
east coast.Note also that according to Ptolemythe Golden Peninsula ran due
south.Compare the over-all shape ofPtolemy’s route with that of theChryse
δsos pictured in ourfirst map.THE SAILORS’ ROUTE

Page 28
28The name Takkola probably had a short-lived popularity in Indianliterature,
since it is mentioned in the Milindapaήha (first century)and Mahάniddesa (second
or third century).It could have been un-earthed from an old Tamil epic by the
author of the Tanjore inscrip-tion in the eleventh century, or its use may have
been continued inTamil maritime circles.This could explain why the inscription
doesnot mention Barus. The mysterious “QSqullah” of the earlier Arab ad-venture
stories is, at least by Buzurg, given a location far distantfrom Fansur98and
thus from the location of Tak
δla.With regard to this location we seem to have
two indications apartfrom the route outlined by Ptolemy. The first is that it is
the begin-ning of a gold-bearing country, the Chrys
δ Chersonesos, and that at
thesame time its name conveysMthe idea of some aromatic plant such asbdellium,
s’andalwood, camphor or cardamom.”99This rather unique com-bination of a site
famous for a certain aromatic product exactly at thebeginning of an equally
famous gold belt, is the very situation exist-ing along the Tapanuli coast.The
second indication is that here isfound the most likely name in Southeast Asia to
have given rise to orto be connected with the Ptolemaean toponym. The location
of such atoponym has sometimes been decided upon on the basis of a far less
con-vincing similarity.The name Angkola or Akkola, which also in India isknown
as a place name (like Takkola) and the name of a tree100may in-deed have
provided people in search of an appropriate Sanskrit namethat was both an
allusion to an existing name and to its product, witha satisfying
solution.Without the incentive of a double-entendre,they would hardly have named
the place after such a general and appar-ently not very common word for
aromatics.Exactly what point of the Tapanuli or Angkola coast they
visited,however, is not very clear.The features mentioned along the whole ofthis
coast are few and far between, viz., apart from Tak
δla one inlandtown and one
river mouth.Besides, though Ptolemy goes on calling hisfeatures emporion, town,
etc., their names, in contrast to those alongthe southeast coast, point to vague
entities and look like genericnames, as if no real centers were in existence and
the travelers tookpot-luck at a certain portion of the coast where they knew
that indige-nous praus would come out to them and where wind and tide at that
momentmade access and shelter possible.After passing aMpromontorium!lat two
degrees south of Tak
δla,which must have been the twin headlands of Ujung Tuan
and Biang rightbefore the bight of Air Bangis, the voyagers came abreast of an
inlandtown, which occupied the same longitude as Tak
δla.If the
comparativebearings of Ptolemy are about right, this must have been the Rao
coun-try.The name of thisMtownflisMKongkonagara,Mwhich we consideredabove to be
the equivalent ofMKanakanagaraπor Gold Town (compare7.1.50 Kognandaua for
Kakanada or KakanSva, an old name of Sanch
ί).101The name calls to mind the
ί (gold country) of the Dharma^raya98. Wheatley, The Golden
Khersonese, pp. 224-28.99. Ibid., pp. 268-72.100. J. F. Dastur, Useful Plants
from India and Pakistan (Bombay: D. B. Taraporevala,1964), p. 16.101. Eggermont,
“TheMurundas,” p. 263.

Page 29
29inscription (Minangkabau) and especially the Kanakapυrί on an islandin the
ίpSntara of the Kathasaritsagara.l°2In order to go there onehad first to
approach the king of Utsthaladv
ίpa. Batak legend knows as Raja Uti who had a
boar’s head and lived on an island off the coastof Barus. All rulers, the
Singamangaraja as well as Tuanku Barus andthe king of Pagerruyung, had to offer
him presents.103We may alsopoint to the Chinese name “Gold Island” for Pulau
Pini off AirBangis.104The name Kanakanagara may indicate the residence of
aproto-Tuanku Rao, who claimed a portion of all the gold that was mined,but we
do not have any real evidence to confirm the existence of anysingle prominent
center in these parts at the beginning of our era.One degree south the voyagers
passed the mouth of a river (or itsdelta; Ptolemy always uses the plural
ekbolai*) . Though no great storecan be set by Ptolemyfs distances, we may
nevertheless conjecture thatit must have been the Masang or Antokan River.Its
name is “Chrysoana”or “Chrysoanas.” It would seem that the corrupting influence
of goldis felt even in the text of Ptolemy.The name could be explained asan,
according to the standards of normal Greek, impossible or at leastcurious,
combination of ehvyso-, golden, and ana or ano, an adverbwhich means upwards,
above, on high, up country.Another curious thingis, however, that its Greek -as
ending is declined with the Doric geni-tive -a (Chrysoana potamou ekbolai*) ,
which commonly is used in non-Greek names. This would mark it as a pseudo-Greek
word, formed from aSanskrit or Prakrit name by a copyist who was impatient to
see somegold emerge. It could be, therefore, that its matrix either had
some-thing to do with gold (e.g., g
ίrisuvanna, mountain-gold river, a gen-eral
name for all rivers along this coast), or had nothing to do withit (e.g.,
ίshyarna, pounded river, Antokan). Since it is possiblethat some stretching
had to be done in order to obtain the desiredgold-effect, it will be difficult
to decide anything on the basis ofthis toponym alone.The end of the voyage to
the south comes at Sabara (other manu-scripts have Sabana, Sabala, etc.)>
f°rhere the course makes a sharpturn to the east and even to the northeast.We
may note that in nearlyall variants of the name the element “Saba” remains
constant, and thatthis element is also found in the name of the “Sabadibai” or
JavaIslands to the east of the Agathou Daimonos Island. The second partmay
derive from Sanskrit –
αrnα, stream, strait, sea (compare Dosara =Da^arna, the
present Orissa and the Desarena of the Periplous).x°5This SabSrna, Java Strait
or Sea, would have been a sensible appella-tion for the modern Sunda Strait and
even for the northwest coast ofJava.That it was an “emporion,” a place or coast
where customarilybarter was going on, is in any case evident from the
archeological re-mains in and around the Sunda Strait. Han ceramics have been
found inBangkahulu, Lampung, Tulangbawang and West Banten, next to other
pre-102. S. Levi,MIKfouen-Louen et Dv
ΐpantara,f” BKI^ LXXXVIII (1931), p.
623.103. J. H. Neumann, “Pustaka Ginting,” TBG, LXX (1930), pp. 92-93.104. Ma
Huan, Ying-yai, p. 282.105. Eggermont, “The Murundas,” pp. 269-70.

Page 30
30T!ang ware in Kerinci, Bangkahulu, West Lampung and South Sumatra.1 0 6A great
number of so-calledMRoman beads11were unearthed in Lampung,not in an isolated
hoard, but in connection with megalithic burialsites and bronze artifacts.1 0
7These conditions seem more in line withthe area being the principal gateway to
the archipelago and more apt tolead up to the Hinduized kingdom of Tarumanagara
and the fifth centurytrading communities of South Sumatra mentioned in the
Chinese annals,than the desolation and late development of the Malaka Strait.1 0
8Against the identification of the name as the Java Strait, itmight be objected
that Ptolemy interposes three Sindai Islands betweenFortuna Island and the
Sabadibai.We may note in the first place, thatthe identification would be
sufficiently justified by the fact that itgave access-to the north coast of
Java.But secondly, this interposi-tion may in reality have to be taken with a
pinch of salt.It is veryunlikely that Ptolemy’s Sindai can be connected with
Sunda as a whole.The latter name does not make its appearance before the
eleventh cen-tury and is thought by Berg and Gonda1 0 9to be derived from
suddha,”white11or “bright,1 1originally an epithet of the sacred Sunda
Moun-tain.Sindai would thus more likely correspond with siddha, theblessed
spirits who have “arrived” and have been admitted amongst thesemi-deities.1 1
0They are the nέnέfe moyang or rahyang of the Indone-sians.The Sindai must,
therefore, most probably be located in theregion of Pa-rahyang-an.l l lWe may
note that the toponym “Siddhapura”is not only known to the author of the
Aryabhatiya,112but occurs alsoin the topography of India.VI.The Bay of
PerimulaFrom Sabara the voyagers coursed to the northeast until they metwith the
delta of the Palandas river (Palandou potamou ekbolai}, where106. E. W. Orsoy de
Flines, Gids voor de keramische verzameling (Batavia: KoninklijkBataviaasch
Genootschap van Kunsten en Wetenschappen, 1949), and Orsoy deFlines, passim in
Jaarboek Batavia Genootschap (Bandung: Nix, 1937-38), IV andV.107. A. Th. a Th.
van der Hoop, Megalithische oudheden in Zuid-Sumatra (Zutphen:W. J. Thieme,
1932), pp. 92, 133-39; van der Hoop, “De Praehistorie,” in F. W.Stapel (ed.),
Geschiedenis van Nederlandsch Indie (Amsterdam: N. v.
uitgevers-maatschappij!Joest van den Vondel,11938), I, pp. 70-73; R. Broersma,
DeLampongsche districten (Batavia: Javashe Boekhandel § Drukkerij, 1916),
pp.321-22; G. P. Rouffaer, “Waar komen de raadselachtige moetisalah’s
vandaan?”BKI, L (1899), pp. 409-674.108. See Lamb, “Miscellaneous Papers,11p.
72.109. J. Gonda, Sanskrit in Indonesia (Nagpur: International Academy of Indian
Cul-ture, 1952), p. 223.110. A. Danielou, Hindu Polytheism (New York: Bollingen
Foundation, 1964), p. 303.111. That Sindai could have been mistaken for islands
was precisely because it wasnot situated at the thoroughfare but on the barely
accessible south coast,accidentally visited by an unhappy vessel that was blown
past the thoroughfareand wrecked on the shore.112. Kern, “Java en het Goudland,”
p. 308.

Page 31
31somewhere inland the kampung Palanda was situated. Both the masculine(-as) and
the feminine (-a) endings are simply Greek finishings, corre-sponding
respectively to potamos and polis . That here the genitive -ouis used, does not
mean, however, that we are necessarily confronted inPalandou with a Greek
name.The name derives most probably from theSanskrit palandu which means onion
and is thus the equivalent of Malaybawang. A petty kingdom of this name is
credited with considerablediplomatic activity by the Chinese imperial annals
around the middle ofthe fifth century.Its whereabouts cannot be determined,
because thename can be found wherever an Indonesian language is spoken.
When,however, in later sources a seventh century To-lang-p!o-wang
(Talangba-wang) is mentioned, it is thought to be the former Bawang with an
ex-panded or completed name; it could, moreover, with probability be con-sidered
as an inexact transcription of Tulangbawang, the name of ariver in South
Sumatra.113If the name Bukit Bawang Ujung north of theSemangka Bay has any
historical significance, the Bawang principalitymay have stretched across the
whole of South Sumatra. At the end ofthe seventh century it probably became a
part of Srίvijaya, since thelatter also placed on the south coast one of its
famous inscriptions.But Ptolemy’s voyagers may still have known it only as a
river and anindigenous settlement.Passing by for the moment the Maleoukolon
cape, we go on until wereach the delta of the third river the Chrys
δsos, the Attabapotamou ekbolai or the delta of the Attabas river. Though
the compoundmay be a bit uncommon, Sanskrit attambhas means “without water,”
whichequates Malay keying or kumering.Though according to KernlliftheMalay infix
“urn” was already “dead” at the time of the Sr
ίvijaya in-scriptions, it is still
a living element in this name. For the Kome-ring river, in contrast to its
rival, the Ogan, has indeed “periodic11spells of dryness, which makes it one of
the less useful waterways forinland travel. But where did those voyagers pass
its mouth? For atthe present the river empties itself not in the sea, but in the
Musiriver, nearly opposite the town of Palembang. Between the time ofPtolemy and
the present, however, the yearly silt deposits of therivers have added a marshy
strip that in some places may measure abouta hundred kilometers
across.Nevertheless, the path our sailors took is still a common way forsailing
craft. It consists of a combination of channels through themarshland which runs
from the mouth of the Masuji at the southern endto the mouth of the Komering at
the northern. The first river comesfrom the southwest and, taking a sharp turn
behind Talang Batu, emptiesin this channel creating in the act its southeastern
extension to thesea. The latter empties in the same way in this channel near
KayuAgung and has appropriated the last part of the gully as its own outletinto
the Musi. In the time of Ptolemy this channel must already haveexisted (with the
exception of the end part of the Masuji), but as anuncharted gully through the
coastal mangrove belt. Both these riv-ers emptied in this gully, which for all
practical purposes meant intothe sea. Since these gullies were no doubt
difficult to locate andsubject to changes, the sailors may have used local
guides (from113. Krom, Hindoe-Javaansche geschiedenis, p. Ill; Wolters, Early
Indonesian Com-merce, pp. 162, 202, 206.114. R. A. Kern,MEnkele aanteekeningen
op Coed
δs1uitgave van de Maleische in-schriften van Crivijaya,” BKI, LXXXVIII
(1931), p. 510.

Page 32
32Bawang?), lest the ship should run aground on submerged mud shoals.115Thus
after they passed the mouth of the Komering at Kayu Agung,they continued along
the coast until they reached the estuary of theMusi, which at the time was
located not much less than 100 kilometersinland of the present mouth.They turned
into this estuary at a pointbetween Plaju and Sungaigerong, where once a ship
with Buddhist mis-sionaries may have come to grief, since in the present mouth
of theKomering a collection of ancient Buddhist statues were dug out of
themud.116Here they were confronted by K61§ polis. The name Kδlg isprobably
derived from Sanskrit k
ΐila, “hill” (with a Greek feminineending corresponding
with polis). Kampung Bukit or Hill Town is thename of the most ancient site of
Palembang.When our voyagers turned into the estuary, they saw across thewater a
spit of higher wooded land (both the land and the settlementson it are called
talang} jutting out into the sea, and on it a row ofhillocks of which the
highest (26 meters, an enormous height in thiskind of country), the sacred Bukit
Siguntang, was the birth place andcornerstone of Malay leadership, according to
the tradition incorpo-rated in the Sejarah Melayu,When this text was written,
and still atthe time of Marsden (1783) the lower part of the Musi was called
theTatang River, while “upstream of its mouth” the Malayu River emptiedinto the
Tatang. On this Malayu River lay the Siguntang Hill. It cantherefore have been
none other than the small rivulet that is nowcalled the Tatang and which takes
its rise from the Siguntang. At sometime in the eighteenth or nineteenth century
it must have taken overthis name and lost that of Malayu River.Part of it had
been canalized,probably long before this change, in order to facilitate access
to thehill.This part is called Kedukan Bukit or Hill Canal.117It con-nected
Kampung Bukit with the Musi estuary.From here our voyagers continued in a
northeasterly direction toPerimula and subsequently to the bay of the same name.
We are not in-formed about the character of Perimula, whether it was a town, or
trad-ing center, etc.From similar instances we are still to meet, we canbe
reasonably sure that such an unqualified name indicates a region.The name is
most probably derived from the Sanskrit pr
ΐ or prίya (lov-able, pleasant), and
mula (root, foundation, beginning).In Malay,which uses a reverse word order, the
second notion is aptly renderedby the word mat (socle, base), and the first by
ayu (pleasant). Werepeat that it is immaterial whether this explanation is
etymologicallycorrect or not. We need only point out that the older form is
evidently115. See Geographical Handbook, pp. 84, 98-99; R. Sukomo,
“Geomorphology and theLocation of Crivijaya,” Madjalah Ilmu-Ilmu Sastra
Indonesia, I (1963), pp. 79-92; Mohr, De Boden, pp. 555-56.116. Amerta. Warna
warta kepurbakalaan (Jakarta: Archaeological Department of Indo-nesia, 1955), p.
10; F. D. K. Bosch, “Verslag van een reis door Sumatra,”Qud-heidkundig Verslag,
(1930), pp. 153-57.117. Sedjarah Melayu menurut terbitan Abdullah (annotated by
T. D. Situmorang andA. Teeuw) (2nd ed.; Jakarta; Djambatan, 1958), p. 22;
Sedjarah Melayu compiledby W. G. Shellabear (revised edition) (Kuala Lumpur:
Oxford University Press,1967), p. 20; L. C. Westenenk, “Boekit Segoentang en
Goenoeng Mahameroe uit deSedjarah Melajoe,” TBG, LXIII (1923), pp. 212-226; F.
M. Schnitger, ForgottenKingdoms in Sumatra (Leiden: E. J. Brill, 1939), pp. 7-8.

Page 33
33Malayu, and not Melayu.This seems to be confirmed by an inscriptionof 870,
though its Malayu is the name of a desa on the Dieng plateau.118The same form is
used for Sumatra by the Nagarakι?tagama.Besides, the”mauli-” element of royal
names in the DharmaSraya inscriptions119maybe connected with a similar
interpretation as that given by the “moulaMof Ptolemy.The course which we have
to follow in order to reach this Malayuseems plain enough.Ever since Rouffaer
wrote his article,120the cityof Jambi has been the undisputed seat of Malayu.
Besides, Ptolemy’scourse points straight northeast.The answer is, however, not
as sim-ple as that. Of course, the whole of Sumatra north of the Lampung
dis-tricts and south of the Batak mountains was or became part of the
“tanahMalayu11and each part could at one time or another have played a
promi-nent role. This could, however, hardly be the tanah Malayu which ismeant
here, nor for that matter the Malayu region I-tsing was speakingof. This was a
more limited feature that for some reason was especial-ly entitled to this name,
probably because it was considered to^be thecradle of the race. According to
I-tsing it was distinct from Srivi-jaya and on the way between^this royal
residence (kedatuan} and Kedah.121Before 685 itMhad become11Sr
ίvijaya. If he and
his compatriots Ch!angChin and Wu-hsing took the present usual way by sea, there
would besome reason to think about the environs of modern Jambi, but did
they?For there is scarcely any other substantial evidence to connect thisplace
with the original Malayu.Every tradition there is points to the banks of the
Musi. Srivi-jaya may have tried to replace the name Malayu with the new
“interna-tional” name of its kedatuan and to force this innovation on the
semi-independent neighboring demang and datu, centralizing in the processthe
organization of the state and fortifying its “royal” authority,traditionally a
shared affair, along Hindu lines of magical deifica-tion.This Buddhist-backed
enterprise enjoyed an enormous succcess inthe international (especially
Buddhist) press. But it left no tracewhatever in Malay tradition, which has
completely forgotten its nameas if its memory was obliterated on purpose.Its
twentieth centuryrestoration had to be unearthed from other sources.It is only
men-tioned in the Sundanese Carita Parahyangan in connection with the
expe-ditions of Sanjaya, but as Siwijaya (sio} , a king alternately of Malayuand
Keling.122Thus Sr
ίvijaya itself may not have lasted very long,though the name
remained in use for representative purposes.What really lasted was Malayu.It did
not “become” Sr
ίvijaya, asI-tsing boasted, but soon thoroughly shed the name
that was forced uponit. Apart from the just-mentioned Sejav
άh Melayu tradition
which con-118. See the review of Poerbatjaraka!s “Riwajat Indonesia11by L. C.
Damais in BEFEQ,XLVIII (1957), p. 627.119. Krom, Hindoe-Javaansche geschiedenis,
pp. 335-37, 414-16.120. “Was Malaka emporium v
όόr 1400 A.D. genaamd Malajver?”
BKI, LXXVII (1921), pp.1-174; 359-604.121. Wheatley, The Golden Khersonese, pp.
42-45; Wolters, Early Indonesian Commerce,pp. 185, 207-9, 227-28, 263.122.
Poerbatjaraka, “De Batoe-Toelis bij Buitenzorg,” TBG, LIX (1920), p. 403.

Page 34
34nects the origin of Malayu with Bukit Siguntang and Palembang, we havesome
traditions from Palembang itself, related by Westenenk.123 Onesays that before
the Muslim period Palembang was under the dominationof four princes. Along with
the Demang Lebar Daun of the SejarahMelayu are mentioned: a raja of Gunung
Mahameru, a raja of BukitSiguntang who resided in Malayu, and a demang of Tambun
Tulang. An-other tradition comes from the people along the Rawas River, who
main-tain that the region around Muara Limun and the Asai River (the habitatof
the so-called-Batin tribes) is the authentic Malayu.Schnitger onthe other hand
points to a district Tanah Malayu along the lower reachesof the Musi and an
island Pulau Wijaya between this region and Palem-bang.124This is probably the
“Tana Malayoπmeant by Pires and deBarros,125which is located north of the
Sekampung (Cacampom, Sacampam,and the Acampar of Descelliers) and of the
Tulangbawang (Tulimbavam,Tulumbauan) country.The maps of Diego Ribeiro (1529)
and Descelliers (1546) note, someway south of Palembang, the legend “Salida
(de)l CanalπandMSaida duCanalM(Exit of the Channel), while Diego Homem (1568)
also depictssome kind of channel going inland.It seems likely that this is
thechannel we described above, and at whose entrance (or exit) the Maleou-kolon
Cape must be located.For Maleoukolon means most probably MalayuChannel (from
Sanskrit kulya\ the -on ending is again a Greek featurecorresponding with
akron,Mcape?!), the channel that conveyed the voyag-ers directly to Malayu and
the Gulf of Malayu.The cape which at thetime must have marked its entrance is
the high ground around present-day Talang Batu. Thus, though the evidence on the
position of Malayumay not yet be absolutely convincing, we are prepared to
chance sailingup the Musi River (as was probably done by I-tsing) in the hope
thatthe voyage itself will add its own evidence.VII. From Kdlβ Polis to
ZabaiToday the Musi “carries a tongue of marshy land 25 kilometers widefar
inland to Sekayu, over 200 kilometers from the mouth of the river,but still only
9 meters above sea level.”126For the time of Ptolemywe have to cut down the
length by nearly 100 kilometers, so that aftera voyage of some 100 kilometers
through the region of Malayu, the trad-ers arrived at a bay that is currently
called Sekayu.This name ismost likely a contraction of Suak-ayu, Pleasant Bay,
though if Ptolemy’sinformation was right, its original name must have been Suak
Malayu.On the other hand, if the Chinese “W
βn” bay127can be identified withthis
same feature, the present name may be of ancient origin, becausewen represents
also, inter alia, the idea “well-ordered, pleasant.”Ptolemy announces that this
is the end of the Chrys
β Chersonβsosand moreover assigns Perimula to the same
latitude as Tak
δla. We may123. Westenenk, “Boekit Segoentang,” p. 223.124.
Schnitger, Forgotten Kingdoms, p. 8.125. M. Dion, “Sumatra through Portuguese
Eyes: Excerpts from Joao de Barros1!Decadas da Asia,111Indonesia, No. 9 (1970),
p. 145.126. See Geographical Handbook, p. 82.127. Wolters, Early Indonesian
Commerce, p. 52.

Page 35
35wonder, however, whether he based the announcement on his sources andthen
determined the latitude, because the Greek insistence on symmetryrequired that
both sides of the Peninsula should match, or whether ac-cording to his
calculations he had attained the same latitude and con-cluded that therefore
this must be the end of the Chersonβsos. Thelast sequence seems the more
probable one. His calculations were theresult of his custom to allot for every
sailing-day reported by hissources the same amount of stadia, i.e., the amount
he considered tobe a fair average (516 stadia a day). What, however, were the
conse-quences of this kind of calculation?It is most likely that the voyag-ers
sailed along the west coast starting early in the year in order toprofit from
the prevailing northeast to northwest wind so that theycovered the distances
between stops in a relatively small number ofdays until they reached Strait
Sunda.There the difficulties wouldhave started, because they would have met with
a strong current to thesouthwest. Having conquered this slowly, probably by
hugging the Javacoast, they could have waited (dependent on their objectives) in
somesheltered spot, most likely on the northwest coast of Java, until theturn of
the monsoon wind, and then continued their voyage up the Suma-tra coast.If,
however, they needed the rainy season when the riverswere much easier to
navigate, they could have continued at once underprotection of the coast and
navigated the vital Malayu channel. Al-though this last part of the journey
covered many days and few stadia,the voyagers were nevertheless credited with a
large amount of stadiafor each day. Thus when they arrived at Palembang, Ptolemy
imaginedthem to be at a latitude corresponding with Tak
όla and (on the otherside
of Sumatra) with some spot that in reality was not far south ofMedan.Going on at
the same rate, Jambi (and the south end of theMalay Peninsula) would have to be
placed north of Aceh. This processof prolongation of the east coast northward
was of course reinforcedby the corresponding shortening of the west coast
(though maybe not inthe same measure) and consequently the shifting of the
southern tip tothe north. This same miscalculation must also have been the main
rea-son for the enormous displacement of the Great Gulf northward to thelatitude
of the Ganges mouth and thus to what we now know to be SouthChina.Thus,
Ptolemy’s statement notwithstanding, we continue to thenortheast up the Musi
River until its junction with the Air Rawas, andfrom there up the latter river
to Muara Rupit (rupit
α?) and Surulangun.During the rainy season, the Musi is
navigable even by small steamersfor about 300 out of its present total of 553
kilometers. Several ofits tributaries are navigable too; the Rawas, for example,
is open tosailing craft up to Surulangun, a voyage of 15 days reckoned
fromPalembang.128We must note that about 35 kilometers north of thisSurulangun
(or Sorolangun), we find a place with a nearly identicalname, Sarolangun, at the
confluence of the Asai and Tembesi rivers.These twin places lie thus on
different river systems, the first onthat of the Musi and its tributaries, the
second on that of the Batang-hari-Tembesi. The land between them forms a trough,
an example of whathas been called a “sub-Barisan depression,”129low and marshy,
andmoreover, for the greater part of its length traversed by a tributaryof the
Asai, which descends into it at about 8 kilometers distance fromSurulangun.128.
Encyclopedie van Nederlandsch-Indie, vol. IV, pp. 192-94, vol. Ill, pp.
556-57.129. Geographical Handbook, p. 82.

Page 36
36Though at the moment water-traffic of any significance along thistrough is
impossible (but so is land-traffic; the road runs along thewestern foothills)
and there seems to be no oral tradition about acrossing in former times (the
sparse population are fairly recent immi-grants from elsewhere), nevertheless
these twin names seem very muchlike signposts indicating such a crossing.No
sensible indigenousderivation can be given (Sarolangun is derived by the people
fromsaro- or saru-melangun, and said to meanMto migrate wretchedly),130while a
Sanskrit«saro-ίahgana (pool or marsh crossing), that fits bothnames and the
situation, presents itself as rather obvious. These twonames are not on
Ptolemyfs map. He notes, however, a Samarand
β(samara-nadΐ, “battle river11). The
region itself, on the other hand,conserved the district name Penegah (adversary,
contester) between thetwo signposts of the crossing.Ptolemy’s Samara-nadi may
thus havebeen the translation of Air Penegah,Mthe river that was difficult
toconquer.f!We may note that SamarandS is again given without any quali-fication
and therefore is meant to indicate a region.It would be false, however, to think
that the voyagers were com-pletely lost in the backwoods. The foothills to the
west were famousfor gold, a fame that still held good at the time of the Jambi
War(1902). Since the Dutch found no crossing, they built the road alongthe
foothills in order to create access to this part of the country.At the time of
ίvijaya, however, this crossing may still have been ofvital strategic
importance:this is the more likely because one of itsdecrees was found at Karang
έ, some way north of this area on thebank of the Marangin, which we will
meet presently.It may have markedthe subjection of this part of the Malayu
heartlands and be included inI-tsingfs laconic annotation.Having conquered this
crossing, the voyagers entered another re-gion whose name seems badly garbled in
the manuscripts and is givenvariously as Patrasa, Pagrasa, Paprasa,
etc.Tentatively we conjecturethat these were floundering efforts to reproduce
the Sanskrit compoundpratapt
άsa (gold dust), the more so because they had indeed
arrived inthe Mas-urai (gold dust) country at the foot of the Masurai
mountainfrom which the Tembesi River (the river they had just entered)
origi-nates.In the framework of this estimate of their voyage route, themouth of
a certain S
δbanas river which they passed on their way down-stream must have
been the mouth of the Batang Marangin orπWindy River11which we mentioned above.
It comes down from Lake Kerinci and is themain northern tributary of the
Tembesi. Its Ptolemaean name S
δbanasmay have been derived from the Sanskrit
άna (luck or lucky). Inthat case its connection with Malay berangin is far
from obvious, ex-cept maybe for those incorrigible optimists who like to believe
thatevery wind is angin baik (a good wind, a windfall, a piece of
goodluck).Continuing downstream they arrived at an emporium whose name bythe
most trustworthy manuscripts is rendered as Thipinobastai. We willleave them for
the time being to their negotiations, however, and in-spect first the next item,
which is again a region and is calledMAkadraπor “Akandra.” The kappa could be a
substitute for a Sanskritaspirate in the middle of a word for which the Greek
alphabet has nograph, the more so since the Latin manuscripts generally read a
“chi”(Achandra or Akhandra). Whatever the origin of the compound Batanghari,130.
Encyclopedie van Nederlandsch-Indie, p. 705.

Page 37
37its components are undoubtedly batang (staff, pillar, treetrunk, etc.)and hari
(day). In Sanskrit, using the reverse order, this could berendered as aha(n)
(day) and dhara (bearing), with the latter elementmaybe containing an allusion
to dhάra (stream), the special meaninggiven to the word “batang11in the Central
Sumatran dialects.It seems likely that thisMPillar-of-the-day” country was,
underthe variant form Ahandhar
ί, also mentioned by the Liang Shu, where itwas
transcribed as Kan-t’o-li, a country that by Ferrand, Przyluski andCoed
located in the region of Jambi, and by Wolters132at leastalong the southeast
coast of Sumatra.Its recollection may even havebeen preserved by Diego Ribeiro
who notes on his map, between purobahala(Pulau Berhala) and andaragire
(Inderagiri), not Jambi but manderika(ahan-dh
άrika means “pillar of the day”).
The same may be true of thepresent name of the river Mandahara, though tradition
ascribes its ori-gin to immigrants from Johore.133Returning from the Batanghari
region (Ptolemy does not mention a king-dom) to Thipinobastai, we are certainly
confronted with one of those”typonymes £ consonance sanskrite,” though it is
difficult to knowwhich.The -ai ending may be a Greek plural as in the names of
peoples,or may (like -g) represent a Sanskrit –
ί as in
δlinggai/Bhauliήgί.*34The initial t(h)- could derive from Sanskrit c- as in
Tiastanes/Cas fanaand in the Thinai polis of the Periplous and Ptolemy (though
they speakabout a different polis), which is undoubtedly the Sanskrit
ίnanagara,the city of the Chinese. This does not lead us to any Sanskrit
wordthat, as far as I can see, could (even if we take into account
possibledeviations of copyists) have been the matrix of Thipinobastai. A
com-pound with dvtpin
ΐ– (river) would at least make sense, but the
commontranscription of dvi- and dva- seems to be simply di- and do-.Iftherefore
errors by copyists have to be taken into account (somethingthat in a name of
this length may certainly be expected), my best guesswould be an original
ίptinabhasί (light in [from] the sky). Though atthe present no place name Hari
(day), Matahari (sun), Dinihari (dawn)or the like appears on the map, it is
possible, however, that a traceof the Sanskrit name itself has been preserved in
(Muara) Tembesi, aname of which the origin is in any case obscure.After the
voyagers left the Bay of Perimula, they sailed througha region which, according
to Ptolemy, was inhabited by the L
δistaipeople. These were clearly
forest-dwellers, who hadMthe appearanceof animals, lived in holes or caves and
had a scaly skin that could notbe pierced by arrows” (7.2.21). The region they
crossed, starting fromthe Rawas and the Struggle River, is indeed the
millennia-old rightfuldomain of the Kubu tribes. Some of them still fitted the
above descrip-tion less than a century ago, though their “scaly skin” was more
realis-tically attributed to inveterate scabies and to the coat of dirt that131.
δs, Les Etats, p. 108.132. Wolters, Early Indonesian Commerce, pp. 211-12;
The Fall of Sr
ίvijaya, pp.181-83.133. Encyclopedia van Nederlandsch-Indie, vol.
VIII, p. 1859.134. Renou, La Geographic, p. ix.

Page 38
38habitually covered their bodies because they were as afraid of water asthey
were of strangers.The name is said to be derived from (dialec-tal) ngubu, “to
hide” (behind trees, etc.).135Ptolemy’s source, how-ever, seemingly provides
another explication of the name (since he didnot know the local dialect). The
Sanskrit lestas (slight, cursory,fleeting) portrays exactly the furtive ways of
these people and thefleeting glimpses the voyagers sometimes had of these silent
barterers.Moreover, it is the exact translation of the Malay kubίt (sekubίt
means”slightly,” mengubit “touching cursorily,” “appearing fleetingly”).The very
last item of the voyage is Zabai. Since it is supposedto be still inside the
territory of the L
βistai, it will have to belocated somewhere in the estuary of
the Batanghari and not on the is-lands across the Berhala Strait.It is a simple
polis, an indigenoussettlement, probably not of forest-dwelling Kubu, but a
fishing villageof boat-dwellers somewhere in the middle of the mud flats, who
mayoccasionally have functioned as guides through the shallows.Ptolemyfstoponym
could represent the name Jambi. Nasalization or de-nasaliza-tion of borrowed
words is common enough.136From the fact that thename appears in a text ascribed
to the pilot or trader Alexander onecould conclude that the indigenous name was
already well known and thatthe authors of the source might waive a
sanskritization.If they didnot, however, the Greek transcription would probably
be the same, sincethey had to look for a sound-simile, the word “jambi” having
no meaningin common Malay.They could have made a compound with sambin
(rower,boatman), such as samb
ΐpura (the Greek zέta can, as we noted,
representboth “j” and “S”). The reason for this supposition is that the
P!u-leichou of the third-century Chinese sources137can then be given a
loca-tion.It lay to the north of the W
δn bay and was evidently inhabitedby “boat
people” as the name indicates.Its naked black people withwhite teeth “look for
ships passing by and come flocking to them withfowls, pigs and jungle fruit
(which they offer) in exchange for metalarticles.” In that case Ko
(-ying)138could not have been far off.The name might indicate Aha(-pura), a
variant of D
ίptinabhasί.We may wonder what could have motivated the traders in
choosingthis way through the interior, which took about a month to accomplish?In
the first place, we don’t know what alternatives were available tothem. Through
the absence of a direct channel like the one that ledto Palembang, the road by
sea would have been, if not longer, in anycase more difficult and dangerous
because of the prevailing winds, theuncharted shoals and the danger of
pirates.It is probable that whatmade this coast “most favoured” was as much the
peaceful attitude ofits agrarian, gold-washing and forest-products-selling
population thanits geographical position.In the second place, if they had to
waitfor a change of monsoon, no better use of this time could be made thanby
trading in the interior for gold, silver, lignum aloes, yellow wax,benzoin and
other products and accomplishing part of their voyage intothe bargain. Whether
the traders used their own ships or made use of135. G. F. van Dongen, “Een en
ander over de Koeboes,” BKI, LXXXYIII (1931), pp.538-42.136. Gonda, Sanskrit,
pp. 233-34.137. Welters, Early Indonesian Commerce, pp. 52-53.138. Ibid.

Page 39
39indigenous river craft, while the ships tried to manage the sea route,is a
matter for speculation.I-tsingfs itinerary139and later Chineseaccounts140seem to
point to the second possibility.In any case, thefact that the ports at the
mouths of the Musi and Batanghari commandedthe entrance to the interior may,
more than anything else, have ini-tiated their rise to prosperity and power.In
regard to the controversy about the location of Srίvijaya, theJambi-party seems
to me to hold all the geographical trumps141whilethe Palembang-party commands
the historical arguments.142This impassearises because both parties start from
the same premise, i.e., theStrait of Malaka as the main thoroughfare. Without
the Sunda Strait,Palembang could not have taken the lead.By opening the Straits
ofMalaka for safe traffic it undermined its own position, though it couldstill
maintain its lead for nearly two centuries.Then it had to stepaside for
Jambi-jaya, which was known to the Chinese as San-fo-chfi(Samb
ί-jaya or
Sam-vijaya, the “victory of welfare”?) and to the Arabsas Zabaj. Many royal
compounds left their names in the entitlature ofsubsequent courts. Thus “jambi”
or “jamba” was used as a title forcertain leading poets and magicians (mantri
bhujangga’) at the Majapahitcourt.143If the supposition of Pigeaud144is true,
that rajajambiconsists of an honorific ra- and -jajambi, the second syllable
wouldseem to be the common abbreviation of jaya.In any case the Ming an-nals are
fairly explicit in affirming the continuity between Kan-t
San-fo-ch1i,145while both the Sung and Ming annals report that theking and his
residence were called Chan-pi (Jambi).146It was thereapparently that Balaputra
(who called himself Suvarnadvipadhipamaharajalike the later kings of Upper
Jambi) and the Sailendra dynasty put uptheir residence around the middle of the
ninth century. Besides, theArabs do not leave any doubt about the distinct and
superior positionof Zabaj with regard to both Sr
ίvijaya and KalSh.147We still
have to inspect the inland features of the Golden Cherso-n
βsos. Ptolemy asserts
that his “Peninsula” was dominated by variousanonymous mountain ranges. The
rivers flowing out of these mountainsmerged with one another further south.So
far nothing unexpected. Hegoes on to explain, however, that these united rivers
formed first theAttabas (the most northern according to his calculations),
secondly the139.Ibid., pp. 207-8.140. W. P. Groeneveldt, Historical Notes on
Indonesia and Malaya, Compiled fromChinese Sources (Jakarta: C. V. Bhratara,
I960), p. 73.141. Sukmono, “Geomorphology,” pp. 81-84.142. Welters, Early
Indonesian Commerce, pp. 208-9, 227.143. Th. G. Th. Pigeaud, Java in the 14th
Century (5 vols.; The Hague: M. Nijhoff,1960-63), V, p. 376.144. Ibid., IV, p.
373.145. Groeneveldt, Historical Notes, p. 68.146. Ibid., pp. 63, 72-73.147.
Krom, Hindoo-Javaansche Geschiedenis, pp. 112, 134, 204, 254, 304.

Page 40
40Chrysoanas, and last the Palandas.Because his sources mentioned noother
outlets, he took it for granted that only these three existed.This may have been
a reasonable assumption for an Egyptian, but if hehad known anything about the
region he was charting, he would have seenthat such an assumption was quite
absurd in relation to any part ofSoutheast Asia, even those parts that can boast
of a large main riverwith a mighty delta.It seems moreover, as far as can be
deduced fromthe rather obscure and probably corrupted text, that his Nilotic
frameof reference led*him to imagine that the streams first all united intoone
river which subsequently split in three branches. This makes thewhole outlay too
fantastic to be anything but an abstract interpreta-tion on paper. It was,
however, the best Ptolemy could do in his cir-cumstances.Of the “inland towns”
we have tried to locate Kanakanagara andPalandu. The other two, i.e., Kalongka
and Tharra, are either in theBataήghari region or south of Palembang, dependent
on whether theirposition was originally determined in relation to the west coast
or theeast coast data. With nothing definite to start from, an attempt
atidentification seems unpromising. The two places lack the first andfundamental
requisite, namely that their position is upheld by thenames before and after
them and that the whole pattern fits in with aknown geographical and historical
framework.It is evident, therefore, that the most definitive element intracing
the course of the Ptolemaean voyagers is not the interpretationof names or a
fixation on “coastlines,” but an interpretation of the”pattern of settlement as
set forth in the Geography”148against thebackground of the historical geography
of Southeast Asia. The namesmay be helpful, but their interpretation will always
be subject to mis-givings. This is especially true in connection with “the most
favouredcoast of early Indonesian commerce.”149The Malays were evidently
lessthan interested in literary Sanskrit substitutes for their own names.They
were very likely more concerned with commerce than with “culture”and may have
felt themselves equal to these intruders on sea routesthey had dominated long
before the appearance of the Indians.Hinducultural influence seems in any case
to have been more restricted herethan in other regions. The first inscriptions
of Sr
ίvijaya were not,as elsewhere, in Sanskrit, but in Malay, though mixed with
Sanskritwords. Even the Sanskrit name of the kingdom was eliminated by
Malaytradition. Nevertheless I think that with the help of the originalMalay
names most of Ptolemy’s appellations found their appropriateplace. It is somehow
more satisfying, however, if the sound of theappellation seems to fit the name,
in spite of the many false identifi-cations which in this way have been made. It
is possible that alongthe east coast of the Malay Peninsula and along the rest
of the GreatGulf Sanskrit names were more easily received and had a more
tenaciouslife than along the coast we just left. A study of the voyage in
theGreat Gulf area may be the subject of a later article.148. Wheatley, The
Golden Khersonese.149. Wolters, Early Indonesian Commerce.


  1. sandrar said,

    Hi! I was surfing and found your blog post… nice! I love your blog. 🙂 Cheers! Sandra. R.

  2. fuck you said,

    fuck your ass
    god hate you,,,,,

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: