Memorial of the Republic of Botswana

INTERNATIONALCOURTOF JUSTICE

CASE CONCERNING KASIKILIISEDUDUISLAND

(BOTSWANA/NAMIBIA)

MEMORIAL OFTHE

REPUBLICOF BOTSWANA

VOLUME 1

28 FEBRUARY1997 VOLUME1OFTHE MEMORTAL

TABLE OFCONTENTS

CHAPTER1 -THENATUREOF THE
DISPUTE

(a) The Basis of Botswana's Clah as
regards the Boundary between
Botswana and Namibia around 1

KasikiliISedudu Island
The Determination of the Boundary
(b) determines the issue of Sovereignty 2

(c) The Centrality of the Anglo-German
Agreement of 1890 has been accepted
3
by the Governent of Namibia
(d) The Character of the Dispute reflected
in the Provisions of the Special 9

Agreement
CHAPTERII -THE GEOGRAPHICAL
AND SOCIALCONTEXT AND THE

CHARACTEROF BOTSWANA'S
INTEREST
Description of the Chobe River in the
(a)
last 50kilometres (30 miles) pnor to
its confluence with the Zambezi River11

(b) Seasonal Flooding in the Flood Plain
of whichKasilcililSeduduIsland forms 12
a Part

(c) The Interest of Botswana in
KasikiliISedudu Island and the 13
Vicinity(d) The Relation of the People of the
Caprivi toasaneand its Facilities

CHAPTER m- TIIEHISTORICAL
BACKGROUNDTO THEDISPUTE

(a) Introduction

(b) Precolonial relations the Northern
Botswana -Caprivi Region
(i) Precolonial contacts

(ii) The process of colonial
partition

(iii) The 1890Anglo-German
Agreement

(iv) Events from 1890to 1914

(v) Questionsover the status of
the CapriviStrip
(c) The Administrationof British

Bechuanaland in the ColonialPeriod

CHAPTERIV - THERELATIONSOF
BOTSWANA WITH NEIGHBOURING
STATES SINCEINDEPENDENCEIN
1966

(a) The Geo-Politicsof Botswanaat
Independence

(b) The Impact of Botswana's
Independenceon NeighbouringStates

(c) The Economic Dependenceon South
Africa(d) Botswana's Political and Economic

Relations with Neighbouring States
(e) The Influx ofRefugees to Botswana
from South Africa and Rhodesia

(0 The Confrontation with South Africa

(g) The Frontline States and the Southem
African Development Community
(SADCI

(h) Further South African Pressure on the
Frontline States

(i) Acts of Aggression by South Africa
against Botswana

(i) UN Security Council Resolution 568
(1985)

(k) Developments in Namibia
(1) The General Pattern of Peaceful
Relations

(m) The Boundary Dispute

CHAPTERV - THEINTERPRETATION
OF THE ANGLO-GERMAN
AGREEMENTOF 1890

(a) The Agreement of 1890Constitutes
the Applicable Law

(b) The Interpretation of thenglo-
German Agreement of 1890

(i) The Text

(ii) The Position of Botswana

iii (iii) Interpretation in accordance

with theobject and purpose of
the Agreementof 1890
The SubsequentConductof the
(c)
Parties asConfirmatoryEvidence
(i) The Principle

(ii) CaptainH.V. Eason'sSurvey,
1912

(iii) The Joint Reportby Trollope
and Redman, 19January 1948

(iv) Official Admissionsby the
South African Govemment,
1948-49

(v) The British High
Commissioner'sAffiiation of
theLegal Status Quo, 10May
1951

(vi) The Opinionof the Surveyor-
General of Bechuanaland,18
October 1965

(vii) The Botswana-South African
Agreement Concludedat
Pretoria on 19 December 1984
and the Joint Surveyof 1985

(d) Conclusion

CHAPTERVI - THEPRETORIA
AGREEMENTOF 1984 AND THE
JOINT SURVEY OF 1985

(a) The Background The Meeting in Pretoria on 19
(b) December 1984

(c) The Intergovernmental Agreement of
1984

(d) The Joint Survey Report of 5 July
1985

(e) The Government of Botswana
regarded the Joint Survey Report as a
confirmation of title

CHAPTERVI1 - THE ISSUE OF FACT:
THE MAIN CHANNELIS TO THE
NORTH AND WEST OF

KASlKILYSEDUDUISLAND
(a) The Legal Context

(b) Criteria by which to identify the main
channel of the Chobe River

(i) Sole channel

(ii) More than one channel
(iii) Navigability

(iv) Thalweg

(v) Deepest channel

(c) The Evidence
(i) The hydro-morphological

evidence relating to the main
channel of the Chobe around
Kasikili/Sedudu Island (ii) Evidence supportingas a
matter of fact the northern and
western channel as the main
channel

(iii) Direct observation

(iv) Historical evidence
(v) Aerial photographs

(vi) Sedimentaryevidence

(d) Absence of evidence of change
(i) Absence of historical evidence

of change
(ii) The Presumption of Continuity

(iü) Absenceof scientificevidence
of change

(e) General Conclusion

APPENDMTO CHAPTER MI - THE
HYDROGEOMORPHOLOGICAL BASIS
FOR BOTSWANA'SCLAIMON THE
SEDUDUIKASIKILI ISLAND:OPINION
PREPAREDBY DR. F.T.K. SEFE

1. Introduction

2. The present morphology of the Chobe
River

3. Decipheringchannel changesin the
Chobe River System

3.1 The 1985Hydrographic
Survey 3.2 The hypothesisof floods in 6

1899
3.3 Evidence from Aqhotographs 11

4. Conclusion 12

CURRICULUMVITAEOF DR. F.T.K.
SEFE

CHAPTERVIII -THE RELEVANCEOF
MAP EVIDENCE

(a) The General Context 111

(b) Role of Maps as evidence 111
(c) Maps available relating tothe Caprivi 112

strip
(d) The Main Channel at the present time 113

(e) Evaluation of the Maps available 114

(i) Official and private maps 114

(f) Categorisationof the availablemaps 115
(i) Maps contemporarywith the
Anglo-Geman Agreement 116

(ii) Maps publishedby the German
Colonial Administration 117

(iii) BritishOfficia1Maps 118

(iv) South African OfficialMaps
published since 1949 121

(v) Botswana Officia1Maps 123

(vi) Third State Maps 123

vii(g) Generalanalysis of the available maps 124

(i) No map as an integralpart of
theAnglo-GermanAgreement 124

(ii) Too small scaleof maps 125
(iii) Unreliability of boundaries
drawn on maps 126

(h) Conclusionon the Map Evidence 127

CONCLUSIONS 129

GRAPHICSIN THEMEMORIAL
1. Affected sectionof the Chobe River opposite
referred to in Documents page 12

2. Chobe National Park Boundary opposite
page 14

3. Political Map of SouthernAfrica opposite
page 33

4. Opinion of Dr. F.T.K. Sefe 109

Figure 1. Extract from KasaneSheet
No. 1725 C3 and Part C4 showing
the Chobe River System

Figure 2. 1985 Hydrographic Survey
Cross-sections

viiiFigure 3. Palaeo Lake Levels

Associated with the 940-945 M and
936 M Elevation:

(A) 940-945 m level - Lake
Palaeo- Makgadikgadi

(B) 936 m level -Lake Palaeo
ThamalakaneVOLUME IIOF THE MEMORIAL
ANNEXES

(ANNEXES 1 TO 12)

ANNEXES
(ANNEXES 13 TO 62)

MEMORIAL:APPENDM 1:DOSSIER
OF AERIALPHOTOGRAPHS:

MEMORIAL:APPENDM II: DOSSIER
OF MAPS

ATTACHMENTTO THEMEMORIAL:

CROSS-SECTIONS OFTHERIVER

CHOBEAROUND KASIKILUSEDUDU
ISLAND BASED ON SOUNDINGS
TAKENBY THE 1985JOINT SURVEY CHAPTER1

THE NATURE OF THE DISPUTE

(a) The Basis of Botswana's Clairn as regards the
Boundary between Botswanaand Namibia around

KasikiliISeduduIsland

1. In accordance with Article III(2)of the Anglo-
German Agreement of 1890the boundary in the relevant sector
descends the River Chobe and is defmed as 'the centre of the

main channel of that river'. The German text refers to the
'Thalweg des Hauptlaufesdieses Flusses.There appears to be
no inconsistency between the English and German texts.

2. On this basis, as a matter of fact and law, the
main channel of the Chobe is the northern and western channel

in the vicinity of KasikWSedudu Island and hence the median
line in that northern and western channel is the boundary
between Botswana and Namibia. This is established by the
facts as evaluated by expert opinion (see Chapter VI1below).
Moreover, the identification of the northern and western
channel as the main channel, and hence as the boundary
between the two countnes, was the subject of an

intergovemmental agreement of 1984 and the Joint Suwey
Report of 1985 adopted on the basis of that agreement (see
Chapter VI below).

3. At no stage did Botswana or its predecessor in
title relinquish sovereignty overthe south eastern channel of the

Chobe River or in respect of KasikWSedudu Island. On
independence Botswana inhented British title in accordance
with the principles of State succession. The independence of
Botswana on 30 September 1966generated a continuing title by
virtue of its statehood. The title thus inherited was confmed by the Declaration of the Assembly of Heads of State and
Governrnent at Cairo in 1964 emanating from the First

Ordinary Sessionof the Organisationof African Unity. In this
instrument theAssembly:

"Solemnly declares that all Member States pledge
themselves to respect the borders existing on their
achievementof national independence." (Annex 35)

4. This principle, in the Declaration, is explicitly
'based upon Article 3(3) of the Charter of the Organisationof
African Unity, accordingto which Member States

"solemnly affm and declare their adherence to the
principles...
following

Respectfor the sovereigntyand territorial integrityof each
State and for its inalienable right to independent
existence."

5. Both Parties to the Memorandum of

Understanding (MOU) dated 23 December 1992 (Annex 57)
acceptedthe principlethat onthe achievementof independence
Botswana and Namibia succeededrespectivelyto the rightsand
obligations provided for in the Anglo-German Agreement
signedon 1July 1890. The prearnbleto the MOU (as set out
in paragraph 11 below) Statesthe position inclear terms.

(b) The Determinationof the Boundary determinesthe
issue of Sovereignty

6. In the circumstances of the present case the
determination of the boundary, in accordance with the

provisions of ArticleIII(2)of the Anglo-GermanAgreement,
will logically and inevitably also determine the question of
sovereigntyin respect of the island. 7. It has at all tirnes been accepted both by
Botswanaand Namibia, and by their predecessorsin title, that
the alignments defined in the Anglo-German Agreement
constitute an international boundary with the normal

concomitants of such a boundary. For present purposes the
pertinent characteristic of theboundary is its role in the
allocation of sovereignty. This role has equai operationai
efficacyboth in relation to landerritory and withina riverine
milieu. Indeed, for purposes of legal analysis and the
disposition of sovereignty, thereis no distinction. What is

involved is the territorial ambit of the States concemed, and
stateterritory includesrivers (or part thereof) and islands.

(c) The Centrality of the Anglo-GermanAgreement of
1890 has been accepted by the Government of
Namibia

8. TheGovemmentofBotswanawouldrespectfully
draw theattentionof the Court to the factthat the centralityof
the Anglo-German Agreementof 1890to the determinationof
the statusof the islandand river waters surroundingit hasbeen
acceptedby the Govemmentof Namibia on severaloccasions.

9. The first occasion on which this occurred was
the Headsof StateMeetingat Kasane,at whichHis Excellency
President Mugabe of Zimbabwe performed the role of
mediator. In the resulting Communiquéof 24 May 1992, the
relevant paragraphsare as follows:

"TheirExcellenciesPresidentSamNujomaof Namibiaand
Sir Ketumile Masire of Botswanamet inKasane, on 24th
May, 1992 in the presence of His Excellency President
Robert G. Mugabe of Zimbabweto discuss the boundary
between Botswanaand Namibia around SeduduIKasikili
Island. After the amival of President Mugabe and

President Nujoma, the three Presidents went on a tour of
the Chobe River and viewed the SeduduIKasikiliIsland, afier whichthey examinedvariousdocumentsdejining the
boundary around the Island. These included the 1890

Anglo-GermanTreaty, the1892Anglo-Gemzan-Portuguese
Treaty'and Maps. Thetwo treatiesdejine the Botswana-
Namibia boundaryalong the ChobeRiver asthe middleof
the main channelof that river."

"The Three Presidents after a frank discussion, decided

that the issue shouldbe resolved peacefully. To this end
they agreed that the boundary between Botswana and
Namibiaaround SeduduIKasikiliIsland shouldbe a subject
of investigationby ajoint team of six (6)technical experts
- three from eachcountrytodeterminewherethe boundury

lies in terms of the Treaty. The team should meet within
three (3) to four (4) weeks. The team shall submit its
fmdings to the three Presidents. The Presidents agreed
that the fmdings of team of technicalexperts shallbe final
and binding on Botswanaand Namibia."
(emphasis supplied) .

(Annex 55)

10. As a consequence of the decisions taken at
Kasane, the Governmentsof Botswana and Namibianegotiated
the Memorandum of Understanding (MOU) of 23 December
1992 (Annex57). The first substantive provisionin the MOU

is Article 2, whichprovides for the establishmentof 'ateam of
TechnicalExperts todetemine theboundarybenveenBotswana
and Namibia around Kasikili/SeduduIslandinaccorhnce with
the Anglo-GemzanTreatyof 1890' (emphasissupplied).

'~he referencehereto the Anglo-German-PortugueseTreatywas due
to a misunderstanding.No suchinstrumentexistsandthePartiesarenotdivided
on this issue. Il. The content of the preamble to the MOU
confms the role of the Anglo-German Agreement. The
matenal paragraphs are as follows:

"WHEREAS a Treaty between Britain and Germany
respectingthe spheresof influenceof the two countriesin
Africa was signed on 1stJuly, 1890;

WHEREAS Article m(2) of the said Treaty between

Britain and Germany respectingthe spheresof influenceof
the two countries in Africa, signed at Berlin on 1st July
1890, describes the boundary line defming the sphere in
whichthe exerciseof influencewasreservedto Britain and
Germanyin the south-westof Africa;

WHEREAS the Territory of the British Bechuanaland
Protectorate achieved Independence on 30th September
1966 as the sovereign Republic of Botswana which
succeeded to the rights and obligations under the
aforementioned Treatybetween Britain and Germany;

WHEREASthe Territory of South West Africa achieved
Independence on 21st March 1990 as the sovereign
Republic of Namibia which succeeded to the rights and
obligationsunder the said Treaty;

WHEREAS a dispute exists relative to the boundary
between the Republic of Botswana and the Republic of

Narnibia as described by Article m(2) of the said
Treaty ..."

12. TheMemorandumof Understandingdefmedthe
terms of reference of the proceedings of the Joint Team of
TechnicalExperts in Article 7 as follows:In the execution of its functions, the Tearn shall

have authority to:

(a) examine the Anglo-German Agreementof
1890 and the Anglo-German- Portuguese
Treaty of 1892 defining the boundaq
between Botswana and Namibia around
KasWSedudu Island and any relevant

maps andlor other materials relatingto the
boundary between Botswanaand Namibia
around KasikiliISedudu Island and to
determinewherethe boundary lies in terms
of the said Treaties;

(b) examine, evaluate and compile any
documentaryevidencerelied on by experts
from each side, each document properly
indexedand certifiedby the party producing
it and acknowledgedby the other party;

examine,evaluateandcompile detailedjoint
(c)
surveys made, includingplans, graphs and
statisticsandprovideany explanatorynotes,
clearly demonstratingthe resultsof the said
joint surveys;

do, perform or carry outany act or function
(d)
necessaryand relevantfor the determination
of the boundary between Botswanaand
Namibia around KasWSedudu Island,
taking into account, and subject to, the
provisions of the Treatieseferred to in (a)
above;

(e) secureandexamineanyrelevantdocuments,
charts, maps, plansand diagramsproduced
before and after 1890 relating to the said boundary ;

(f) hear, without prejudice to the 1890 and
1892 Treaties, any oral evidence from any

competent person in Botswanaand Namibia
or from any other country which the Team
may consider necessary to enable it to
arrive at adecision on the KasikiliISedudu
Island dispute;

(g) request, jointly or severally, historical
documentsfrom the archivesof Botswana,
Namibia or any other country;

(h) physically inspect the area of
KasikiliISedudu Island with a view to

obtaining first-hand information on the
boundary in dispute; and

(i) submitto their Excellencies,Presidents Sir
Ketumile Masire, S. Nujoma and R.G.

Mugabe of the Republics of Botswana,
Namibia and Zimbabwe respectively, the
fmdingof their investigationsin the form of
a report incorporatingthe conclusions and
justifications forhose findings andlor any
recommendations, including al1

documentation referred toin paragraphs (b)
and (c) above and any other matter or
material considered essential by the
Tearn."

21ndue course the Joint Team of Experts reportedtotheir Excellencies
at a Summit Meeting in Harare that they had failed to reach agreement: see the
Communiqué of the Summit Meeting (Ann59). 13. These provisions make it clear that the task of
the Joint Tearn of Technical Experts is confined to the
interpretationand applicationof the Anglo-GermanAgreement
signedon 1July 1890. Article 7(l)(a) refers exclusivelyto the

Anglo-German Agreement of 1890 and the (non-existent)
Anglo-German-PortugueseTreatyof 1892. Article 7(l)(d) is
signifïcantin this regard. This provision mandates the Joint
Tearnof TechnicalExpertsto 'do,perform or cany out any act
or function necessary andrelevantfor the determinationof the
boundary between Botswana and Namibia around
Kasikili/SeduduIslandtakingintoaccount,andsubjectto, the
provisionsof the Treaties referredto in (a)bove)'. (emphasis

supplied).

14. Article 7(l)(f) provides for the taking of oral
evidence and includes the proviso 'without prejudice to the
1890and 1892Treaties' .

15. Of substantialassistanceare the Minutesof the
Meeting on the TermsofReferenceofthe JointTeamofExperts
on the Boundary Between Botswanaand Namibia around

Kasikili/SeduduIsland, heM in Windhoekon 8th December
1992(Annex 56). These Minutes werejointly approvedby the
delegations. In theMinutes it is recordedthat the head of the
Namibiandelegation, Dr. Albert Kawana,

"...pointed out that the Governmentof South Africa has
confirmedthat the questionof Kasikiliwas never resolved
prior to the Independenceof Namibia. Hemaintainedthat

the intentionof the parties(Britain,Gemny and South
Afnca) to the 1890 Treatyand their subsequent practice
beforethe independence of the two countriesarethe most
importantfactorsto be takenintoaccountin determining
thedispute." (para.27). (emphasis supplied). 16. These characterisations of the dispute are
reflectedin the provisionsof the SpecialAgreement. However,
the reported position of the Governrnentof South Africa is of
doubtful reliabilityin the light of the Joint Surveyof 1985(see
Chapter VI).

(d) The Character of the Dispute reflected in the
Provisionsof the Special Agreement

17. TheSpecialAgreementisthenaturalculmination
of the transactions of the year 1992. The first preambular
paragraph refers to the Anglo-German Agreementof 1 July

1890. The text of the preamble as a whole refers to no other
instrument as being relevant to the determination of the
boundary .

18. The request to the Court in Article 1 is
compatible with all that hasgone before:

"The Court is asked to determine, on the basis of the
Anglo-German Treaty of 1 July 1890 and the rules and
p~ciples of international law, the boundary between
Namibia and Botswanaaround Kasikili/SeduduIsland and
the legalstatusof the island."

The provisionsof Article 1are to be understood
19.
in the light of the first two paragraphs of the preamble:

"Whereas a Treaty between Great Britain and Germany
respecting the spheresof influenceof the two countries in
Africa was signed on 1 July 1890 (the Anglo-German
Agreement of 1890);

Whereasa disputeexistsbetweenthe Republicof Botswana
and the Republic of Namibia relative to the boundary
around Kasikili/SeduduIsland,...."together with the fourth:

"Whereas the two countries appointed on 24 May 1992a

Joint Team of TechnicalExpertson the Boundary between
Botswanaand Namibiaaround KasilcililSeduduIsland "to
detemine the boundurybeîweenNamibia and Botswana
aroundKasikili/SeduduIsland" on the basis of the Treaty
of 1 July 1890 between Great Britain and Germany
respectingthe spheresof influenceof the two countriesin
Africa and the applicable principles of international
law ;.." (emphasissupplied) .

20. The preamble of the Special Agreement, the
provisions of Article 1 thereof, and the inter-governmental
transactions of 1992 rehearsedabove, all point unequivocally
to the character of the disputeas a boundarydispute. CHAPTERJI

THE GEOGRAPHICALAND SOCIALCONTEXTAND
THE CHARACTEROF BOTSWANA'SINTEREST

(a) Description of the Chobe River in the last 50
kilometres(30miles) priorto its confluencewiththe
ZambeziRiver

21. Kasikili/SeduduIslandisa smailislandabout3.5
Km2(1.5 sq.miles) in area in the Chobe River, located within

the area bounded by approxirnately 25O.07' and 25O.08'E
longitude and 17O.47' and 17O.05' S.latitude, and is
approximately 20km (12.5 miles) from Kazungula wherethe
Chobejoins theZarnbezi.The town of Kasanelies on the south
bank some 1.5km downstream from KasilcililSeduduIsland
irnmediatelyto the north-east, and the river, after continuinga

further2km. in a northerly direction, then flows east, after
encountering the rapids (of approximately 9m. drop) at
Commissioner's Kop, for some 14 km. (8.75 miles) till it
reaches Xahuma Island (Impalera on early maps) and its
confluencewith the Zambezi at Kazungula.

22. The Chobe River inits lower reachesisconfined
as to its southernbank by sandyridges rising to 1000m., dong
which run cattle tracksand the modem roadto Livingstonevia
the feny at Kazungula.

23. In the 10km.(6.5 miles)stretchof the river west

of KasikiliISedudu Island three meander loops are to be
observed, al1 of the characteristic shape of an otter's head
facingeast; the mostwesternat Kabulabulalooping some 4km.
north and two moremeander loops, one (sirnilarin the size of
its area toasilcililSeduduIsland) at Serondela 5 km. west of
the island. By reason of the sand ridge to the south these

meanders are uniformly made by the main Streamof the river
flowing in a northerly direction,l1have backwater lagoonsassociated with the outside bends of the meander loops.
Kasikili/SeduduIsland isitself formed by sucha meander loop

origindy made round a sand bar and like the other loops the
main chamel flows to the west andnorth clockwise,round the
island.

24. Immediately south of the island on the south
bank is locatedthe ChobeNationalPark headquartersand from

this point up toKasaneand the rapids at Commissioner's Kop
a sandy ridge of some 930m. above mean sea level forrns the
southernbank ofthe river rising steeplyfrom thewater'sedge.
To the westof the ChobeNationalParkHeadquartersthe sandy
ridge increases to 1000m. above mean sea level and is
separated from the river by a band of swamplandabout lkm.
in width. It is at this point that the sand ridges are cut by a

highvalley ,well wooded, known asSeduduvalley. Thisvalley
runs northlsouth with a strearn entering the swamplandabout
lkm. from the Headquarters building.

25. The Court is respectfullyreferred to the map.

(b) Seasonal Flooding in the Flood Plain of which
KasikiliISeduduIsland forms a part

26. The flowdong the Chobe (usuallyknownas the
KwandoILinyantilChobe)systemis normallyas a result of the
runoff from the catchmentarea of the system. The catchment

is shared with Angola, Botswana, Namibia and Zambia with
the larger portion of theareain Angola. The back flow from
the Zambezi is another contributor to the flow but this is
minimal when the Chobe is in flood from the catchment.

27. The runoff usually follows the pattern of the

rainfall regime which occurs over certain seasonsof the year.
The runoff therefore results in a lateral inflow with a lag of
severalmonths. The systemhas lagoonsand flood plainswhich
arefilledduringtheprocess. The flood plainsare of two types.Theperennial flood plainsarealwaysunderwater and seasonal
flood plains only receive water over certain seasonswhen the
water startstofill. The islandispart of the seasonalfloodplain

which is normally covered with water for a particular season.

28. The highestpoint onthe Kasikili/SeduduIsland
is approximately 927m above mean sea level and a gauging
station atKasane has recorded water levels from 922.05111 to
927.40m, which meanspart of the islandis subjectto seasonal
flooding, normally of four months duration: refer to the

hydrographs reproducedin Annexes60 and 61.

29. In this context it is to be emphasised thatthe
availability of the island does not depend exclusively on
whether the island is wholly or partially flooded. The surface
of the islandis veryuniform in level and, whenthere is partial

flooding, the whole island becomesintractableswampland.

(c) The Interest of Botswanain KasilcililSeduduLsland
and the Vicinity

30. It wiUbe obviousto the Court thatthe principal
interest of Botswanalies in the stability of its boundaries and

the confumation of the bounckuyprovided for in the Anglo-
German Agreementof 1890. In this respect Botswanahas a
common interest with other African States in the stabilityand
finality of boundary regimes.

31. The clarificationof the questionof sovereignty

is of particularignificanceinthecontext ofthe Chobedistrict.
Trans-border poaching activities and ivory smuggling,
involving both Namibian and Zambian tenitory, have been a
problem for at least twenty years. Inevitably, a disputed
boundary leads to risks of incidents between units of the
respectiveDefenceForcesofBotswana and Narnibia,operating

on the basis of conflictingmaps. 32. The island of KasikiliISeduduforms part of the
Chobe National Park established in 1967 (Annex 37) and,

before that, was part of the Chobe Game Reserve created in
1960 (Annexes32 and 33). The island is regularly frequented
by a variety of wildlife. It is the seasonal home of herds of
buffalo. The grazing on the island is excellent and there is a
daily elephant migration tothe island.

33. It is to be noted that the establishmentof the

Game Reserve in 1960 did not involve any change in the
existing use of the island. The island was not in use for
cultivation and this is evidenced by the dossier of aerial
photographs (Appendix 1 of the Mernorial). This includes
photographsfor the years 1943, 1962 and 1972.

34. The role ofthe islandas a major wildlifefeeding
ground readily accessible to safe viewing by tounsts has
assistedin the developmentof highvolumetourism basedupon
the Botswana side of the Chobe River.

35. In this context it is necessaryto emphasise that

the northernand western channelof the Chobe River isthe sole
channel navigable in relation to the needs of the regional
economy. It follows that the boundaq betweenBotswanaand
Namibia is the centreof the northem and westernchannel and
consequently the title of the Republic of Botswana over
KasikiliISedudugoes with accessto the navigable charnel of

the Chobe.

36. The southemandeastem channelissubstantially
impairedby a weii-establishedand persistent growthof Kariba
Weed. Moreover, thischannel isessentiallya backwaterlagoon
carrying floodwater flowing from theZarnbezipart of the year

and involvinga very low energy flow toward the Zambezi at
other times. If this channel were to form the boundary
Botswana's access tothe main channel of the Chobe as a
navigable river would be denied.(d) The Relationof the People ofthe Caprivito Kasane
and its Facilities

37. The geographical and social context of the

disputeinvolvesthe relationof the peopleof the neighbouring
area of the Caprivi (in Namibia)with the town of Kasane on
the Botswana side of the Chobe. The general stability of local
conditions, the existenceof Basubiafamilies on both sides of
the international boundary, the location of thenearest road
bridge across the River Chobe to Namibia, Ngoma Bridge,

some60 km (40 miles)to theWestandthe availabilityof social
amenities and trading stores in Kasane, have conduced to a
regular traffic of people and goods betweenthe Caprivi and
Kasane.

38. Kasanehas a populationof approximatelyseven
thousandpeople. Apartfrom shops andtradingstores, thetown

has a primary hospitalwithan outpatients department.There is
also a bank and a pharmacy. People from Kasika (in Namibia)
enter thetemtory of Botswanato sellfishand vegetablesdaily.
It is a common occurrence for Caprivians to use health
facilitiesat Kasane, Ngoma, Parakarunguand Satau, all places
in the Chobe District. Namibian children form part of the

enrolment in the primary school in Kasane. It is well known
that, at times of officially declared drought, people from
Namibia are the indirect beneficiariesof drought relief.

39. The good relations between the people of the
Caprivi and the people of the Chobe District of Botswanaare

evidenced by the conference held at KatimaMulilo in 1981.
This consistedof a meeting betweenthe Executive Cornmittee
of the Administration for Caprivians and a delegation from
Botswana led by Mr. Gambule, District Commissioner at
Kasane. The meeting was chaired by Chief J.M.M.
Moraliswani, Chiefof the Basubiapeople living inthe Caprivi

region. The Minute of the Conferenceare set forth in Annex
39. 40. For formal purposes within this Mernorial use
has been made of the convention adopted in the Special
Agreementto refer to the islandinvolved as 'KasikiliISedudu' .

41. The sources available variously describe the
island as either'Kasikili'or 'Sedudu'. The term 'Kasikili'is

recognisableas aBasubian locutionand inthat contextthe term
is not to be associated exclusivelywith onebank of the Chobe
River. Place namesof Basubianorigin are to be foundon both
sides of the Chobe River, reflecting the fact that Basubian
families live on both sides of the boundary.

42. It is not uncommonfor the narne'Sedudu'to be
employed in reference to the island. 'Sedudu'may also be a
Basubian term. The name is also appliedto the valley running
up into the plateau to the south of the island. As a matter of
general background, it is well known that, in an African
context, the sameplace may have severaldifferent names. The HistoricalBackgroundto the Dispute3

(a) Introduction

43. Thepeoples of Botswanaand Namibia havelong

been drawntogether by common historicalcircumstance aswell
as geographic proximity. During the late nineteenth century
both temtories fellnder colonial occupation following bnef

economic booms based on the export of game products.
Throughout much of the twentieth century their peoples have
collectively struggled to assert their independent national

identities in the face of the expansionist designs of former
white ruled South Africa.

3~heGoverment of Botswana has not considered
that detailed source references arenecessary, given
the function of this chapter. For the informationof
the Court the principal sources relied on are as
follows :

1. Livingstone, David, Missionam Travels and
Researches in South Africa, London, John
Murray, 1857.

2. Schapera, Isaac, The Ethnic Com~osition of
Tswana Tribes, London School of Economics and
Political Science, 1952.

3. Robinson, Ronald, Gallagher, John, and Demy,
Alice, Mrica and the Victorians: The Official
Mind of Imperialism,London, Macmillan, 1961.

4. Shamukuni, D.M., "The baSubiya1Iin Botswana
Notes and Records no. 4, Gaborone, Botswana
Society, 1972, pp. 161-84.

5. Tlou, Thomas, A Historv of Nsamiland 1750 to
1906, The Formation of an African State.
Gaborone, Macmillan Botswana, 1985.

6. Wilmsen, ~dwin N., Land Filled with lies, a
Political Economy of the Kalahari, Chicago &
London, University of Chicago Press, 1989. 44. The greater portions of Botswanaand Namibia
are encompassedby theKalaharisandveld.Notwithstandingthe

popular mythof its until recentpristine isolation, for centuries
communities living in this vast, semi-arid region have been
linkedto eachother and the outsideworldthrough networksof
trade and social exchange. Past contact among and between
various Khoisan, Shekgalagari, Setswana and Thiherero
speaking communities complemented shared lifestyles based
largely on migratory pastoralism, hunting and foraging. For

just over a century the colonially defmed border between
Botswana and Namibia has run roughly North-South. As a
resultmany communities, inparticular such Khoisan speakers
as the //Ai-khoe (Aukwe, Nharon), Nama, !Xoand Zhulhoasi
(Julwasi, Kung), haveseen their histonc hunting and foraging
lands politically divided.

45. Extensive inter-regionalcontact has also been a
long term feature of life in and around the Okavango and
Kwando-Linyanti-Choberiver systems of northern Botswana
and the adjacent Caprivi Strip. In this unique environmentof
water and sandy savanna interrelated communities of

Hambukushu, !mu, Shua-Khoe, Bayei and Basubia (also
known as Bekuhane) have also been divided by the border,
exceptfor theperiodbetween 1915and1930,whenthe Caprivi
was administratively integrated into the Bechuanaland
Protectorate. However, this facthad lirnited local impact as
there were few barriers to the free movementof people, who

also continuedfreely to plant gardensand graze their cattle on
opposite sidesof the border.

46. The 1885-90 British occupation of Botswana
came about largely as a strategic response to the 1884-90
German drive into Narnibia. The two countries, however,

experienced different decolonization tirnetables. Botswana
peacefully regained its sovereignty fromthe United Kingdom
in 1966.In the same year the Namibians,under the leadership
of the South West African People's Organization(SWAPO)began a protracted war of liberation against South African
occupation, whose ultimate success culrninatedin a United
Nations supervisedtransition to fuil independencein 1990. In
their fight for freedom the Namibiansin general, and SWAPO

in particular, enjoyed consistentovert supportfromBotswana's
govemment and people. Political soliddty between the two
nations, however, dates back earlier to the mid-nineteenth
century when indigenouscommunitiesthroughoutthe Kalahari
foundcommongroundintheir collectiveeffortsto resist andlor
adapt to the arrival of European socio-economic andpolitical
influence.

(b) Precolonial relations in the Northern Botswana -
CapriviRegion

(i) Precolonialcontacts

47. Archaeological findings and oral traditions
dovetail with early written accounts in showing that the
communities of Botswana and Namibia have been in close
contact for many centuries. Material evidence, for example,
suggests that pastoralism began in southem Africa, c.2000
B.C., among Khoe communities in northem Botswana, from

whence it spread into Namibia and other parts of the sub-
continent. By the fourth century A.D. iron age technologyhad
been adoptedin the region via northern Namibiaand north and
north eastem Botswana. Material evidence, such as rock
paintings and thepresenceof common patternsof ceramicsand
metal working, suggests that there was a generally peaceful

diffusion of goods and livestockculminating, cl ,000 A.D., in
the emergence of sophisticatedlate iron age trade settlements,
such as Toutswamogala, and mining complexes, such as at
Otjikoto.

48. First hand written evidence of the Kalahari

interior begins with the 18th century Dutchjoumals of Carel
Brink, Hendrick Wikar and Willem van Reenan, whichtogether describe trade routes linking Botswana, Nama and
"Samgomamkoa" ("San cattle people") in the southem
Kalahari. These and other early European writings and maps

alsooften refer to Batswanaas "Briquas"or "Birinas". Arnong
other things the Nama, dong with the related Cape Khoikhoi,
acquiredgoats (Nama: "biri") from the Batswana, whom they
thus called "goat people".

49. British Captain James Alexander's published

account of his pioneering 1836 expedition across the Namib
andKalahari providesadditionalfirsthanddetailaboutthe north
eastem as well as southeasternKalahari. He also reports that
Batswanaregularly crossedthe desertto trade with the Nama,
exchanging axeheads for cattle. Trans-Kalahari contact
increased during the mid-nineteenthcentury as a result of a
boom in exports through WalvisBay of ivory emanatingfrom

northem Botswana and the interior of centrai Africa. DuMg
this period various Batswana, Nama and Ovaherero
communities, dong with invading groups of mixed Euro-
Africandescent, notablythe Rehobother"Basters",and Jonker
Afrikaner's "Orlams", were able to establish their local
hegemonyover neighbou~g groups throughthe acquisitionof

firearms in exchangefor ivory and other gameproducts.

50. DuMg the 1860s a longstanding alliance was
forgedbetweenNamibia'sOvahereroand theBatawanabranch
of the Batswana. The Ovaherero ruler Maharero (r.1861-90)
and the Batawanaruler Letsholathebe(r.1848-76) concludeda

treaty of friendship, sealed through an exchange of cattle,
whichprovided for the right of refuge by membersof the two
communities in each other's country. This agreement was
upheldby Maharero'ssuccessorSamuel (UereaniKatjikumua)
(r.1890-91) and Sekgoma (r.1891 1906). Ultimately it served
as the basis on which large numbers of Ovaherero, including

Ovambanderu, were settledin BatawanaruledNgamilandfrom
the 1890s' more especially during the great 1904-8 war of
resistance againstthe Germans. 51. For many generations the Basubia of northem
Botswanaand the Caprivi were united as a smallpolity known
as Itenge after its founder ruler. However, it is necessary to
recall thatince 1800 at the latest this region was ruled either

by the Malozi from north of the Zambezi or by the Makololo.
In 1876 their last paramount ruler Nkonkwena, also known as
Liswani II, fled to the Boteti to escape the Malozi. He settled
at Rakops where he accepted the overrule of the Bangwato
ruler Khama III . ccording to Basubiaaccounts, Khama was
generousto Nkonkwena,giving hirncattleas weii as refuge, in
gratitude for the role played by Nkonkwena'sfather, Liswani

1, in rescuing Khama's father, Sekgoma 1, from captivity
among the Bakololo. Tothis day descendants of Nkonkwena
remain at Rakopsin central Botswana.

(ii) The processof colonial partition

52. The coming of colonial rule to Botswana and
Namibia involvedinterconnected episodes in the broader late
nineteenth century European scramble for Africa. Although
European overrule was formally imposed inboth temtories in
1884-85,pnor to 1890 there was little actual colonial control
on the ground.

53. In 1884the Germans caughtthe Britishoffguard
by proclairning a Protectorate over the Namib coast south of
the Cape Colony's enclaveat Walvis Bay. This proclamation
was preceded by the negotiation in 1883-84 of a number of
dubious concessions from minor coastal rulers by a merchant
namedAlbert Luderitz, who used themas a basis for claiming

the temtory on behalf ofhis German Colonial Society, whose
legal successorwas the German Colonial Company of South
West Afnca or "DKG" (Deutsche Kolonialgesellschaft fur
Sudwestafrika).By the end of 1884the German Govemment's
protection ofclaims negotiatedby Luderitz had been extended
dong Namibia's northem coast from Walvis Bay to just north

of Cape Frio (18'24') and some20 milesinland. This areawasthe nucleusof German South WestAfrica.

54. The German move into Namibia was audacious,

given that between 1876 and 1878 the British Special
Commissioner, William Palgrave, had signed treaties with a
numberof Namibiangroups,includingMaherero's, Ovaherero,
Afrikaner's "Orlams", the Rehobothers and Bondelswarts
Nama. In the later years Palgrave's effortshad been rewarded
withtheproclamationof a Britishprotectorateover a largearea

of central Namibia then labelled Damaraland, as well as the
annexation to the Cape Colony of Walvis Bay. To the east
Batswana lands up to the Molopo River were also occupied,
while moves were then under way to establish imperial
authority over Botswana proper. But, in 1880, a change of
govemmentinLondon, accompaniedby militaryreversesvis-à-

vis theTransval Boers and Arnazulu, resulted in a temporary
Bntish pullback throughout theregion. While the protectorate
over Damaraland lapsed, Namibia was still regarded by
officialswithinHer Majesty'sColonialOfficeand CapeColony
administration as lying within Britain's natural sphere of
influence.

55. Notwithstanding its past commitments, the
British Cabinetdecided to accept Germanterritorial claims in
South-West Afnca. This followedcommunicationsbetweenthe
German Chancellor, Otto von Bismarck, and the British
Foreign Secretary, Lord Granville, inwhichthe former Linked
his govemment's attitude towards the British occupation of

Egypt with Britain's stance toward Germany's Namibian
expansion. In renouncing its Namibian claims, the Bntish
government rejected calls by both the Cape Colony's
Pariiament and the Ovaherero that the Namibian interior be
placed under British mle.

56. Throughout 1885 German agents sought to
expandtheir country's jurisdiction into the Namibian interior.
By the end of the year a number of local rulers, includingMaharero, had signed treaties accepting Germanprotection,
though others, such as the Nama leader Hendrick Witbooi,
refused to recognise Berlin's authority in any way. With the

Anglo-Geman boundary in the Kalahari still undefined north
of the 22ndparallel of southlatitude, a German scouting party
toured parts of northem Botswana but did not conclude any
treaties.

57. In the German-Portuguese Declaration of 30
December 1886 (Annex2), Article 1provided as follows:

"The boundary which separates the German from the
Portuguese possessions in South West Afnca follows the
course of the Kunene River from its mouth to those
waterfallswhichare formed to the southof Humbe where
the Kunene breaksthrough the Serra Canna; thence it runs

along the parallel till it reaches the Kubango; thence it
followsthe course of this river as far as the placedara
which is left in the sphere in which the exercise of
influence is reserved for Germany; thence it runs in a
straight line to the east till it reaches the cataracts of
Catima on the Zambezi. "

58. Germany7sexpansionof its territorial claims in
Namibiabefore 1890was not matchedby the establishmentof
an effective administration on the ground. At the time
Bismarck'sgovernmentexpected the DKG to take the lead in
developing the area. But the undercapitalized DKG only

initiateda few smallscaleprojectsbefore 1890.For its part the
Imperia1 Governmentlimitedits expenditurebetween 1885and
1889 to the posting of a mere three officiais,including the
Govemor Heinrich Ernest Goenng. Working out of a mission
schoo17sclassroom inthe Ovaherero villageof Otjirnbingwe,
Goenng and hissubordinatesissueda total of sixregulationsin
their first three years, none of which were enforced. When, in

late 1888, Maharero nullified his Protection Treaty the
Otjimbingwe office was hastily evacuated. The Germanofficialsthen fled to Britishprotectionat WalvisBay, advising
other Europeansto do the same.

59. Facedwiththe optionsofpullingout of Namibia
or deepening their commitment, the Germans decided to
establisha militarypresence. On 24 June 1889, 21 men under
the command of Captain Curt von Francois landed at Walvis
Bay, followedby another40 man contingentin January 1890.
With the permanentdepartureof Goeringin August 1890,Von

Francois became the senior German official in the temtory.
German authority in the interior was limited to Windhoek,
where Von Francois established his headquarters in October
1890.

60. Whilst accepting German expansion in south-

west Africa, the British Cabinet decided to respond by the
occupationof Bechuanaland.Atthe endofJanuary 1885Berlin
was informedthat Bechuanalandhad beenplaced under British
protection. In the Proclamationof 30 September 1885 (Annex
3) declaring Bechuanaland and the Kalahari to be British
temtory, the northem limit was stated to be the 22nd parallel

of south latitude.

(iii) The 1890Anglo-GermanAgreement

61. The contemporaryboundary between Botswana
and Namibia was establishedby Article III of the 1 July 1890

"Agreement between Great Bntain and Germany, respecting
Zanzibar, Heligoland and the Spheres of Influence of the two
countries in Africa". As suggested by itstitle the agreement
was a comprehensive understandingbetween the two powers
that resolved overlapping claims throughout Africa. The
agreement was the product of bilateral diplomatic discussions

held in Berlin and London, most especially during May and
June of 1890. 62. Germany'sdesire to uphold its claim to access
to the Zambezi River for its South West Africa Protectorate
had been communicatedto Britain'sAmbassadorin Berlin, Sir

Edward Malet, by Bismarck on 24 November 1888. In a
subsequent despatch of 2 September 1889 by the Foreign
Secretary, the Marquis of Salisbury, to the German
Ambassador, CountHatzfeldt (Annex4), it was suggested that
Lake Ngami be seen as equally under German and British
infiuence "and thatGermany shallbe securedfree accessfrom

thatlake to the upper watersof the Zambezi". In this respect
itis notable that Lake Ngami, which in the late nineteenthas
well as twentiethcenturieswas often dry, is not linkedby any
permanent water course with the Zambezi.

63. On 30 September 1889 the German Chargé
d'Affaires,CountLeyden, proposedto the Foreign Officethat

Germanyacquirerightsto Ngamilandwestof 24Oeastlongitude
andnorth of 22' southlatitude(Annex5). But, on 9 November
1889, the British Colonial Office, having consultedwith its
High Commissioner for South Africa, expressed its strong
opposition to recognizing any German clairnsto Ngamiland
(Annex6). This standwascommunicatedto Maletin Berlinon

17 February 1890. The following day Malet was further
informed that Leyden's proposal was being discussed among
othercolonial mattersbetweenHatzfeldtand Salisbury. During
the second half of 1889 there was also considerableconflict
over competing clairnsbetween Germany and private Cape
Colonyinterests intheformerBritishDamaralandProtectorate.

64. In expressing its opposition to Germany's
Ngamiland claims the Colonial Office wished to uphold the
interests of concessionaireslready active in the areaand, in
the process, avoidany partition of the tenitory under the rule
of the BatawanaKing Moremi. By 1890 rival private British

claimsto the Ngamiland-Chobe regionwere beingput forward
by a numberof concessionaires.In September1889theBritish
South Africa Company(BSACO)of Cecil Rhodes had beenawarded the rightin its royal charterto rule the entire area on
Her Majesty's behalf. Subsequently, in June 1890, the

Company negotiatedthe Lochner Concession fromthe Malozi
King Lewanika, which consolidated its grip on Bulozi
(Barotseland).Earlier, in August 1888and 1889,the Batawana
King Moremihad grantedmineralrightsto the AustralAfrican
Exploration and Mining Syndicate and Messrs. J. Stronbom,
J.A. Nicholls and R.J. Hicks. These Ngamiland-Chobe

concessionswere later consolidated, becomingthe basis of the
BritishWest Charterland Company's commercial clairnsto the
area.

65. On 18March 1890 GermanKaiser Wilhelm II
acceptedBismarck's resignationfollowingpolicydifferenceson
a number of issues including German colonial interests in

Africa. Thereafter the new Chancellor, General (later Count)
Georg Lm Von Caprivi and his Foreign Minister, Baron
Marschall, accelerateddiplomaticdiscussionswithBritainover
Africa. After a conversationwith Marschall, Malet on3 May
1890 communicated Germany's desire to reach a
comprehensivesettlementsooner rather than later. Africa was

the subject of both Von Caprivi and Marschall's inaugural
addresses to the newly elected Reichstag. On 14 June 1890
Malet was informedby the Foreign Secretary that:

"The communications w hich have been in progress
betweenHer Majesty's Govemment andthat of Germany

have now reached apoint sufficiently advancedto justify
me requesting Sir Percy Andersonto retum to Berlin for
thepurpose of discussing the necessary details with Dr.
Krauel. The negotiations which he conducted in Berlin
during his previous visit, togetherwith the conversations
whichwehavehadwith the German Arnbassadorsincehis

return, have enabledthe two Governmentsto draw with
sufficient completenesstheoutline of an arrangementfor
the adjustmentof the matters that are in dispute between
them." "The frontier betweenNgamiland andDamaraland is not

definitivelydrawn, and the detailsof it must be left to the
negotiationsof Sir Percy Anderson andDr. Krauel; but it
is agreedthat in the latitudeof Lake Ngami, and up as far
as the 18th degreeof S. latitude, the Germanfrontier shall
coincide with the 21st degree of E. longitude from
Greenwich. The character of this country is very
imperfectlyknown, andthe very position of Lake Ngami

has been the subjectof considerableuncertainty.There is,
however, Littledoubtthat the21st degreeof longitudewill
amply clear it, and will encloseMoremi's country within
the British Protectorate.
(Annex7)

66. TheoutlineoftheAnglo-German Agreementwas
publishedin a semi-official GermanGazettein mid-June1890,
receiving favourable press coverage. TheAgreement was
subsequently signedin Berlin on 1 July 1890by Von Caprivi,
Krauel, Maletand Anderson.Botswana'sfrontierwith Namibia
was defined by ArticleIII of the Agreement (seeChapter1).

67. The Anglo-German Agreement coincidedwith
the forma1establishmentof Britishjurisdiction over Northem
Botswana through the Order-in-Council of 30 June 1890
(Annex 8), which conferred on Her Majesty's High
Commissionerfor South Africa, in his capacityas Govemorof
the BechuanalandProtectorate, the power to:

"provide for giving effect to any power or jurisdiction
which Her Majesty, her heirs or successors, may at any
time before or after the date of this Order have withinthe
limits of this Order."The limits of the Order were:

"The parts of South Africa situate north of British

Bechuanaland;west of the SouthAfrican Republicand of
Matabeleland;east of the German Protectorate; and south
of the RiverZambezi andnot withinthejurisdiction of any
civilizedPower. "
(Annex 8)

In a subsequentOrder in Council of 9 May 1891 (Annex 13)
the northern limits of the Bechuanaland Protectorate were
specificallydeclaredto include theChobe. The southern limits
of Southern Rhodesia were indicated in an Order in Council
dated 20 October 1898 (Annex14).

(iv) Events from 1890to 1914

68. During the quarter century following the 1890
Anglo-Germanborderagreementthe impactof colonial rule on
the peoples of the Bechuanaland Protectorate and German
South West Africa differed greatly.In the former the loss of

political independence was accompanied by economic
underdevelopmentand the introduction of elements of racial
subordination. But, with most of its land reserved for the
indigenous populationwho continuedto be ruled through their
traditional rulers, the territory remained at peace. While the
stationing of significant numbers of pararnilitary colonial

police, along with threatened military campaigns to crush
resistance to colonial administrative masures in 1892 and
1894, had an intirnidatingeffect, British authority ultimately
rested on the emergence of an understanding with the
Batswana.

69. By the mid-1890s a political consensus had
emerged among leading Batswana that imperial rule from
Londonwaspreferableto the likelyalternativeof incorporation
into the neighbouring white settler dominated Statesof SouthAfrica andlor SouthemRhodesia. Thisconsensuswas reflected
in the popularly supported 1895 petitions to keep the
Protectoratefree of British SouthAfrica Company(Rhodesian)

control, culrninating inthe 1895visit to theUnitedngdomby
DikgosiBathoen,Khama, andSebele,andthe 1908-12petitions
and mass agitation to remain outside the Union of South
Africa. Thus, at least until the last decade of the colonial era
nationalist sentiment withinBotswana was ironically equated
with the retentionrather than rejection of British rule.

(v) Questions overthe status of the Caprivi Strip

70. Before 1914 the statusof the Caprivi Strip was
unique in that de facto British authority existed in spite of
German clairns within the region. A token German police
presencewas establishedin the Eastern Capriviin April 1909,

but remained under British supervision.

71. British administrative interventionw ithin the
Caprivi before 1914was brought about by circumstance. In
1893 Protectoratepolice wereposted in Ngamilandto restore
the status quoafterthe Batawanaruler Sekgomaconquered the
western Caprivi. In the Eastern Caprivi jurisdictionntinu4

to be exercisedby the Balozi KingLewanika, who appointed
headmen to rule over local Mafwe and Basubia. By 1905,
however, order amongthe Basubialiving onthe north bank of
the Chobe river was becoming increasingly compromised by
the lawless activitiesof white bandits, many of whom were
renegadeBritishsubjects,despiteoccasionalpolicepatrolsfrom

the BechuanalandProtectorate and Rhodesias. In addition to
incidents of robbery, shootings, and poaching, a number of
these local warlords began to establish farms in the region,
forcing locals to work for them. Affidavits about undesirable
European activities in the territory, including instances of
individualstrying to pass themselves offas German officials,

werecollectedbythe British DistrictCommissionerat Sesheke.
In October 1908 a permanent Bechuanalandpolice post wasestablishedat Kazungulaunder CaptainEason, by which tirne
the establishmentof a German police presence acrossthe river

was anticipated.

(c) The Administrationof BritishBechuanalandin the
ColonialPeriod

72. With the outbreak of the First World War in
August 1914, the BritishGovernrnentmade the occupationof

German South WestAfrica an immediatepriority. Except for
the Caprivi Strip,this operation was carried out by the new
Union of South African Defence Force, which invaded the
temtory under the overallcommandof the SouthAfrica Prime
Minister and Commander-in-Chief General Louis Botha.
Among General Botha's victorious multiracial armyof some

50,000 was a contingent of 1,325 Protectorate Batswana.The
last Germantroops in Namibia surrenderedon 9 July 1915.

73. The Eastern Caprivi was occupied without
resistance by paramilitary police from Southern Rhodesiain
September 1914, while the western Caprivi was secured with

equal ease by Bechuanaland Protectorate police and local
Batswana underthe directionof the BritishResidentMagistrate
for Ngarniland. In November 1914the High Commissioner
provisionallyplacedthe entire Capriviunderthejurisdictionof
the Resident Commissionerof the BechuanalandProtectorate,
in a move designed to forestall potential postwar Rhodesian

claimstothe area. Until31 December 1920theStripwas mled
as a de facto part of the Bechuanaland Protectorate under
martial law.

74. Martial law in the Caprivi was ended by the
Governor General's ProclamationNo. 12 of 1922 and High

Commissioner's Proclamation No. 23 of 1922 (Annex 19),
which, retroactively from 1 January 1921, placed the Strip
under civilian Protectorate authority. Also from 1 January
1921, the whole of South West Africa formally passed underSouth African control as a League of Nations Mandate. A
notable feature of the South African regime was its
unwillingnessto allow the general repatriation of Ovaherero

and Nama refugees living in the BechuanalandProtectorate.
Until the end of 1929 the Western Caprivi was administered
from Maun, while the Eastern Capriviwas administeredfrom
Kazungula-Kasane.

75. Caprivi residents, however, wereexemptedfrom
the Protectorate7sHut Tax during the 1920s.As a result many

Basubia shifted their residences from the south to the north
bankof the Linyanti-Choberiver, while continuingto plough,
graze and hunt inside Bechuanaland. Such activities were
generallymet with officialindifferenceAn exceptionoccurred
in 1924whenthe DeputyMagistrateat Kasane, CaptainNeale,
is reportedto have givenverbal permission toKasikaresidents

to plough on KasikililSedudu Island. This intervention
apparently occurred dueto the factthat Governmentoxenwere
already grazing there.

76. The Caprivi was administratively reintegrated
into SouthAfrican administeredSouth West Africathroughthe

Governor-General'sProclamation no.196of 1929 (Annex20)
and the High Commissioner's Caprivi Zipfel Proclamation no.
27 of 1930 (Annex21). This developmentcame about in the
contextof renewedpressurefromPretoriafor theincorporation
of theBechuanaland Protectorate itself, intothe Unionof South
Africa. The new arrangement soon proved inconvenient. In
1937 the authonties in Windhoek askedto be relieved of the

burden of administeringthe Eastern Caprivi. Thus, in August
1939,theterritorypassedunderthe directadministrationof the
Native Affairs Department in Pretoria, which maintained its
administrative control until 1980. A Magistrate for Eastem
Caprivi was posted at KatirnaMulilo. To assist hirn provision
was made by the Govemor General'sProclamation no. 38 of

1940, for BechuanalandProtectorate Police officers to patrol
the Eastern Caprivias "Peace officers. CHAPTERIV

THERELATIONSOF BOTSWANAWITH

NEIGHBOURINGSTATES SINCE INDEPENDENCE IN
1966

(a) The Geo-Politics ofBotswanaat Independence

77. On attaining independence on 30 September
1966, Botswana, besides beinga land-lockedcountry, found

herself almost completely surrounded by politically hostile
neighbours. There was South Africa to the south and to the
east,pursuing a policy of aparthei- a form of racism which
enforced minority white supremacy over the majority black
population; South-West Africa (later Namibia) to the west,
which wasadministered bySouth Africa, and Rhodesiato the

north, in which minority settler-European descendants ruled
over a majorityindigenousblack population.The situationcan
readily be appreciatedby reference to the accompanyingmap;
the reader will recall that, at the relevanttime, Zimbabwewas
ruled by the regime of Mr. Ian Smith, then designated as
Rhodesia.

78. Southem Rhodesia, a British colony, had
declared itself independentof Britain (as Rhodesia)on 11
November 1965. Britainhad, as a result, soughtandgotUnited
Nations-backed economicsanctions. Britainhad in fact placed
a lonely battle-shipong the eastem Coastof Mozambique, a
Portuguese colony, in order to enforce economic sanctions

against Rhodesia. The United Nations sanctions initiated by
Britain were intendedto end the rebellionby Rhodesia, which
was commonly known as the Unilateral Declaration of
Independence, or UDI. The efforts of the international
community to enforce economic sanctions against Rhodesia
were in vain, as the most important country neighbouring

Rhodesia, namelySouthAfrica, did not participatein the UN-
backed sanctions. On the contrary, South Africa became themain trading partner of Rhodesia. South Africawas also the
transit country of more sophisticated goods from dubious
sources overseas destinedfor Rhodesia. Zambia, a relatively

weak and underdeveloped country tothe north of Rhodesia,
could only watchhelplessly.

(b) The Impact of Botswana's Independence on
NeighbouringStates

79. South Africa, on the other hand, had been

excluded fromthe (British)Commonwealthof Nations, which
comprisedBritainandal1her former colonies,straddlinal he
five continents. South Africa was progressively becominga
pariah statebecauseof her apartheidpolicy and her decisionto
supportthe Unilateral Declarationof Independence(UDI),by
Rhodesia.BothSouth AfricaandRhodesia were condemnedby
the rest of the world as outcasts; and were in consequence

therefore bound to cooperate in all spheres of human
endeavour.The minority regirnesof Rhodesiaand SouthAfrica
had no legitimacy to govern, as they had been elected by
minority white electorates. The majority black populations
could only watch fromthe sidelines.

80. Theonlyfriendlyneighbou~g statewasZambia
to the north with whichBotswanashared a minimalboundary.
Despiteher militaryandeconomicweaknessvis-à-visRhodesia
and SouthAfrica, Zambiaaccorded Botswana al he necessary
politicalanddiplomaticsupport.AngolaandMozambiquewere
still Portuguese colonies, whilst Lesotho and Swazilandwere
also emerging countriesbut completely surroundedby South

Africa and economically dependenton her.

81. Thebirth of Botswana,asa newlysovereignand
independent country, was not a welcome development tothe
regimes in Rhodesia, South Africa and the South African-
administered territory of South West Africa. These minority

regimes viewedthe independenceof Botswanaas a new threatto their political stabilitywhich was under constant challenge
from their respective oppressed majority populations.In fact,
the oppressedblackpopulationhad resortedto the prosecution

ofthe armed struggle,supportedby the Organisationof African
Unity, the Scandinaviancountries and organised groups, such
as churchesand non-govemmentalorganisationsin Europeand
the United States. Al1the liberation movements of Southem
Africa, namely the African National Congress (ANC), Pan-
Africanist Congress (PAC) and South West Afnca People's
Organisation (SWAPO) had their headquarters located in

Lusaka, Zambia.

82. The independenceof Botswana wasperceivedas
an inspiration to the black majoritypopulations of Rhodesia,
South Africa and South-West Africa to attain their own
independence and self-detemination. More importantly, if

Botswana could become a successful non-racial democracy,
there was no reason why blacks inthe neighbouring countries
could not assume responsibility in govemment. The white
oligarchiesin Rhodesia and SouthAfrica were promoting the
fallacy that blacks couldnot govem responsibly. Botswana's
successful policy of a non-racial democracy, surroundedby

minorityracist regimes,would thus serve as an exarnpleto be
emulated by the oppressed majority populations,and a rude
awakeningto the whiteoligarchies.Rhodesiaand South Afnca
could not tolerate a successfulblack majoritygovemment on
theirdoor-step,as it wouldwhetthe appetiteof their ownblack
populationsfor majorityrule. A successful non-racial Botswana
would also belie the propaganda that black people could not

hold the reins of govemment.

(c) The EconomicDependenceon South Africa

83. After aperiod of more than 70 years asa British
Protectorate, Botswana had gained independence as a poor

undevelopedcountry bereftof any natural resources.It was an
expanse of territory the size of France, with half-a-millioninhabitants. When the country asked for independence from
Britain, observers concludedthat the inhabitants were either

foolhardy or sirnplynaive. There was only one government-
builtsecondaryschool,supplementedby a fewmissionary-built
and tribally-builtschools.There were lessthafift kylometres
of all-weather roads and few basic medical facilities. There
werenofactoriesor any sourcesof employment.Botswanawas
completely dependenton SouthAfricafor al1her manufactured
goods and for employment opportunities, particularly in the

gold mines around Johannesburg. Any serious medical cases
had to be attendedto eitherin SouthAfrica to the south or in
Rhodesiato the north.

84. It is no exaggeration to state that southem
Botswanawas an economic appendageof South Africa whilst

northem Botswana wasan economic dependencyof Rhodesia.
Thus Botswana at independencewas not only politically and
militarily weaker than her neighbours; she was also
economically dependenton them. In terms of United Nations
economicparlance, Botswanawasa "leastdevelopedcountry".
The very railway line that transited Botswana from South

Africa to Rhodesiawas owned by Rhodesia. This railway had
been built by the British South AfricaCompanyof Cecil John
Rhodes, thefounderofRhodesia,who drearnedofjoining Cape
Town in South Africa, and Cairo in Egypt by a railway line.
Botswana,a newlyindependentAfricancountry, that wishedto
serveas a goodexampleof a non-racial democracyin the midst

of minority racist regirnes, was virtually a hostage. The
countries neighbouringBotswana were not friendly to her, as
their forms of govemment were political anathemas.

(d) Botswana's Political and Economic Relations with
NeighbouringStates

85. The set of circumstancesthat prevailed at the
time of independencedictatedBotswana'sforeign policybased
onpragmatismandnational survival. Botswana'sforeignpolicyposture had to take into account her geopolitical situation of
being landlocked and almost completely surrounded by
minority-ruled countries on which she was totally dependent
economically. Thus, at independence, Botswana did not enter

into any diplomatic relations with South Africa, but maintained
trade and commercial links, and intergovernmental technical
co-operation, with that country.

86. The other African countnes which had imposed
economic sanctions on both South Africa and Rhodesia readily

understood Botswana's predicament and exempted her from
following suit. It would have been an act of suicide for
Botswana to attempt to cut off her trade, commercial and
communications links with SouthAfrica and Rhodesia. Even if
Botswana had tned to impose selective sanctions on her
neighbours, it is alrnost certain that they would have retaliated
with fulelconomic sanctions so as to strangle her to death. So,

although political relations with her neighbours were
deliberately avoided, economic relations had to be maintained
out of economic necessity.

87. As a matter of principle, Botswana did not
accept any official aid or investment from Rhodesia and South

Africa. Private investment, particularly by multi-national
companies based in South Africa was welcome, such as
investment in mining by the Anglo-Amencan Company. As
expected, the minonty regirnes of Rhodesia and South Africa
irnposed policies and regulations that discouraged any
investment by multi-national companies in Botswana. It would

not have been in their interest to have a black majority-ruled
country flourish on their door-step.

88. The Anglo-Amencan Company which later
invested in diamond mining in Botswana gave very little in
return until theerms of this investmentwere revised radically,

in 1975, when Botswana increased her shareholding from 15%
to 50%. The South African currency was legal tender inBotswana until 1976 when the country introduced its own
currency. The United States and the countries of Western

Europe did everything possible to help Botswana survive by
providing her with economic and technical assistance. They
could not, however, offer any substitute for the cheap and
readily available manufactured goods and foods from South
Africa andRhodesia.

89. By asserting its independence,Botswanafuially

closedthe doorto the long-standingthreat of incorporation into
SouthAfrica which was containedin the 1910Act of Union of
SouthAfrica. This clausehad left it openfor the inhabitantsof
Botswanatojoin SouthAfrica. Although, properlyinterpreted,
the clause meant that the will of the inhabitants had to be
expressed freely, it is well-known that the Boers in the

Transvaal and the settlers in the Cape have at differentirnes
tried to coerceBotswanainto acceptingincorporation.

90. The independence of Botswana in 1966 thus
closedthe door to the possible optionof incorporationinto the
Union of South Africa, which had become a minority-ruled

republic in 1961. South Africa was pursuing the policy of
apartheidin whichthe majorityblackpopulationhad no say in
the political and economic life of their otherwise independent
and sovereigncounty.Botswana tookto condemningthe moral
banktuptcy of apartheid as a policy that denied the majority
people their human rights, particularly politicaland economic

rights. The black people were not allowedto participate in the
politics of their country and couldnot perform certain jobs
which were reserved for white people. Although South Africa
was irritated by Botswana's condemnation of the apartheid
policy, she could not take any punitive action against heras
Botswana was not participating in the economic sanctions

imposeclby African countries. Besides, South Africa did not
wishto deny herselfa marketfor her goods ina world that was
increasingly hostile toher. So, Botswana remainedan irritant
to be toleratedwithin the Customs Union of 1910, which alsoincludedLesotho and Swaziland,besides South Africaitself.

The Influx of Refugees to Botswana from South
(e)
Africa and Rhodesia

91. Onassumingindependence,Botswanasignedthe
1951 Geneva Convention Relatingto the Status of Refugees.
Even before Botswana's independence, refugees from South
Africa had aiready settled in or passed through to safer
independent African countries to the north. In order to

circumvent the logical role of Botswana asa country of first
asylum for refugees from the neighbouring countries, South
Africa, in particular, decided not to give any effect, either
director indirect, to the 1951 GenevaConvention.ThoseSouth
Africans who sought and were granted refugee status in
Botswanawere regardedas enemieswhosimplyused Botswana

as a springboardfrom whichto attack SouthAfrica. Botswana
protested vigorouslythat onlyrefugees were allowedto settle,
andthat thosewho pursuedthe armed struggle toliberateSouth
Africa by force of arms were resettledin other countries,uch
as Zambia and Tanzania. If freedom fighters were found
bearing arms in Botswana territory, they were arrested and

chargedwith illegalpossessionof arms. The lawsof Botswana
forbadeanyoneto carry arms without apermit. Asfar as South
Africa was concerned, and to some extent also Rhodesia,
refugees in Botswana werepotential enemies.

92. In order to safeguardher security,SouthAfrica
operateda clandestinesecurity andsurveillancenetwork within

Botswana, even amongstrefugees themselves. Some security
agents from South Africa simply infiltrated the refugee
population, posing as birds of the same feather. At the height
of the refugee crisis in Southem Africa, Botswana hosted no
lessthan30,000 refugees, whenher own people numbered only
half-a-million. In recognition of her services to the refugee

cause, the then President of Botswana, the late Sir Seretse
Khama, was awarded the Nansen Medalin 1978. Throughoutthe turbulent years of Southem Africa, Botswanacontinued to
host refugeesfrom SouthAfrica, Rhodesia,South WestAfrica,
Angola, Mozambiqueand from other distant countriesto the

north. The international community and non-govemmental
organisations in Europe and America supported Botswana
fmancially and moraily in her chosen cause to host and assist
refugees. It was inevitable that Botswana'sdecision to host
refugees from South Africa would precipitate a confrontation
with SouthAfrica and Rhodesia.

(f) The Confrontationwith South Africa

93. On attaining independence in 1966, Botswana
adopteda policyof non-aggressionagainst ai lhe neighbouring
States, including South Africa. This policy rested on the

principle of internationallaw that a statemay not permit on its
temtory activities for the purpose of carrying out acts of
violence on the temtory of another state.Prior to, and after
independence,Botswanapursued and administeredthis policy
to the letter. In the case of refugees admitted into Botswana
under the 1951 Geneva Convention Relating tothe Status of

Refugees, the 1969 OAU ConventionGoveming the Specific
Aspects of Refugee Problemsin Africa and under Botswana's
own Refugees (Recognitionand Control) Act, offenderswere
charged, prosecutedand sentencedor resettledin otherAfrican
countriesto the north, butnever retumed to their countries of
ongin. To the extent that South Africawas not a signatory to

the 1951 Geneva Convention, she tried in vain to force
Botswana to retum refugees to South Africa. When this
strategy failed, South African agents simply abductedsome
refugees from Botswana or sent them parce1bombs, which
exploded and killed them as they opened mailseeminglysent
by relatives athome. On 2 February 1974, OnkgopotseAbram

Tiro, a South Africanrefugee schoolteacher at SaintJoseph's
College, was killed by a parce1bomb. Again on 14May 1985
another South African refugeewas killed by a powerful bomb
explosionat Flat No. 2602 in Gaborone.(g) The Frontline States and the Southern African
Development Community(SADC)

94. It willbe noted that whilstZambia servedas the

headquartersfor the liberation movementsof Southem Africa,
Tanzania provided training bases. The advantageof Tanzania
was her safe distance from possiblemilitary attacks by South
Africa. Botswana, Tanzania and Zambia held informa1
consultations throughoutthe mid-1970sin an effort to promote
anegotiated settlementoftheRhodesianconstitutionalproblem.
These informal consultations crystallised into what became

known as the Frontline States forum. The agenda of the
Frontline States wassimply to promote the independenceand
sovereignty of those territories that were still under
colonialism, such as Angola, Mozambique, Namibia and
Rhodesia, as well as to eradicate apartheid in South Africa.
Botswana, Tanzaniaand Zambia were later joined by Angola

and Mozambique when these two Portuguese territories
achieved independencein 1975.

95. South Africa was apprehensiveof the activities
of the Frontline States.Botswanabecame increasinglyvocal at
the OAUand UN advocatingtheindependenceof Namibia and

Rhodesia, as well as the eradication of apartheid in South
Africa. By 1979, the Frontline States,then numbering six,
decided to set up an economic arm to promote the economic
liberation of Southem Africa. In April 1980, in Lusaka,
Zambia, the six Frontline States of Angola, Botswana,
Mozambique, Tanzania, ZambiaandZimbabwewerejoined by
establishing a forma1
Malawi, Lesotho and Swaziland in
organisation called the Southem African Development
Community, or SADC for short. Botswana became its first
chairman and led the organisation till August, 1996, when
South Africa was electedas the new Chairman. The twin aims
of SADCwereto reduce theeconomicdependenceof Southem
African states on South Africa and to CO-ordinateeconomic

development amongstthe member states. 96. Again, Botswana's role as leader of an
organisationwhich soughtto reducethe economic dependence

of Southern African stateson SouthAfrica did not go down
well with the latter. Southem African states had reached the
conclusionthat politicalindependencewithoutthe concomitant
economicindependence was hollow. Botswana sought andgot
the support of theEuropean countries and institutions, as co-
operating partners, for the objectives of SADC. Through
SADCactivitiesand its Programmeof Action, roads,railways,

port facilities and telecommunications were rehabilitated to
inter-connectal1the Southern African states.The efforts of the
frontline statesto promote political liberationand the effortsof
SADCto promote economic liberation, werebeing rewarded.
SouthAfrica was not arnused.

(h) Further South African Pressure on the Frontline
States

97. In April 1984 South Africa had pressured
Mozambique into signing a non-aggressionpact at a border
town calledNkomati, fromwhichthe Accord took itsnarne.In

effect, the pact forbadeMozambiquefrom givingany form of
assistance to the South African liberation movementsand, in
retum, South Africa would terminate support for the
Mozambican rebel movementcalled RENAMO.South Africa
then pressured Botswanato sign a sirnilarnon-aggressionpact
and, at the same tirne, revealed that a similar pact had been

signed secretly with the Kingdom of Swaziland in 1982. The
strategywasto clincha seriesof non-aggressionpactsalongthe
northem border in order to deny the liberation movements
accessto the territory of South Africa. As could be expected,
Botswana refused to sign any pact on the ground that there
were no guerilla bases in her territory, and that she had no
intention or the capacity to threaten the security of South

Africa.(i) Actsof Aggressionby South AfricaagainstBotswana

98. Like a woundedbuffalo, SouthAfrica began to

pursue what sheherselfdescribai as a policyof total onslaught
against her neighbours. The policy was predicated on the
premise that al1the neighbouringstates had combined against
her through the activities of the Frontline States and SADC,
and were harbouring guerillas to attack South Africa, which
was alreadyenduringeconomic sanctionsand anarmsembargo

imposedby the internationalcommunity.SouthAfricabeganto
launch military attacks against the neighbouring statesas well
asto destroytheireconomicinfrastructure.Themilitary attacks
were justifred as pre-emptive strikes against guerillas coming
from the neighbouring states. The destruction of economic
infrastructure,suchas bridges, wasclearlymeantto perpetuate
the economic dependenceof SouthernAfrican states on South

Africa.

99. In 1985, South Africa resortedto open acts of
aggression as an instrument of policy to force Botswana to
agree to enter into a non-aggression pact. Fromtime to the,
car bombsexplodedin the cityof Gaborone,targeted on South

African refugees.On 14June 1985a lightningmilitary attack
on the capitalcity left twelve people, including a six-yearold
child,killed. On 19May 1986helicoptersof the SouthAfrican
Airforce fired at the Botswana Defence Force Barracks in
Gaborone, to distractthe attention of the Defence Force from
a simultaneous attack on a civilian housing complex in the

neighbourhood. A civilian was killed and three others were
injured. On 8April, 1987,threeBotswananationalswere killed
in their sleep when a powerful bomb detonated anddestroyed
a vehicle parked nearby. It is signifïcantto note that these acts
of unprovoked aggressionwere perpetuated withoutany prior
complaint or waming through diplomatic channels, or

othenvise, to indicate any perceived threat to the security of
South Africa.(j) UN Security Council Resolution568 (1985)

100. The UnitedNations SecurityCouncil Resolution
568 (1985) (Annex49) stronglycondemnedtheunprovokedand
unwarranted military attack on the capital of Botswana by
South Africa as an act of aggression againstthat country, and
a gross violation of its territorial integrity and national
sovereignty. It further condemned all acts of aggression,
provocation and harassment, including murder, blackmail,

kidnappingand destructionof property committedby the racist
regime of South AfricaagainstBotswana.The Resolution also
demandedfull and adequate compensationby South Africa to
Botswana for the darnageto life and property resulting from
such acts of aggression.It also reaffmed Botswana's right to
receive and give sanctuary to the victims of apartheid in

accordancewithitstraditionalpractice, humanitarianprinciples
and international obligations.edlessto saythat SouthAfrica
chose to disregard the Resolution.

101. Confrontation between Botswana and South
Africa endured till 1990when the racist regime took concrete

stepsto negotiatea newconstitutional dispensation.The regime
allowed political activity andreleased Nelson Mandela from
prison after 27 years of incarceration. South African acts of
aggression againstBotswanawere put on holdfrom 1988when
South Africa began to withdraw its troops from southern
Angola, which it had occupied since 1975176.The process of

constitutional negotiationsin SouthAfrica, begun inearnest in
1990, was clearly incompatible with continued acts of
aggression againstBotswana andother neighbou~g States.

(k) Developmentsin Namibia

102. Namibiawasa German colonytillthe end of the
First World War when it becarne a League of Nations
mandated territory administered by Britain. Britain in turn
delegated that responsibilityto its own colony, South Africa.With the demise of the League of Nations, South Africa
systematicallytreatedNamibia as itsfifth province and resisted
the efforts of the successor authority, narnely the United
Nations, to re-assert its mandate over Namibia. As early as

1903-4, the inhabitants of Namibia had been fleeing into
Botswana to seek refuge from colonial persecution at home.
Anotherinflux occurredin 1913,and becamea constant feature
during the war of liberation, which ragedfrom 1966to 1990,
when the territoryfinaiiy gained independence.

103. Thus, when Botswana gainedindependence, in

1966, a large number of refugees from Namibia had already
been integrated as citizens. In 1967, the United Nations
establishedthe Council for Namibia as the legal authority for
the administrationof the territory. The Council remained ade
jure authority whilst South Africa continued to exercise de
facto authority until independence,in 1990. Botswana spared

no efforts to press for the independence of Namibia whilst
hosting quitea sizeablepopulationof refugeesand settlersfrom
that temtory. Naturally, Botswana became a mernber of the
UN Council for Namibia.

104. In 1978, Botswana, together with the other

Frontline States, negotiatedUN Resolution 435(1978) (Annex
38) which established the independence plan for Namibia.
Meanwhile, South Africa had been trying without successto
promotean intemal settlementwith its clientparties, excluding
SWAPO. The wa. in southem Angola became unpopular in
SouthAfrica, asthe casualtiesmounted. The irnplementationof

UN Security CouncilResolution 435was triggered off only in
1988 when a three-comered Agreement was signed in New
York. The Agreement called for the withdrawal of South
African troops from Angola, the departure of Cuban troops
from Angola, and for the implementationof Resolution435. 105. The implementationof Resolution 435 wasnot
without incident. For some inexplicable rasons, SWAPO
poured troops into Namibia from Angola on the eve of the

implementation of the UN Resolution in April 1989, with
heavy losses. The UN Secretary-General's Special
Representative, Marti Ahtisaari, the current President of
Finland, clearly neededassistance to tackle theSouth African
Administrator-General,with whom he was expected to co-
operate in implementing the UN Resolution. Botswana's

PermanentRepresentativeto the United Nations was appointed
DeputySpecialRepresentativein Namibia.The Resolutionwas
then implemented successfullyand launched Namibia as a
sovereign and independent country on 21 March 1990.
Throughout the implementation of Resolution 435 (1978)
(Annex 38), Botswana, together with the six other Frontline

States, had maintainedObserverMissionsin Namibia,in order
to monitor and assist the processof decolonisation.

106. Meanwhile,the UN had decreedthat in the run-
up to electionsin Namibiaall political partiesshouldbe treated
equally without anypartiality. But it wasan open secret that

South Africa channelled millions of dollars for election
campaigns by internal client parties, to the exclusion of
SWAPO. The OAU decidedat its Summit inAddis Ababa, in
1989, to set up an OAU Fund for Namibiafor the purpose of
assisting SWAPO during the election campaign. There is no
record of the Fund being honoured, as individual OAU States

preferred to assist SWAPO directly. Botswana was no
exception. Botswanadelivered twenty 4-wheeldrive vehicles
which could traverse the desert temtory without difficulty.
SWAPOwontheelectionshandsomely, andexpressedgratitude
to Botswana.The Presidentof Botswana becamethe first Head
of State ever to pay a state visit to Namibia and did so only

four months after its independence. A generai pattern of
peaceful relations and co-operation between Botswana and
Namibia was bom.(1) The General Pattern of Peaceful Relations

107. ThroughouttheyearsofNamibia'soccupationby

South Africa, Botswana was a strong advocate of that
territory's independence andself-determination.Botswanaalso
providedrefuge and succourto those fleeingpersecutionin the
territory.Dunng the visit of the President of Botswana to
Namibia in July 1990, three CO-operationagreements were
signed. These related to CO-operationin defence and security
issues, including anti-poaching; economic CO-operationand

cultural CO-operation. otswana and Namibia have managedto
forge close linksbetweentheir respectivearmed forces, police
and immigration authorities. Namibians were attached to
Governent Departmentsfor on-the-jobtraining in Botswana
and there is still a Streamof visiting Ministers and officials
from Namibia, to this day, to learn from Botswana's
experiencein development.The two countriescontinue to co-

operatewithinthe SouthernAfricanDevelopmentCommunity ,
which Botswanaled as Chairman from April 1980 to August
1996.

108. The Presidentof Namibiaalso paida returnvisit
to Botswana in September 1990, during which Batswana of

Namibian origin, who wished to return to their ancestral
country, were allowed to do so. The voluntary returnof the
descendants of Namibian refugees was to be conducted en
masse, which involveda waiver of the rules of emigrationand
immigration. This was indeed a true reflection of the good
relationsbetweenthe two countries. The voluntary repatriation

was carried out from 1993 until completion in 1996. The
exerciseincludedthemovementofpeoplewho wantedto retum
to Namibia, together with their livestock. Under normal
circumstances, suchan exercise wouldnot have beenallowed,
as it representedthe depletionof another country's resources.
It was a true expressionof mutual goodwill.(m) The Boundary Dispute

109. Botswana and Namibia found themselves at
loggerheads in Mach 1992 when a boundary dispute was
recognised to exist around Kasikili/SeduduIsland dong the
ChobeRiver. The boundarybetweenBotswanaand Narnibiais
basedon the Anglo-GermanAgreementof 1890which defmes
the boundary as the centre of the main channel of the River.

Again, as two friendly countries,Botswana and Namibiahave
decided to resolve the dispute by peacefulmeans and have
referred the matter to this Court. CHAPTERV

The Interpretation of the Anglo-German
Agreement of 1890

(a) The Agreement of 1890 Constitutes the Applicable
Law

110. In Chapter 1of the presentMemorial the nature

of the dispute was exarnined andthe conclusion reachedwas
that the dispute concerns the interpretation andapplicationof
ArticleIII(2) of the Anglo-GermanAgreementof 1890.Thus,
in the provisions of Article II of the Special Agreement, the
Court is askedto determine theboundary 'onthe basis of the
Anglo-German Treaty of 1 July 1890 and the rules and
principles of internationalla'

111. The relevanceof the Anglo-German Agreement
isalsoconfirmedby theKasaneCommuniqueof 24 May 1992.
The relevant paragraphsare as follows:

"TheirExcellenciesPresidentSamNujomaofNamibia and

Sir KetumileMasire of Botswanamet inKasane, on 24th
May, 1992 in the presence of His Excellency President
Robert G. Mugabe of Zimbabweto discussthe boundary
between Botswanaand Namibia around SeduduIKasikili
Island. Afterthearriva1ofPresidentMugabe and President
Nujoma, the three Presidentswent on a tour of the Chobe

River and viewed the SeduduIKasikiliIsland, after which
they examined various documents dejining the boundary
aroundtheIsland. Theseincludedthe 1890Anglo-German
Treaty ,the 1892 Anglo-German-Portuguese Treaty4and
Maps. The two treaties dejine the Botswana-Namibia

4~he reference here to the Anglo-Gennan-Portuguese Treaty was due
to a misunderstanding. bounduryalong the ChobeRiveras the middle ofthe main
chunnelof thut river."

"The Three Presidents after a frank discussion, decidd
that the issue should be resolvedpeacefully. To this end
they agreed that the boundary between Botswana and
Namibia aroundSedudu/KasikiliIsland shouldbe a subject
of investigationby ajoint teamof six (6)technical experts
- threefrom eachcountryto determinewheretheboundary

lies in terms of the Treaty. The team should meet within
three (3) to four (4) weeks. The team shall submit its
findings to the threePresidents.The Presidentsagreed that
the findingsof team of technicalexperts shallbe final and
binding on Botswanaand Namibia." (emphasissupplie.).
(Annex55)

112. It must follow that the provisionsof the Anglo-
German Agreement constitute the applicable law for the
purposes of the determination of the present dispute. The
additionalreferenceto 'the rules andprinciplesof international
law' is pleonastic, in light of the fact that an international
agreement is nomally interpreted taking into account 'any
relevant rules of international law applicable in the relations

between the parties' (Vienna Convention on the Law of
Treaties,Article 31(3)(c)).

(b) TheInterpretationof the Anglo-Geman Agreement
of 1890

(i) The Text

113. The text of the Anglo-German Agreement is
reproduced in the annexes from two official sources, one
German and one British. The German source is Die deutsche
Kolonial-Gesetzgebung, Part 1 (Berlin, 1892) (see Annex 11).
TheBritishsourceis the BritishandForeignStatePapers,Vol.

82 (seeAnnex 12). The text appearsalso in Hertslet, TheMapofAfnca by Treaty,3rd ed., 1909, Vol. DI, p.899 (the English
text). The Treaty entered into force immediately upon
signature: see An Inda of British Treaties1101-1968, Parry
and Hopkins, London, H.M.S.O., 1970, Vol. 2, p.428.

114. The materid part of the text (in English) is as
follows:-

"III I.South-westAfrica the sphere in which the exercise
of influence is reserved to Germany isbounded:-

1. To the south by a line commencing at the mouth of the
Orange River, and ascending the north bank of that river
to the point of its intersection by the 20th degree of east
longitude.

2. To the east by a line commencing at the above-named
point, and following the 20th degree of east longitude to

the point of its intersection by the 22nd parallel of south
latitude, it runs eastward along that parallel to the point of
its intersection by the 21st degree of east longitude; thence
it follows that degree northward to the point of its
intersection by the 18th parallel of south latitude; it runs
eastward along that paralleltill it reaches the River Chobe;

and descendsthe centre of themain channelof that river
to itsjunction withthe Zambezi, where it terminates."
(emphasis supplied)

115. The German text employs the term Thalwegas
the counterpart to 'themain channel' in the English text. In due
course the connection between these terms and the purpose of

such a provision (topermit access by both riparian States to the
navigable channel) will be elaborated.(ii) The Position of Botswana

116. The position of the Governrnentof Botswanais

that the main channel of the Chobe in the vicinity of
Kasikili/SeduduIsland isthe northern and westernchannel, the
principal criterion on which this assessmentis based being that
of navigability. In the absenceof evidence to the contrary, the
presumption must be that this was also the main channel at the
time of the conclusionof the Anglo-Gennan Agreement.

117. The alternative position adopted by the
Governmentof Botswanais that, in accordancewith the object
and purpose of the Agreement, the main channelis constituted
by the navigablechannelat any giventime, and that at present
the northem and western channel is the main channel on this
basis.

(iii) Interpretation in accordance with the object and
purpose of the Agreementof 1890

118. The Vienna Convention on the Law of Treaties
provides as follows in Article 31:-

"Generalruleof interpretation

1. A treaty shall be interpreted in good faith in
accordance with the ordinary meaning to be given to the
terms of the treaty in their context and in the Lightof its
object and purpose.

2. The contextfor the purpose of the interpretationof
a treaty shall comprise,in additionto thetext, including its
preamble and annexes:

(a) any agreement relating to the treaty which was
made between al1the parties in connectionwith the

conclusionof the treaty; (b) any instrument which was made by one or more
parties in connection with the conclusion of the
treaty and accepted by the other parties as an
instrument related to the treaty.

3. There shallbe taken into account, togetherwiththe
context :

(a) any subsequent agreement between the parties
regarding the interpretation of the treaty or the

applicationof its provisions;

(b) any subsequent practice in the application of the
treaty which establishes the agreement of the
parties regarding itsinterpretation;

any relevantrules of internationallaw applicablein
(c)
the relations between the parties.

4. A special meaning shall be given to a term if it is
establishedthat the parties so intended."

119. As the Report of the International Law

Commissionto the GeneralAssemblyindicated, the successive
paragraphs were not to be taken as layingdown a hierarchical
order for the application of the various elements of
interpretationin the article (Yearbookofthe International Law
Commission, 1966, II, pp.219-20 (paragraph 8)).Moreover,
'the Commission desired to emphasise that the process of

interpretation is aunity and that the provisions of the article
form a single, closely integrated rule.' (ibid).

120. Neither Botswananor Namibia is a party to the
Vienna Convention on the Law of Treaties. However, it is
generally accepted that the principles set forth in Articles 31
and 32 of the Convention are declaratory of general

international law. This view has been adopted by thejurisprudence of international tribunals.

121. In the CaseConcemingtheArbitralAwardof 31

July 1989 the Court referred to certain principles of
interpretationand then observed:

"Theseprinciplesare reflectedin Articles31 and 32of the
Vienna Conventionon the Law of Treaties, which may in
many respects be consideredas a codificationof existing
customary internationallaw on the point."

(IC.J. Reports, 1991, p.53 at pp.69-70(para. 48)).

122. IntheELSICaseJudgeSchwebelapproached the
questionof interpretationon the following premises:

"In its pleadings, Italy relied upon the rules of treaty

interpretation set forth in Article 31 of the Vienna
Convention on the Law of Treaties, as reflective of
customary international law, a position which was not
questionedby the UnitedStates. Article 31 provides that,
"Atreaty shall be interpreted in good faith in accordance
with the ordinary meaningto be givento the terms of the
treaty in their context and in the light of its object and

purposes. "It providesthat "Thecontextfor the purpose of
the interpretationof a treaty shall comprise, in additionto
the text,...its preamble ..." It specifiesthat there shallbe
taken into account, together with the context, "any
subsequent agreement between the parties, regarding the
interpretation of the treaty or the application of its
provisions". And it provides in Article 32 that:

"Recoursemay be had to supplementarymeans of
interpretation, including the preparatory work of
the treaty and the circumstancesof its conclusion,
inorder to confii the meaning resultingfrom the
application of article 31, or to determine the

meaning when the interpretation according to article 31:

leaves the meaning arnbiguous or obscure;
or)

(b) leads to a result which is manifestly absurd
or unreasonable."

In the cunent case, the Parties attached radically
different interpretations to the provisions of the
Treaty and its Supplementary Agreement which

were at issue between them. It is undeniable that,
when their conflicting arguments are matched
together, the meaning of some of the Treaty's
provisions are ambiguousor obscure; indeed, each
of the Parties maintained that the opposing
interpretationled toresults which, if notanifestly
absurd, were unreasonable.Thus, according to the

ViennaConvention,this is a casein whichrecourse
to the preparatory work and circumstances of the
Treaty's conclusionswas eminently in order."
(Dissenting Opinion of Judge Schwebel, Case
Conceming Electronnica Sicula S.p.A (ELSI),
I.C.J. Reports, 1989, p.97)

123. Other international tribunals have adopted the
sameview of Articles31and 32, as in the following examples:

(a) The Inter-Amencan Court of Human Rights in
Restrictions of the Death Penalty, Advisory

Opinionof 8September1993,InternationalLaw
Reports, Vol. 70, p.449 at pp.465-6 (paras. 48-
49).

(b) The European Court of Human Rights in
Lithgow and Others, Intemutional LawReports,

Vol. 75, p.439 at p.483 (para. 114). (c) The Iran-United States ClaimsTribunal in Case
No. A/18; International Law Reports Vol. 75,
p. 176at pp.187-8.

124. In any event the principle adopted inparagraph
3(c)of Article31, that account shouldbe taken of 'any relevant
rulesof internationallawapplicableinthe relationsbetweenthe
parties', haseen generally recognisedas a principleof general
international law both before and after the conclusionof the
Viema Convention.

125. In his examination of the problems of treaty
interpretation Judge Hudson states the following (under the
rubric 'legal background'):

"Any international instrumentmust be interpreted in the
light of the prevailing international law, by which the

parties must be taken to have charted their course."

(ThePemuznentCourtofInternational Justice192û-1942,
New York, 1943, p.655 (para.573)).

126. The Judgment of the International Court in the

Right of Passage Case (Preliminary Objections) contains the
following affirmation: 'It is aule of interpretationthat a text
emanating from a Govenunent must, in principle, be
interpreted as producing and as intended toproduce effects in
accordancewith existinglaw and notin violationof it'. (I.C.J.
Reports, 1957, p.125 atp.142).

127. The passagejust quoted was prefigured by one
of the 'rulesof interpretation'setforth in the eighth editionof
ûppenheim, edited by Sir Hersch Lauterpacht. The third rule
is as follows:

It is taken for granted that the contracting parties
"(3)
intendsomethingreasonableand somethingnotinconsistent with generally recognised principlesof InternationalLaw,
nor with previous treaty obligationstowards third States.
If, therefore, theeanhg of a provisionis ambiguous, the

reasonable meaningis to be preferred to the unreasonable,
the more reasonable tothe less reasonable, the consistent
meaning to the meaning inconsistent with generally
recognised principles of International Law and with
previous treaty obligations towardsthird States."

(Oppenheim 's International Law, Vol. 1 Peace, 8th ed.,
London, 1955, pp.952-3).

128. The sameprincipleis reformulatedbySirHersch
in TheDevelopmentof International Law by the International

Court of Justice (London, 1958) :-

"In fact, it would bea mistaketo assumethat the function
of interpretation of treaties, consisting as it does in
ascertaining what was the intention of the parties, is a
process divorced from the applicationand developmentof

customaryinternational law. The elicitingof the intention
of the parties is not normally a task which can be
performed exclusivelyby meansof logicalor grammatical
interpretation. As a rule, the established canons of
construction - which themselvespartake of the nature of
customary law - must be supplemented by the principle

that whenthe intentionof the partiesis notclear it mustbe
assumedthat they intendeda result which is in conformity
with generalinternationallaw. Undoubtedly,conventional
international law may derogate from customary
internationallaw, but it is no lesstrue that the formermust

be interpreted by reference to international custom. In
many a caseof treaty interpretationthe effect of the treaty
will depend on our view as to the position of customary
internationallaw on the question."
(pp.27-28) 129. The outcome is that there is a presumption that
the parties to an agreement intendeda result which was in

conformity with general international law. In relation to the
relevant provisionsof the 1890Agreement(Article III (2)the
pertinent principlesof general internationalLaware those
relating to the detemination of boundanes in navigable
watenuays. But before this aspectof the question ispursued, it
is necessary to address the question:what was the object and
purpose of the Agreementof 1890overall?

130. The object and purpose was in essence a
partition ofseveralimportant regionsof Africa into spheresof
influenceof GermanyandGreat Britainrespectively,with some
territorial 'compensation'for Germanyelsewhere.

131. The relevant diplomaticdocumentsalso indicate

a connected purpose, which was to maintain and facilitate
accessto the principallakes andriversof the Africancontinent.
This significantaspectof the exercisein territorial divisionand
boundary-makingconsistsof two distinctbut related elements.
The first of these consisted in the general interest which the
European powers had in 'free navigation' in major African
rivers. This interest was manifested in the General Act of the

Conferenceof Berlin, signedon 26February 1885 (Annex 1).
This instrumentincluded an Act of Navigationfor the Congo
and an Act of Navigation for the Niger. The importance of
freedom of navigationto the Europeanpowers was evident in
the proceedingsof the conferenceand the theme wascontained
in Bismarck's opening speech: see Geoffrey de Courcel, in
Forster, Mommsen andRobinson(editors), Bismarck, Europe,

and Afnca, The German Historical Institute London, Oxford
University Press, 1988, pp.252-4.

132. The interestin access to Africanriversand lakes
is evidentin the contentsof the collectionsof documentswhich
constitutethemainelementsin the travauxpréparatoire of the

Anglo-German Agreementof 1890. The principal collection isthe Correspondence respectin tgeNegotianonsbenveenGreat
Britain and Gemuznyrelanngto Afnca, April to December
1890 (set forthin extensoin Annex 9).

133. In this collection of correspondence relating to
the negotiations of the Anglo-German Agreement there are
manyreferencesto the issueof accessto andfree navigationon
the major African rivers and lakes. One particular theme was
access to Lake Ngarni: see the Correspondence (Annex 9) at
pages 2 (No. 2), 18 (Inclosure 1 in No. 27), 27-28 (No. 41,

para. 2), 29 (No. 43), 29-30 (No. 49, 32-33 (No. 48, para.
II), 52 (Inclosure 1 in No. 84, Article III), and 53-54
(Inclosure 2 in No. 84, ArticleII).

134. A concomitanttheme wasthat of Germanaccess
to the Zarnbezi, and the outcome was the acquisition by
Germany of the Caprivi strip as a part of her sphere of

influence. The theme of German access appears in the
Correspondence (Annex 9) at pages 27-28 (No. 41, para. 2),
32-33(No. 48, para. II), 52 (InclosureNo.1in No. 84, Article
III), and 53-54 (Inclosure 2 in No. 84, Article III).

135. The theme of free access by Germany to the
upper watersof the Zarnbezihad first appearedin the exchange

of proposalswhichhad takenplace in September 1889: seethe
Marquis of Salisburyto CountHatzfeldt on 2 September 1889
(Annex4), and Count Leydento the Marquis of Salisbury,30
September 1889 (Annex5).

136. It is of interestthat the successivedraftsreferred
to 'the centre of that river' (i.e. the Chobe) until the Marquis

of Salisbury proposed the inclusion of the words 'the main
channel of' in a telegram to Sir E. Malet on 25 June 1890:
Correspondenc eAnnex 9)(atpage. 63). Thisproposalappears
to have been accepted by the German side without any
oppositionor expressionof doubt. In this way Article(2) of
the Treaty emerged in its finalorm. 137. The reference to the Chobe in Article III(2)
occursin closeassociationwiththe Zambezi andit is clear that

accessto waterwaysandthe generalquestionof navigationwas
a matterof majorconcem. The Zambeziandits majoraffluents
were considered to be navigable, and it is in this context that
the pnnciples of contemporary general internationallaw are of
great importance. They establishthat in the case of navigable
rivers the middle of the navigable channelwas the boundary.
Indeed, the contemporary legalsources reveal the background

of the drafting of the relevant part of Article III(2).

138. Bothbeforeand afterthe conclusionof the 1890
Agreement it was generally recognised that in the case of
navigable rivers the middleof the navigable channel(thalweg)
was the boundary betweenthe riparian States.

The relevant authorities will be rehearsed in chronological
order as follows.

(a) G.F. de Martens, Précisdu Droit des Gens
moderne de l'Europe, edited by Ch. Vergé,
Paris, 1864, Tome 1, pp.135-6.

(b) J.-L. Klüber, Droit des Gens moderne de
1'Europe,revised by M. A. Ott, secondedition,
Paris, 1874, pp.187-90 (para. 133) (see, in
particular, pp.188-9).

JC . Bluntschli,Le Droit International CodiJié,
(c)
translated from the German, second edition,
Paris, 1874, pp.181-2 (para.298).

(d) J.C. Bluntschli, Das moderne Volkerrechtder
civilisienen Staaten, Nordlingen, 1878, p. 180
(para.298).(e) F. de Martens, Traitéde Droit International
translatedfromthe Russian,Paris, 1883,p.456.

(f) Sir Travers Twiss, The Law of Nations

considered as Independent Political
Communities, second edition, Oxford, 1884,
pp.249-50 (para. 153).

(g) William Edward Hall, A Treatise on
International Law, second edition, Oxford,
1884,pp.114-15 (para.38).

(h) Charles Calvo, Dictionnaire du Droit
International Public etPrivé ,erlin, 1885,
Tome II, pp.256-7.

(i) Quaritsch, Compendium des Europaischen

Volkrrechts, Berlin, 1885,p.27.

(j) Franz von Holtzendorff, Handbuch des
Volkerrechts.Auf der GrundlageEuropaischer
Staatenpraxis, vol. II, Hamburg, 1887, p.303
and p.304,n.6.

(k) Peter Resch, Das Volkerrecht der heutigen
Staatenwelt Europaischer Gesittung , second
edition,Graz & Leipzig, 1890, p.70 @ara.65).

(1) William Edward Hall, op. cit., third edition,
Oxford, 1890,p.125 (para.38).

(m) William Edward Hall, op.cit., fourth edition,
Oxford, 1895,p.127 (para.38).

(n) AlphonseRivier, Principes duDroit des Gens,
Paris, 1896, Tome 1, p.168. (O) Frantz Despagnet, Coursde Droit International
Public, secondedition, Paris, 1899, p.407.

(p) William Edward Hall, op.cit., fifth edition by
J.B. Atlay, Oxford, 1904,pp. 122-3.

(q) Franz von Liszt, Le Droit International, French
translationof the ninth German edition of 191 3

(by Gilbert Gidel), Paris, 1928, p.86.

139. Only two years prior to the conclusion of the
Anglo-German Agreement the Institute of International Law
had adopted a resolution at its Heidelberg session conceming
theregulationof fluvialnavigation.Thefirst threearticles were
as follows:-

"ARTICLE PREMIER - Les Etats riverains d'un fleuve
navigable sont obligés, dans l'intérètgénéral,de régler
d'un communaccord tout ce qui a rapport à la navigation
de ce fleuve.

ART. 2 - Les affluents navigables des fleuves
internationauxsont, à tous égards,soumisau mèmerégime
que les fleuves dont ils sont tributaires, conformément à
l'accord établi entre les Etats riverainset au présent
réglement .

ART. 3 - La navigation dans tout leparcours des fleuves
internationaux,dupointouchacund'euxdevientnavigable
jusque dans la mer, est entiérementlibre et ne peut, sous
le rapport du commerce, êtreinterdite à aucun pavillon.

La frontière des Etats séparéspar le fleuve est marquée

par le thalweg, ce'est-à-dire par la ligne médianedu
chenal."

(Annuairede 1'Institut, 1887-88, p.182). 140. The membership of the Institute at this period
included persons closely associatedwith the actual practice of
diplomacy. It may be recalled that since 1876and particularly
in the years 1884to 1890the questionof free navigationon the

Zambezi and other major African rivers had been the subject
of considerable diplornatic controversy: see Clive Parry
(editor),A British Digest of International Law, Phase 1, Vol.
2b, London, 1967, pp.176-90.

141. Consequently , in theperiod inwhichthe Anglo-

German Treaty was concluded, the reference to the centre of
the deepestchannel was very clearly linked to the navigability
of the river concernedand the interestsof theiparian Statesin
navigation. This connection between the Thalweg and
navigationcontinuedtobe recognisedin the doctrine ofthe law
in subsequentdecades, as the following sources indicate:G.H.

Hackworth, Digest of International Law, U.S.G.P.O.,
Washington, Vol. 1, 1940, pp.570-5; Oppenheim's
International Law, Vol. 1, Peace, eighth edition by Hersch
Lauterpacht, London, 1955, p.532.

142. Referencemay also be madeto theAward ofthe

King of Italywith regard to the Boundarybetween the Colony
of British Guiana and the United States of Brazil, given at
Rome, 6 June 1904, where the Award States:-

"The frontier dong the Ireng (Mahu) and Tukutu is fixed
at the "thalweg" and the said rivers shall be open to the

free navigationof both conterminousStates". (Britishand
Foreign State Papers, Vol. 99, p.930 at p.932).

143. This connection between the Thalweg concept
and navigation is indicated by a leading French treatise: see
Rousseau,Droit internationalpublic, DI, Paris, 1977,pp.255-

6, where the extensive treaty practice is set forth, including
some of the practice relating tofrica. 144. The conclusion is thus that the object and
purpose of the provisions of Article m(2) of the Agreement
was two-fold:

(a) To affm the rights of navigation of the two
self-appointedriparian States in respect of the
Chobe; and

indoing so to applythe contemporary standards
(b)
of general international law relating to
boundariesin navigablerivers.

(c) The Subsequent Conduct of the Parties as
ConfirmatoryEvidence

(i) The Principle

145. The conductof the parties to a treaty may have
probative valueas a subsidiary methodof interpretation. As
the Court will recall, such evidence of conduct may be
considered as co~iatory or corroborative evidence of the
parties to the treaty at thetirne of its conclusion : see the
ReportoftheCourtof Arbitration,Beagle ChannelArbitration,

1977, International Law Reports, Vol. 52, pp.220-26,
paras.164-75 ; Fitzmaurice, British YearBook of International
Law, Vo1.33(1957),pp.223-5; Thirlway,ibid., Vo1.62(1991),
pp.48-57.

(ii) CaptainH.V. Eason'sSurvey, 1912

146. In 1911 CaptainH.V. EasonoftheBechuanaland
Protectorate Police was ordered to carry out a reconnaissance
of the River Chobe and to report on the main channel. His
report, dated 5 August 1912, is entitled 'Report on the Main
Channel of the Linyanti (or Chobe) River'. In the Report he
States: 'Two miles above the rapids lies Kissikiri Island. Here 1
consider that undoubtedlythe north shouldbe claimed as
the main channel. At the western end of the island the
North Channel at this period of the year is over one

hundred feet wide and 8 feet deep, the South Channel
about forty feet wide and four feet deep. The South
Channelis merelya back water, what current there is goes
round the North'.
(Annex 15)

147. The Eason Report was commissioned at the
highest level and onginated in a despatch dated 14 Januq
1911 from Lord Harcourt to the High Commissioner(Annex
16). The Report wasaccompaniedby a seriesof maps (Annex
15). The Report was forwardedto the Secretary of State in a
despatch dated 23 September1912. (see Annex 18 for the
reference).

148. The Eason Report remained in the Archives in
Gaborone and is quoted in the Boundary Survey of July 1985,
the status and significance of which are considered later in
Chapter VI of this Mernorial. The reliability of the Eason
Report has at no tirne been questioned intemally and the

reference in the Report of the Survey of 1985 confis the
reliability of Eason's work and places it on the public record.

(iii) The Joint Report by Troiiope and Redman, 19
January 1948

149. TheBotswanaArchivescontainacopy ofa 'Joint
Report' by Messrs. L.F.W. Trollope and N.V. Redman,
respectively Magistrate of the Eastern Caprivi Zipfel, and
District Commissionerat Kasane, BechuanalandProtectorate' .
(Annex 22). The key paragraphs of the Joint Report, which
was signed on 19January 1948, are as follows:- '2. We attach heretoa sketchmap (not drawnto scale)

of the Kasikili Island in the Chobe River and the
waterwaysrelativethereto.

3. We fmd after separate examination of the terrain
and the examinationof an aerialphotographthat the "main
channel" does not follow the waterway which is usually
shown on maps as the boundary between the two

Tenitories.

4. Weexpressthe opinionthatthe "mainchannel"lies
in the waterwaywhichinclude the islandin questionin the
Bechuanaland Protectorate'

150. The Reportconcludeswiththefollowingproviso:

"8. We record the foregoing facts particularly
recording that we have neither amved at, nor expressed
any joint opinion on the effect of those facts on the

question of the ownershipof the island."

151. It is reasonableto assumethat the preparationof
the Joint Report resulted from a mandate from the two
Governments,but no documentto thiseffecthas been seen. At
any rate, when the Report was sent to Pretoria, the South
African Governmentdid not seek to repudiate it: see below,

paragraphs 153to 156.

152. The Joint Report is of particular significance
because it resulted from a practical question concerning the
commercial importance of navigability. As Mr Redman

reported (Letter dated 26 January 1948)-

'1have the honour to attach a joint report by the Native
Commissioner, Eastern Caprivi Strip and myself
concerninga disputewhich has arisen over the ownership of the island shown on the enclosed sketchmap.

2. The questionhasarisenas a result of anapplication
by Mr Ker to transport timberby barge from Serondelato

Katombora, which necessitates the use of the channel
running to the North of KasikiliIsland since the Southem
Channel is not navigable by his Barges when the river is
not in flood, and it is even difficult for small craft to
navigate it.
(Annex22)

(iv) Official Admissions by the South African
Government, 1948-49

153. The appearance of the Trollope-Redman Joint
Report led to a correspondence between the British High
Commissioner's Officeand the South African Govemmentin

the course of whichthe latter conceded the principle that the
main channel of the Chobe was the northem channel.
Moreover, this concessionwas made in the particular context
of the needs of river transport and navigability.

154. On the 14 October 1948 a letter was addressed

on behalf of the South African Govemment to the High
Commissioner's Officein Pretoria as follows: -

'1have the honour to enclose a copy of a Joint Report by
the Magistrate of the Eastem Caprivi Zipfel and the
District Commissioner at Kasane, Bechuanaland

Protectorate, regarding the boundary between the
Protectorate and the Eastem Caprivi Zipfel.

It is understood thatthe necessityfor considerationof the
matter arises from the fact that a certain river transport
venture, whichproposesto transport timberdownthe river

from a sawmillin Bechuanalandhas raised the questionof
the correct boundary both in representations to the Magistrate, Eastern Caprivi Zipfel, and to the
Bechuanalandauthorities.

The Report discloses that while the main channel of the
Chobe River is shownon maps as passing to the Southof
KasikiliIsland it infact passesto the North of that Island.

It has beenconfmed, as a result of exhaustive enquiries,

that there has been no shiftingof the main channel of the
river from South to North within living memory. The
facts, therefore, point to the maps being incorrect.

As against the foregoing there is evidence that the Island
has been cultivated by Caprivi Tribesmen since at least
1907andthat theirright tothe occupationof the Islandhas

at no tirne been disputed.

The UnionGovemmentisanxiousto preserve the rightsof
the Caprivi Zipfel tribesmen on the Island and it is
understoodthat theBechuanalandauthoritiesdesirethe use
of the Northem channelfor navigationpurposes. As there

would appear to be no conflict of interests it should be
possible to come to an anangement which is mutually
satisfactory. Your views in the matter would be
appreciated.'
(Annex 23)

The Governmentof Botswanareserves its position in relation
to the final two paragraphsof this document.

155. A furtherSouthAfricanletterdated 14February
1949 underlines Pretoria's acceptanceof the view that the
northem channel of the Chobe is the channel which is

significantfor navigationpurposes. Thefuiltext isasfollows:- '1have the honour torefer to your letter No. 9628of the
4th November, 1948 (Annex 24), regarding the boundary
between the Bechuanaland Protectorate and the Eastern

Caprivi Zipfel.

While noting that your Administration is prepared to
authorise Caprivi Zipfel tribesmento cultivate land on
Kasikili Island on an annual renewable permit, 1 am to
state that this is not what the Union Governmenthad in

mind.

From the available information it is clear that Caprivi
Tribesmenhave madeuse of the Islandfor a considerable
numberof yearsand thattheir right to do sohasat no tirne
been disputed either by Bechuanaland Tribesmenor the
Bechuanalandauthorities.

It was further understood that the interests of the
BechuanalandauthoritiescentredintheuseoftheNorthern
Channelof the Chobefor navigation purposes.

My object in writing to you was therefore to ascertain

whether agreement could not be reached on the basis of
your Administration recognising the Union's claim to
KasikiliIsland subjectto it issuing ageneralpermit for the
use of the Northernwaterwayfor navigationpurposes.'
(Annex 25)

156. The outcome was a South African proposal
reported in a letter from the High Commissionerdated 6June
1949 (see Annex 26) thatthe boundarybe movedto the South
channel on the basis that Bechuanalandbeguaranteedthe use
of the northern channel for navigation. In fact the proposal
was not proceeded with in the light of the advice of the
Commonwealth RelationsOffice (see letters dated 24 August

1949(Annex 27), 20 October 1949 (Annex28), 19November
1949 (Annex 29) and 10 May 1951 (Annex 30)). TheGovemmentofBotswanareliesuponthis documentexclusively
in relation to its assumption that itis the Northem channel

which is significantfor navigation.

(v) The BritishHigh Commicsioner'sAff"mation of the
Legal Status Quo, 10 May 1951

157. The letter dated 10 May 1951 (Annex 30)

representsthe consideredofficial British reaction (in the light
of legal considerations)to the South African proposai for a
modificationofthe boundary.In materialpart the letter state-:

'3. Thepossibilityof makinga declarationonbehalfof
the Governmentof the BechuanalandProtectorate to the
effect that the Island is not claimed as lying within the

boundaries of the Protectorate has been examined by the
Legal Advisers to the Secretary of State for
Commonwealth Relations. 1 am afraid that they have
found this proposal to be beset by legal complicationsof
an international nature,the solutionof which would entail
difficultiesdisproportionatetothe importanceofthematter

at issue.

4. The BechuanalandProtectorate Govemment might
possibly wish to arrange for some land on the Island at
some time to be cultivated by the few African public

servants at Kasane. Apart from this minor matter, 1
venture to suggest thatit is unlikely that any development
in the foreseeable future will damage the interests of the
Caprivi tribesmen, who have in the past used the Island.
It should, 1 think, be possible to adjustby administrative
action any difficulty arisingin connectionwith the Island
and the adjacent waterway without an alteration of the

existing legal position.

5. The instruction to the Assistant District
Commissioner, Kasane,ofwhichyouwereinformedinMr Priestman's letter No. 9628 of the 4th November, 1948
would be maintained, and itis assumedthat the free use of
the main channel of the Chobe, to the north of the Island,

would continue to be assured under the internationalrules
governing waterways that formthe common boundary of
two States.'

158. Paragraphs 4 and 5 of this important document
are concernedto affm the 'existing legalposition' conceming
the commonboundq. Paragraph5, in particular, affms the

status of the northern channel as the main channel for legal
purposes. The letter was addressedto Mr Forsyth at the South
African Department of External Affairs by the High
Commissioner,Sir Evelyn Baring. The South African response
on 29 May 1951 was as fo1lows:-

'Thank you for your letter No. 9628 of the 10th May
1951, regarding Kasikili Island in the Chobe River.

1note that there are difficultiesin the way of recognising
the Union's claim to the island and will inform the
appropriate authorities accordingly'
(Annex 31)

159. This sequenceof exchangesis to be understood
on the basis that it was the South African side which was
proposing a modification of the legal status quo. The
exchanges of the period 1948-51 also involve South African
acceptance of the principle that the main channel lies to the

north of Kasikili/Sedudu.

(vi) The Opinion of the Surveyor-General of
Bechuanaland, 18 October1965

160. The question of the boundary in the vicinity of

KasikiliISedudu Island was raised again within the
administration of Bechuanaland, apparently in relation to theestablishmentof the boundaries ofthe Chobe GameReserve.
In this connectiontheSurveyor-Generalofthe Protectorate, Mr

R R Renew, was consulted and the most material part of his
response to the Permanent Secret., Ministry of Home
Affairs, reads thus-

'1 have gone into the record thoroughly and have
summarisedthe history of the island. If further action is

contemplatedin regard to the use of this island itmightbe
advisablefor you to see the file.

2. Kasikili island becarnethe subject of a dispute in
1947whenthe NativeCommissionerof the eastemCaprivi
Strip was allegedto have chailengedBechuanaland'sright

to the use of the main channel of the Chobe River along
the north side of the island, as a waterway.

3. At that time, the factual position acceptedby both
Governments, andembodiedin a report submitted jointly
by the District Commissioner, Kasane and the Native
Commissioner of the eastem Caprivi Strip, Major

Trollope, was as follows:

a) The boundary between Bechuanaland and the
relevant portion of the CapriviStrip was the centre of the
main channel or Thalweg of the Chobe River, and this
main channel ranto the northof the island. Thistherefore

placed Kasikili Island on the Bechuanaland side of the
intemationalboundary.

b) Since the assumptionof German administrationof
the Caprivi Strip in 1907, Caprivi tribesmen have
cultivated lands and generally had the undisputed useof

the island.

c) There was no evidence of the island havingbeen
made useof, or claimed, by Bechuanalandtribesmen. 4. When the dispute arose Government logically
asserted ownership by virtue of paragraph 3(a), whereas
the South African Govemment based its claim to the island

on prescription in terms of paragraph3(b).'
(Annex 36)

161. TheSurveyor-General'sOpinion is accompanied
by a map of the relevant sector of the Chobe on which is the
inscription: 'Main Channelcoloured BLUE' (this accompanies
Annex 36) (se Appendix II of the Memorial, Map 18). The

Govemment of Botswana relies upon this document but
reserves its position in relation to paragraph 3(b) and (c).

(vii) The Botswana-South AfricanAgreement Concluded
at Pretoria on 19 December 1984 and the Joint
Surveyof 1985

162. The subsequent conduct of the Parties, and their
successors, must include the significant transactions of 1984
and 1985 in which the Republic of Botswana and the Union of
South Africa agreed to conduct a Joint Survey, the results of
which were approved by the Govemments. The survey

confirmed that:

"The main channel of the Chobe River now passes
Sidudu/Kasikili Island to the Westand to the north of it.
(See annexed Map C).

The evidence available seems to point to the fact that this

has been the case, at least, since 1912."
(Annex 48)

163. The legal significance of these inter-
govermental transactions will be explored further in the
following chapter of theMemorial.(d) Conclusion

164. In the light of the continuityof the boundary in

accordance with the principles of State succession and uti
possidetis, the subsequent conduct of the British and South
African Govemments, as predecessors (subject to the
obligationsarising from the Mandatefor South-westAfrica) to
the present riparian States,has evident signifïcance.

165. Boththe Britishand SouthAfrican Governments
were of the opinion that the main channel in the vicinity of
KasilcililSedudupassed to the north of the island. It was
against this background that in 1949 (see Annex 26) South
Africa proposed a rnodzjicationof the boundary on the basis
that Bechuanaland be guaranteed the use of the northem

channelfor navigation. CHAPTER VI

The PretoriaAgreementof 1984 and the JointSuwey of
1985

(a) The Background

166. In October 1984 the Botswana Defence Force
opened fire on South Africansoldiers who were travellingby
boat on a smallchannelof the Chobe Riverwhich runs south
of the Kasikili/SeduduIsland. Three South AfricanDefence
Force soldierswere injuredand South Africaclaimedthat their

soldiers had been shotat on the South-WestAfrica side of the
border.

167. BeforeBotswanacouldhold talkswiththe South
African authorities on the shooting incident, a Botswana
delegation was sent to New York to consult with the United
Nations which was the de jure legal authority for Namibia

(Annex 43). They met the President of the United Nations
CouncilforNamibia,Mr. Paul Lusaka,on27November 1984.
(The Minute sf the meetingare set forth in Annex 42). The
BotswanadelegationconsistedofMr. M.D. Mokarna,Attorney
General,Mr. G.G. Garebamono,Secretaryfor ExternalAffairs
and Mr. L.J.M.J. Legwaila, Permanent Representativeto the

UnitedNations.Mr. MokarnarecalledthatBotswanahadinthe
past avoidedtalking to the South Africans on Namibia before
consultingwith the United NationsCouncilfor Namibia.

168. In reply Mr. Lusaka thanked the Botswana
delegation for involving the United Nations Council for
Namibiaon matters affectingNamibia.He saidthat the Council

was sympathetic to Botswana but that he would first oall
consult with membersof the Council before giving reply.The
Botswanadelegation left both the SouthAfricanand Botswana
maps withMr. Lusaka for ease of reference. 169. Theyalsomet withSouth-WestAfricanPeople's
Organization(SWAPO) officiais. ThMinute osf the meeting
appear in Annex 41. The SWAPO delegation consisted of

Andimba Toivo Ja Toivo, Secretary-General, Peter
Mueshihange, Secretary for External Affairs and Theo-Ben
Gurirab, Permanent Observer at the United Nations. Mr.
Garebamonoreported that during the Organizationof African
Unity Summit in Addis Ababaon 12-15November, 1984, the
Minister for Externalffairs, Dr. Chiepe, had met with Mr.
Sam Nujoma, the SWAPO President, and had informed hirn

about the dispute betweenBotswanaand South Afnca on the
Botswana-Namibia boundary. She had explained that a
Botswanadelegationwouldpay a visit to New Yorkto consult
with SWAPO andthe Councilfor Namibiabefore meetingthe
SouthAfricans to discussthe boundaryproblem.

170. Mr. Nujoma had observed that it is the "main
channel" of a riverwhich is regarded internationallyas the
boundary. Mr. Toivo observedthat SWAPO asa party and a
liberation movementdid nothave jurisdiction over the border
issue. He enquired if the President of the United Nations
Council for Namibia had been consulted on the matter. Mr.
Mokama confmed that this had been done. He therefore

agreedthat Botswana could holdtalks with the SouthAfricans
on the boundary issue. The meeting further discussed other
issues not related to the boundary. Lastly, the Botswanaand
South African maps concerned were handed over to the
SWAPO delegationfor their informationand records.

171. At its Tenth Summitin 1973the Organizationof

African Unity granted some National Liberation Movements
representationstatus and SWAPO was granted such a status.
The position of SWAPO vis-à-vis South-West Africa was
therefore that SWAPOwas recognisedas being delegatedby
the South-West Africa People to fight for the right to self-
determination on theirbehalf. SWAPOthus gave its blessing

to Botswana so that she could discuss with South Africa thedispute over the boundary around the KasikiliISeduduIsland.
The approvalof the UnitedNations Commissionerfor Namibia
was also obtained (Annex43).

(b) The Meeting in Pretoriaon 19 December1984

172. A meeting was thereafter convened by the
Botswana delegationand the South African delegation, which
meeting was held in Pretoria on 19 December 1984. The
Botswana delegationwas composed of Mr. M.D. Mokama,

Attomey-General, Mr. G.G. Garebamono, Secretary for
Extemal Affairs, ColonelRankhudu,BotswanaDefenceForce,
and Mr. J.A Raffle of the Department of Surveys and Lands.
TheSouthAfricandelegationincludedMr. A.S. Mare, Foreign
Affairs(Chairman),Mr. W. Hugo, Foreign Affairs,Mr. D.W.
Steward, Foreign Affairs, Mr. J. Rautenbach, Legal

Department, Mr. J.F. Kirsten, Foreign Affairs, Mr. E.
Fitschen, Public Works and Lands Affairs, Brigadier J.A.
Klopper, South African Defence Force, BrigadierG. Nel,
South African Defence Force, Brigadier A.N. Heuer, South
African Police, and Lieutenant Colonel G.J. Richter of the
South African Police.

173. TheBotswanaNationalArchivescontaina setof
Minutesof the meetingprepared by the Botswanaside (Annex
44). In addition, there is available a 'transcript' submittedto
Botswana by the South African side (Annex 46). In response
the Governmentof Botswanastated:

"The Departmentof Extemal Affairs wishesto advisethat
the transcriptis basically correct but that because of its
brevity it left out certain essential details such as a
reference to the maps and air photographs brought to the
meeting by the Botswana delegation which were used in
the discussions."

(Annex47). 174. At the meetingit wasdiscovered thatmaps used
by the SouthAfricanshad showedKasikiliISeduduIslandto be

on the South-West Africa side of the boundary because they
wereassumingthat the smallsouthernchannelof the Riverwas
theboundary. On the mapsusedby Botswanathe mainchannel
was indicatedas the boundary .

175. Mr. Mokama stated that there was no wish on

the part of Botswana to expand Botswana; and that the
historical backgroundto the boundary was well documented
and was based on the Anglo-GermanAgreement of 1890. He
went on further to state that the question of the thalweg,
navigation andthe original ideaof Count Von Caprivi that the
river would give access to the Indian Ocean, had to be
considered.Mr. Mokama statedthat the "main channel"ranto

the north ofKasikili/SeduduIsland.

176. Mr. Fitschen said that his Directorate had
worked closely withthe Departmentof Surveys and Lands in
Botswanaon common boundarymatters and that an excellent
working relationship had been built up due to the high

professional standardson both sides. He went on further to
state that river boundaries could change and that these
presented problemsof a technicalnature.

177. Mr. Rautenbach suggested that the Island
problem should first be investigated and that instructionsbe

issuedfor the surveyto be extended at alater date. Mr. Raffle,
ably assistedby Mr. Kirsten, displayedthe relevant Botswana
1:50,000 map sheet together with aerial photographs of the
Island dated 1925, 1972 and 1981. A proposal to inspect the
maps in order to appreciatetheproblem on hand wasmadeby
Mr. Garebamono. The meeting broke up to inspect the

exhibits. As a result delegates from South Africappeared to
have understoodwhat the real situationon this matter was. 178. The meeting agreed that a joint survey should
take place as a matter of urgency to determine whether the
"main channel" of the Chobe River was located to the north or

to the south of Kasikili/Sedudu Island. It was further agreed
that officials of the Botswana Department of Surveys and Lands
and the South African Directorate of Surveys and Mapping
would proceed with arrangements for the survey as soon as
practicable.

179. In Note No. 18EA6/4 VI1 (94)A (Annex47) the

Department of Extemal Affairs of the Republic of Botswana
indicated to the Department of Foreign Affairs of the Republic
of South Africa acknowledgement of the latter's Note No.
8715: 11160/3/1, dated 28 January 1985 (Annex 45), under
which cover was sent a transcript in summary form of the
proceedings of the meeting held between the representatives of

the Botswana Government and of those of the South African
Governent on 19 December 1984 in Pretoria to discuss the
boundary in the Chobe area. The Department of External
Affairs of Botswana advised that the transcript was basically
correct but that because of its brevity it left out certain essential
details such as a reference to the maps and aenal photographs
brought to the meeting by the Botswana delegation which were

used in the discussion.

180. The experts finally produced theirJoin teport
on 15 July 1985 (Annex 48). They confirmed that the main
channel passes on the northem part of Kasikili/Sedudu and
therefore the Island lies on the Botswanaside of the border.

(c) The IntergovernmentalAgreement of 1984

181. The Vienna Convention on the Law of Treaties,
Article 31(3)(a), provides as follows under the rubric 'General
rule of interpretation-: "There shall be taken into account, together with the
context :

(a) any subsequent agreement between the parties
regarding the interpretation of the treaty or the
applicationof its provisions;."

182. In the opinion of the Goverment of Botswana
the intergovernmental agreement concluded at Pretoria
constitutes 'an agreementbetweenthe parties regarding .. the
application'ofthe provisionsofthe Anglo-GermanAgreement.

The principles of general internationallaw do not require any
particular formality for the conclusion of an international
agreement. The only criterion is the intentionof the parties to
concludea bindingagreementandthis canbe inferredfrom the
circumstances.

183. The legal background to informal international

instruments is examinedby AnthonyAust in his article in the
InternationaalndComparativL eawQuarterly ,Vol. 35 (1986),
pp.787-812. Mr. Aust points out that the International Court
saw important legal implicationsin informal unilateralacts in
the NuclearTests Case (Australia v France). I.C.J. Reports,
1974, pp.267-9. Aust observes: 'The position may be that
much strongerwhen a declaration iscontainedin a bilateral or

multilateral instrument andthere is, in effect, anexchange of
declarations'.(op.cit.supra, p.809). Certainly, the practice of
Statesincludes many examples of intergovemmental binding
decisions. A recent example was the decision concerning
Denmark in relation to the Maastricht Treaty taken in
Edinburgh in 1992 (view of the British Government, H.C.
Debates, Vol. 216, WA, col. 822; 15January 1993).

184. The evidence of the legal character of the
Pretoria Agreementderives from a whole series of documents
and inferencesto be drawn from the circumstances. 185. (i) Whilst there is no single instrument
embodyingthe intergovemmental agreementit clearly resulted
from a series of dealings including the meeting in Pretoria on

19 December 1984. The Minutes of the meetingare available
in two versions, one prepared in Botswana, the other in
Pretoria. The Minutes do not constitute an agreement as such
butprovide evidencethat an intergovemmentalagreement was
concluded. Both versions of the Minutesprovide unequivocal
evidence of an agreement to carry out a joint survey in order

to identify the main channel of the River Chobe.

186. (ii) The meeting took place at the Department of
Foreign Affairs and involved twohigh level delegations. The
delegations both included a senior legal component, a fact of
great significance for present purposes. The Botswana
delegationincluded the Attomey-General of the Republic.

187. (iii) As the Minutes reveal, the focus of the
meetingin Pretoria wasthe questionof the main channelof the
Chobe. Mr. Mare, in his opening address, stated that the
question of the course of the main channel of the Linyanti
River 'hadnow to be discussed'. InresponseMr. Mokama for

Botswana addressedthe same issue, that is, the identification
of the main channel in the vicinity ofKasikili/Sedudu.

188. (iv) The text of the Report of the Joint Survey
containsthefollowingpassageunderthe heading'Authorityfor
Survey ':

"At an intergovemmental meeting held in Pretoria on 19
December 1984 it was decided that a joint survey should
be undertaken to determine whether the main channel of
the Chobe River is located to the north or the south of
SeduduIKasikiliIsland.

Representatives of the two national survey organizations
accompanied by co-workers from the Departments of Water Affairs have now been to the areato survey the

'Thalweg'in the vicinityof the island. Specificmentionis
madeto the Thalwegin 1890 Agreement betweenEngland
and Germany" .

189. This passage affms the existence of an
intergovermnentalmeetingat which 'itwasdecided'thatajoint
survey shouldbe undertaken witha purpose directly linked to

the application of the provisions of the Anglo-German
Agreementof 1890.

190. (v) In the aftermath of the shooting incident
which gave rise to the Pretoria Meeting, the Govermnent of
Botswana immediately adopteda correct legal perspective.The

issue involved a boundary question and accordingly it was
necessaryto consultboththe UnitedNations Commissionerfor
Namibia and SWAPO 'before we can approach the South
African authorities' (see Telex dated 8 November 1984,
Annex 40).

191. Theresultwasa highlevel meetingin NewYork
betweenaBotswanadelegationanda SWAPOdelegationwhich
includedMr. Gurirab, the Permanent Observerto the United
Nations. The Minutes of this meeting, which involved the
Attorney-Generalof Botswana, revealthe legalcharacterof the
problems being discussed.

192. Duringthe sarneperiodthe Botswanadelegation
also consultedthe Presidentof the United NationsCouncilfor
Namibia and Mr. Mishra, the United Nations Commissioner
for Namibia. The Minutesof these two meetingsare presented
in Annexes42 and 43.

193. These various considerations leaveno room for
doubt that the result of the Pretoria Meeting was a binding
intergovemmental agreement relating toa legal subjectmatter
and concluded with senior lawyers in attendance. The partiesclearly intended to achievean effective and therefore binding
outcometo their deliberations.In the resultthe agreementwas
implemented: the Joint Survey was ched out. Moreover,

consistently with the bindingcharacter ofthese transactions,
neither Botswana nor SouthAfrica sought to challenge the
validity of theoint SurveyReport.

(d) The Joint SurveyReport of 5 July 1985

194. TheJoint Survey Reportwas finalisedon 5 July
1985. At no time subsequently has the South African
Govermnent challenged the validity of the Repon. The
exchange of telexes between Botswanaand South Africa in
1986(seeAnnexes 50-54)is of considerableimportance in this
respect. The Governmentof Botswana affirms the validity and
finality of the Joint Survey. The South African Govemment

proposes that the 'border issue'should be discussedbetween
Botswana and Namibia. However, beyond this procedural
proposa1 no attempt is made to impugn the validity of the
intergovemmental transactionsof 1984-85.

195. It must be emphasised, fmally, that the Joint

Suwey Report is itselfan intergovemental instrument.It was
signed by Fitschen 'for Chief Director of Surveys and
Mapping, Republic of South Africa', and by Raffle 'for
Director of Surveys and Lands, Republic of Botswana'. The
two survey teamswere carrying out the duties delegated to
them by the Pretoria Agreement of 19 December 1984. The

two national elements in the survey tearns were headed by
surveyors holding official rank equal to that of Deputy
Permanent Secretary. As the Note from Botswana dated 4
November 1985 (Annex50) makes clear, the Govemment of
Botswanaregardedthe JointSuwey Reportas bindingon South
Africa.

196. In conclusionit may be recalled that theJoint
Survey Reporthasa legalsignificancequiteapartfrom itsstatusas an intergovemmentalinstrument resulting fromthe Pretoria
Agreement. It constitutes expert opinion evidence on thekey

questionof law and fact, that is, the identificationof the main
channelof the Chobe in the vicinityof KasikiliISeduduIsland.

(e) The Government of Botswana regarded the Joint
Survey Reportas a confirmationof title

197. In 1986,duringdiscussionsofcommonproblems
with South African officiais in Gaborone (for the Record of
Discussions see Annex 51), the Govemment of Botswana
expressed the unquamed position that the issue of the
boundary in the vicinity of Kasikili/Sedudu Island had been
settled as a consequenceof the Joint SurveyReport. As Mr.

Mpuchane, Secretaryof Extemal Affairs, Botswana, observed
in the discussion:

"....there was no more room for negotiations because a
joint Botswana-South Africateamofexpertshadconfiied
that the Island belonged to Botswana. DSEA added that
Botswana's ownership of the Island would never have

arisen if it had not been for the shootingincidentbetween
the mies of the countnes around the Island the previous
year."

198. There followedan exchangeof Notes (Annexes
52 to 54). In its second Note of 25 November 1986 the

Govemmentof Botswana reaffied its position:

"Pula hasthe honourto acknowledgeSecextem'stelex no.
5142 file 1/160/3/1 dated 17 November, 1986concerning
SiduduIsland and to statethat the latter'spositionremains
as communicatedin its telex no. 164EA. 614B1 dated 22

October, 1986. "The joint BotswanaISouth Africa team of experts were
never asked to demarcate an international boundary but to

determine whether the main channel of the Chobe River is
located to the North or South of Sidudu Island." The joint
tearn confmed what had always been the fact, narnelythat
the main channel is located to the north of the island, and
that is where the boundary is.

It is therefore clear that adequate clarification of the matter

has been made to satisfy normal requirements and no
fùrther discussion of the matter is necessary."
(Annex 54)

199. In the result the Government of South Afnca did
not seek to deny the legal validity of either the original
intergovernrnental agreement of 1984 or the Joint Suwey

Report. CHAPTERVI1

The Issue of Fact:
TheMain Channelis to the North and West

of KasikiliISeduduIsland

(a) The Legal Context

200. Theprovisionsof the Anglo-German Agreement
refer to "the centre of the main channel" of the Chobe. The
Court'stask is to identifythe main channelof the Chobe River

in thevicinityof KasikilijSeduduIsland in accordancewith the
1890 Agreement. This is essentially a questionof f.cThere
are other exarnplesof treaty provisions referring to factual or
geographical criteria suchas a crest line, or a watershedline or
an escarpment line: see the Judgment in the Temple Case
(Merits), I.C.J. 1962, p.6 at p.15. In such cases the factual
criterion is adopted as or converted into the legal criterion.

But itdoes not cease to be in essence a question of fact.

(b) Criteriaby whichto identifythe main channelof the
Chobe River

(i) Sole channel

201. Article IIIof the Anglo-German Agreementof
1890 draws the line of the British and German spheres of
interest by reference to parallels of longitude and latitude "till
it reachesthe River Chobe, anddescendsthe centreof the main
channel("thalweg" in German)of that riverto itsjunction with
the Zambesi where it terminates".

202. Where the river flows in only one channel, that
channel will clearly be the "main" channel for the purposes of
ArticleIII of the 1890Agreement. As set outin ChapterII@),
the flows dong the Chobe (usually known as the Kwando,
Linyanti, Chobe)systemVarywiththe run-off in the catchmentarea. During drought years and the low flow regime of the
Zambezithe southem channel dries out. The variation of the

water levels determinesin effect whether an island exists or
not. In very dry seasonsthe southem section of the channel,
in particular at cross-sections15and 20A as more particularly
described in paragraph 218 below, dries out. On these
occasions there can be no question that the northern and
western channel, as the sole course through which the Chobe
River passes, is themain channel.

(ii) More than one channel

203. However, wherethereis more than one channel,
the assessmentof the facts necessdy requires a reference to
criteria by which to determinethe main channel; it will then
not be exclusively a question of fact because the criteria by
which the facts are assessedshouldlogically reflect the object

and purpose of the relevant provisionsof the Anglo-German
Agreement as assessed in the context of the principles of
general intemational law .

(iii) Navigability

204. Theexaminationoftheseprinciplesinthe period
at whichthe treaty was concluded(see Chapter V, paragraphs
129 to 144) indicatesvery clearly that navigabilityand access
to navigable water were primary considerationsin the mindsof
the negotiators.

(iv) Thalweg

205. Thisterm isusedto mean "theline at thewater's
surface verticallyabovethe deepestchannelof the river bed at
low tide" ."The thalweg is the channel most favourable to the
movement of vessels proceeding downstreamwhen the
water is at its lowest. It isubjectto natural alterations,
whereby the boundary suffers corresponding

displacement."
(Hatschek, Outline of International Law, Trans. by
Manning, 1930, p.130.)

206. On the facts, whenthe ChobeRiver is at its
lowest in the month of August only the northern and
western channel is navigable. At other seasons, as the

soundingsof the 1985JointSurveyshow, the northernand
western channel providesan averagedepthof 5.7m. which
is the more favourableto vesselsthan the 2.23m. average
depth of the south channel.

(v) Deepest channel

207. In the context of navigability the primary,
and perhaps the only, criterion would be relative depth,
although width of the channel would also be relevant to
ease of navigation. In the survey undertaken in 1912,
District CommissionerPanzerain a letterto CaptainEason
of 25 July 1911 set out his terms of reference for
surveying the Chobe River:-

"It is desired to obtain a reliable report with sketchesof
the various channels of the Chobeor Linyanti River, with
a view to determine which can be legitimately determined
as the "mainchannel" ... The widthof the channelwould
have little to do with it, and the question under
consideration could only be solvedby following up the

deepest channel in which there is the strongest current?
The most reliable map shouldbe usedas a guide and upon
this should be baseda prismatic compass rough surveyof
the various streams .."
(Annex 17) 208. Following these instructions and writing on 5
August 1912 Captain Eason reported:

"Two miles above the rapids lies Kassikhi
(Ka~ikiliISedudu)~ Island. Here 1 consider that
undoubtedly the north should be claimed as the main
channel. At the western end of the island the north
channelat thisperiod of the year is over one hundredfeet

wide and 8 feet deep, the southchannel is in a backwater,
what current there isgoes round the North."
(Annex 15, pp.125-63).

Thusdepth, strengthof current, and widthof the river were all
criteria applied by Captain Eason andall clearly indicatedthe
northem and western channelas the main channel.

209. The TrollopeandRedmanReportof 1948makes
no reference to criteria as such. (Annex22). However, their
report refers to "the examination of an aerial photograph",
whichwould indicatethe relevanceof width. TheJoint Survey
of 1985 refers exclusivelyto depth (Annex48).

210. A hydrological definitionof the main channelis
given by Dr. Sefe in his Opinion, amexed to this Chapter at
Appendix 1, which is set out below at paragraph 216 and
combinesin scientificterms the four criteria of depth, strength
of current, compositionof the load and width of the charnel.

211. There are occasionalreferences in legal sources
to criteria for the determination of the main channel in the
context of complex Stream and drainage systems. Such
referencesappearto involvequalitativelyhydrological contexts

and substantiallydifferent legal problems: see thergentine-
ChileFrontierCase. Report of the Court of Arbitration, 24
November 1966; UNReportsofInternational ArbitraA l wards,vol.16, p. 109 at pp.177-9.

(c) The Evidence

212. The facts on the ground and the scientific
evidencerelating to the hydro-morphology of the Chobe River,
as evaluated by expert opinion establish the following
propositions:-

1. The northern and western channel around
KasikiliISedudu Island is the present main channel;

II. All available evidence suggests that the northern and
western channel has been the main channel since the
present profile of the Chobe as a mature low energy
river system was established about 10,000 to 26,000
years ago. It, thus, has been a mature system for a
very long time;

m. Direct observation, historical evidence, aerial
photography and sedimentary and dating techniques
establish that the main channel at the time of the Anglo-
German Agreement in 1890was the same, namely the
northern and western channel as it is today.

(i) The hydro-morphologicalevidence relating to the
main channel of the Chobearound KasikililSedudu
Island

213. Dr.F. T.K.Sefe, BA., Hons., PhD, Lecturer in
Hydrology, University of Botswana, describes the hydro-
morphological basis for the identification of the northern and

western channel as the main channel of the Chobe in the
vicinity of Kasikili/Sedudu Island. His Opinion is setout at the
end of this Chapter at Appendix 1. 214. Dr. Sefe's reportconfms thethreepropositions
set out inparagraph 212 above. Each of these propositions is
well supported.

Proposition 1. The northem and western channel around
KasilcililSeduduIsland is the present main channel

215. In his report Dr. Sefe refers to the 1985 Joint
Survey in which 27 cross-sections around the island were
taken. He States:-

"The results clearly show the north channel as the main
channel: itsmean depth of 5.7m. exceedsthe mean depth
of the southchannelby 2.13m. me conclusionfromthis
suwey is thutthe northchanne1is the mainchunnelof the
ChobeRiver. " (Appendix 1, pp.3-4)

216. Dr. Sefe explainsthat the description "main" in
hydrologicaland geomorphological termsrelatesto the energy
of the river and isdefmed in functionalterms by reference to
the "competence" of the river to transport debris and its
capacity measured by the maximum load (sediment of a
particular grain size)t can carry.

"As energy in a river is proportional to the product of
mass (i.e. size) and the bedslope, "main" is synonymous
with size. Thus of two tributaries in the sarneriver, the
larger is consideredthe main channel" (Appendix 1,p.5).

217. The Joint Survey of 1985 (the FitschenIRaffle

report) took depth soundings at 27 cross-sections around
KasilcililSeduduIslandin a clockwise directionbeginningat the
Chobe National Park HQ. Annex 48, Plan D.

218. From these soundings it is possible to draw a
profile of the river bed in the north-western channel(Profile 1)
andthe southchannel(ProfileII) aroundKasikiliISeduluislandand these are shownin two diagramsat Annex 62. Diagram 1
showsthe LongitudinalProfiles around KasikiliISeduduIsland,
extracted from the soundings recorded by the 1985 Joint

Survey, beginning at the most easterly point of each charnel
and ending at the most westerly point. Cross-section 1 from
which bothProfiles are drawn is used as a common datum or
bench mark as the startingpoint. The profile of the bed of the
Chobe River, as there indicated, can, thus, be described, as
follows:-

North-Western Channel.Beginningwith cross-section 1, and
proceedingin a clockwisedirection, after clearingthe sandbar
downstream fromthe small island, the north-western channel
deepens from 3 m. to 6 m. As the river rounds the island the
depth reduces to 3.5 m. but again deepens to 6-7 m. at the
juncture of the back-streamat Kabuta. From cross-section5 at

the most northem point of the island, to cross-section9, being
the point where upstrearn from Kasane the south-eastem
channel joins and merges into the main river, the charnel
exceeds 9 m. in depth; it isneverlessthan 6 m., the shallowest
part being at the exit of a small back-strearn from
KasikiliISeduduIsland itself.

South Channel.Proceeding in an anti-clockwise direction the
entry point to thesouth sectionof this channelis very shallow,
1.5 to 2.m. in depth, and obstructed by reeds. Apart from a
depth of 5.5 m. foundto the northof ChobeNational ParkHQ
at cross-section 16, the channel remains shallow until cross-
sections 14 and 13 where depthsof 3.5 m. are shown. Reeds

again at this point obstmct the channel. Thereafter, in the
eastem section of the south channel to cross-section 10 at a
depth of 6 m., which is located at a point shortly before the
south channel joins the main river, the channel deepens but
never exceeds 5 m. 219. Diagram 2 in Annex 62 provides a Comparison
of the Longitudinal Profiles of the north-western and south

channels, based on these longitudinal cross-sections. In
summary, it is apparent that as regards the south chunnel,
whilst there is a depth of some 4 m. average of water in the
eastem section,boththeeastand Westentrancesto the southern
sectionof this channelare barely morethan 3 m. in depthwith
thick growth of reeds extendingfrom each side, with a deeper
"lagoon" area of 6 m. in the centre of this southem section.

Navigation is thus severelyimpeded, if not impossible, at the
entrancesto the southern sectionof thesouth channel.

As regards the northern and western channel, once the river
has cleared thesmallsandbar to the north of the National Park
Headquarters,the cross-sectionsshowa steady depthof 5 to 7
m. with a noticeabledeepening to 10m. after the entry of the

backwater strearnat Kabuta, andto 13m. at thejuncture of the
eastem section intothe main stream. A navigable channelof 5
to 7 m. therefore exists in the northem and western main
channel throughout.

220. In summary, it is thus apparent that as regards
thesouthchannel, whilstthere is a depth of some4m. average

of waterin the easternsection,boththe eastand westentrances
to the southem section of this channel are barely more than
2.5m. in depth with a thick growth of reeds extending from
each side, with a deeper "lagoon" area of 6m. in the centre of
this southern section.Navigation isthus severelyimpeded, if
not impossible, at the entrances to the southern sectionof the

southchannel.

221. As regards the northern and western channel,
once theriver has clearedthe smallsandbar to the north ofthe
National Park HQ., the cross-sectionsshow a steady depthof
5 to 7m. witha noticeable deepeningto 10m.after the entry of
the backwater strearnat Kabuta, and to 13m. atthejuncture of

the eastem sectioninto the main strearn. A navigablechannelof 5 to 7m. therefore exists in the northem and westem main
channel throughout .

(ii) Evidencesupportingasa matterof fact the northern
and westernchannel as the main channel

222. AU the observers accept, that, as a matter of
fact, the northem and western channelis the "main" channelin
the vicinity of KasikiliISedudu. On 14October1948in a letter

addressed on behalf of the South African Governmentto the
High Commissioner's Office in Pretoria, the South African
Govemment clearlyacceptedthat, as a question offact, it was
the northem and westem channelwhich wasthe main channel.
Referringto theJoint Reportof TrollopeandRedmanthe letter
continued:-

"The Report disclosesthat while the main channel of the
Chobe River is shown on maps as passing to the south of
Kasikili Island it in fact passes to the north of that island.

It has been cofimed, as a result of exhaustiveenquiries,
that there has been no shifting of the main channel of the

river from South to North within living memory. The
facts, therefore point to the maps beingincorrect."

(See further, Chapter V, para 154).

223. TheJoint ReportofTrollopeand Redmansigned

on 19 January 1948 accepts that the main channel liesto the
north and West (para.4) (Annex 22). Moreover, Trollope's
persona1opinion wasto the sameeffect. His coveringletterto
theJoint Report, dated21 January 1948,containstwo passages
in which he clearly acceptsthat it is the northem channelwhich
is the boundary in accordance with the Anglo-German

Agreement (See Annex 22, paras.3 and Il). Indeed, it
expressly refers to the northern and western channelas "the
geographicaland treaty" position (ibid.,para.11). 224. That this was the settled view of the British
authorities isevidenced bya sketch mapof the lower Chobeof

Scale 1:125,000or 2 miles = 1 inch, dated 18 October 1965,
tracedfrom the Dept.of Surveyor-Generalfrom Print Laydown
Sheet 1725C,Mafeking, Oct.,1965,attachedto the opinionof
Mr.R.R.Renew , the Surveyor General of the Bechuanaland
Protectorate. The northern and western channel around the
island is coloured blue and a superscription explains "main

channelcoloured blue" . See Chapter V, para 161, Annex 36,
Map 18.

225. The evidence provided by aerial photographs
covering the period 1925 to 1985 is particularly impressive.
The photographs confirm the significanceof the northern and

westernchanneland show that no substantialchangehas taken
place in the period of sixty years coveredby the photographs.
The evidence is presentedbelow at paragraphs229 to 232 and
in Appendix 1 to the Memorial.

226. Of particular importance is the Report of the

Joint Survey conducted in 1985 (Annex48). The principal
conclusionof the Joint Surveywas that:

"The main channel of the Chobe River now passes
SeduduIKasikiliIsland to the west and north of it."

227.
The legal significanceof the Joint Survey is
exarninedin greater depth in the Memorial, Chapter VI. For
the present purpose it is adopted as expert evidence, the
professional reliability of which has notbeen chailenged by
either Botswanaor Namibia.Proposition II. Al1 available evidence suggests that the
northem and western channelhas been the main channel since
the profile of the Chobe as a maturelow energy river system

was established,and thus has beena mature systemfor a very
long the.

228. Dr.Sefe records the following features of the
morphologyof the Chobe River:-

the geomorphologyof the Choberiver systemhas been
(i)
influencedby the hydrology of two major prehistoric
lakes, Makgadikgadi ,which extendedinto the alluvial
flatssurrounding the Chobe and Zambezi rivers, and
theMaklane, reachingthe Chobeviathe Savutichannel
immediatelyto the west of the Magikwe Ridge. The
topographieelevationsof theselakes, 910to 940m. and
963m. respectively weresignificant in regulating the

overflow of water in the Chobe and Zambezi rivers,
and together with tectonism causeddown-warpingand
deformationat the Chobe-Zambeziconfluence. Carbon
14dates from shells atthe 940-945m. level suggestan
age for this activity of c.40,000 to 35,000
B.P.(Appendix 1 of this Chapter, pp.5-6);

(ii) the eventual drying up of these lakes occurring over
10,000 years ago caused a significant change in the
hydrologyof the area (p.3);

the broad north easterlydirectionof the ChobeRiver is
(iii)
determinedby tectonic actionoccumng some26,000to
40,000 years ago, channelling the river into a
"rectilinear fault-controlled course" north eastward to
join the ZambeziRiver (p.2).

There has been no tectonicactivityin recent centuries,

as demonstratedby the current lowenergy statusof the
river (p.7); the south channel, as in the case of other backwater
(iv)
channels, shows the same lineations as the regular
channel systems. It is, therefore likely that at the time
of the establishment of the present Zambezi course,
about 26,000 to 40,000 years ago @.2), the south
channel was the main channel. It was later abandoned
as a resultof erosionof the sandridges whichdelivered
an excessivehigh sedimentload, and also as a result of

dow nthrows dong the Zambezi's axis (about 10,000 -
26,000 years ago) which changed slopeconfiguration.

229. Dr.Sefe's report continues:-

the presentmorphologyof the Chobe could not
(v)
have been alteredfromthe commencementabout
2000 years ago of the current stable dry phase
of the climate, leading to drastic reductions in
flowvolumes (Appendix 1of this Chapter,p.7);

the present featuresof the Chobe -
(vi)

the meander loops, the numerous backwater
lagoons,associatedwiththe outsidebendsof the
meander loops, SeduduIKasikili and Xakuma
islands and smaller sand bars elsewherein the
river, the rapids in the vicinity of

Commissioner'sKop, -indicate a river with a
lowbedslope,with an imbalancebetweenenergy
and sediment load and which has also been
subjectto tectonism (ibid., p.2.);

(vii) the landscape suggestsa river system whichhas
shifted many times across the floodplain,

continuallyaggradingand degradingits channels
in maintainingthe local base level, narnely the
top of the Victoria Falls (ibid.p.2); (viii) the south channel shows that it was stopped
from shifting further southwards by the
sandridges which generally rise steeply from
930m. to 1000m., there being no evidence of

recent erosion by the Chobe River along these
ridges (ibid.,p.2);

(ix) the stratigraphy of Kasikili/Sedudu Island
suggests that it wasinitially a sand bar deposit
that later became stabilised by vegetation; its

soi1depth was thickenedwith layers of clay and
silt deposited as successivefloodwaters swept
overandwassubsequentlytrappedby colonising
vegetationcover (ibid. p.2);

(x) erosion and sedimentation in rivers are
necessary for rivers to adjust to base levelbut

the processes are slow. Under ideal conditions
it would take thousandsof years to build a silt
and clay deposit in excess of 2m. Fieldwork
around Kasikili/SeduduIsland indicatesthat silt
and clay depositsexceedabout lOm.(ibid., p.8).

PropositionIII. Direct observation,historicalevidence, aerial
photography and sedimentary and dating techniques establish
that the main channel in 1890was the same, namely the
northem and western channel as it is today

(iii) Direct observation

230. As set out in paragraph 228, (ix) and (x),
fieldwork directly observing the present stratigraphy of
Kasikili/Sedudu Island indicates that its formation has taken
hundredsof years and that in the last hundredyears itspresent
shape remainsunchanged.(iv) Historical evidence

231. The historicaldata (set out at paragraphs 222 to
227 above), clearly supports the Botswana position that the
main channel has at al1relevant times been the northern and
westernchannelof the Chobe. In the years prior to and shortly
after 1890 a number of maps and sketches were preparedby
explorers of southern central Africa, including David
Livingstone, Emil Holub, F.C.Selous, Benjamin F-Bradshaw

and A.Schulz and A.Hammer. Most are too sketchy, or smail
scale, to be of value.

232. Of these sketch maps that of Bradshaw's isthe
most relevant; it relates toa sketch rnap of the Chobe River
drawn in 1880 by Dr. Benjamin F. Bradshaw to a scale of

l.lcm to 1 mile. It can be found in Volume IïI of the
Proceedingsof the Royal Geographical Society for the year
1881, at p.256 accompanyingBradshaw "Notes on the Chobe
River, South Central Africa"and is Map 1 (AppendixII of the
Memonal).

233. This is a large scale sketch rnap whichclearly
shows Kasikili/SeduduIsland in a formation similar to that as
shownby CaptainEasonin 1912. Bradshaw'ssketchwasused
by Seinerin his rnapof 1909in which he specificallyrefers to
"Bradshaw080". Bradshaw showsthe northem and western
channel clearly broaderthan the eastem, and the entry to the

southern channel is at a sharp angle suggesting that the
northwest is the main flowof the river.

234. The available maps and sketches indicate that
fromthe timethe Chobewassurveyedwith anyparticularityby
European explorers from the 1860s onwards a north channel

around KasikiliISedudu Island was known and regularly
depicted. Dr.BenjaminF. Bradshaw'srnap of 1880 (Map 1),
the Frankenberg rnap of 1912 (Map 7), and Captain Eason's
rnap of 1912 (Annex 15), clearly indicatethe presence of thenorthern and westernchannel closely similar to its present
configuration.

235. More recent historicalevidenceis to be foundin
a map dated 25 May 1953 showing agricultural and grazing
areas of North Bechuanaland showingKasikili/SeduduIsland
as a peninsula with no southem channel. This rnap entitled

"1953Approximatedelineationof concessionAreas, 1 /HT/ 1",
was drawn by Silby Horrell, chartered surveyor, Salisbury,
S.Rhodesia, to a scale of 1 125,000 (Map 15). A note is
attachedto the map entitled "Noteof agricultural and grazing

areas to be read in conjunction with Map l/HT/l" and the
legend of the map identifies tirnber and cattle concessions,
grazingareasandthe locationof waterholesandpits, andriver
proximity. No grazing area is shown onthe Kasikili/Sedudu
peninsula.

(v) Aerial photographs

236. Aerial photographsare available dating back to
1925 and the Joint SuweyReportof 1985noted that:

"Airphotographsshowingthe channelsof the river in the
vicinity of the island are available in the archives of the
two national survey organisations. They were taken in

1925, 1943, 1962, 1972, 1977, 1981, and 1982. No
substantialchangeinthe positionof the channelsis evident
from the photographs" .

Al1 these photographs resulted from aerial surveys

commissionedon a professional basis, andwith the exception
of thephotographapparently takenin 1977,appearatAppendix

6~ similar configuration is shown on a map prepared by the
Topographical section of the Union Defence forces, South Africa,
Sheets7 and8 CapriviZipfel, No.115, July 1915;Scale = 10.lkm or
6.3miles, see Map 12.1 in the Memorial.The photograph relating to 1977 has not
been traced, and the referencemay be due to an error.

237. A remarkablefeatureapparentfrom the studyof
the photographs is the consistency of the shape of
KasikiliISeduduIsland over a period of nearly six decades.
The relevant photographs available in the archives of the
Departmentof Surveysand Lands in Gaboroneare as fo1lows:-

0) June 1925 Scale 1 16,000approx. (see Appendix 1)
(ii) 1943 Scale 1:10,000, (see Appendix 1)
(iii) 15 August 1947Scale 1:10,000 (see Appendix 1)
(iv) 1962 Scale 1:10,000 approx. (see Appendix 1)
(VI May 1972 Scale 1 :10,000 approx. (dry season) (see
Appendix 1)
November 1972Scale 1 :10,000approx. (seeAppendix

11
(vii) 1981 Scale 1:16,000 approx. (see Appendix 1)
(viii) 3 August 1985Scale 1 :10,000 (see Appendix 1).

238. These photographs callfor careful study. Dr.

Sefe in his Opinionlistssevenof theseaerialphotographstaken
at intervals between the years 1925 and 1985. He conducted
an analysis of these photographswhich shows that the shape,
locationand size of Kasikili/SeduduIsland has not changedin
the 60 year periodbetween 1925and 1985. In fact he reports
"features are so stable that a small cluster of trees on the
water's edgeon the northern spit appeared on all the aerial

photographsand was easily identified"by Dr. Sefe during the
fieldwork of 10and 11March 1994. There was nodiscernible
changein topographical levels. Thesetrees are stillto be seen
there today .

239. The following featuresremain consistent in the
aerial photographs: a) the overall shape of the island; (b) the

alignment of the channels in relation to the island; (c) thecomparativedimensionsof the northemlwestem channelof the
Chobe and the eastemlsouthem channel.

(vi) Sedimentaryevidence

240. Fieldwork around KasikiliISedudu Island
indicates that silt and clay depositsxceed about 10m. The
depth of these depositsprovidesevidenceto date the ageof the
original sandbar depositwhich forms the island. Itsoi1depth

was thickened with layers of clay and silt deposited as
successive flood waters swept over and were subsequently
trapped by colonisingvegetationcover. Under ideal conditions
it would take thousandsof yearsto build a silt and clay deposit
in excess of 2m.

Absenceof evidenceof change
(d)

(i) Absence of historicalevidenceof change

241. Given the strength of evidence that the main
channel is the northem and westem channel, any opposing
thesis mustbe based uponthe hypothesisof a radical changein

the course of the Chobe between the present day and 1890.
Such athesis meets a number of substantialdiffculties.

242. There is a complete absenceof any evidenceof
radical change in the course of the Chobe. Large scale maps
both before and after the 1890 Anglo-Geman Agreement,
prepared by those who had actually surveyedthis stretchof the

ChobeRiver, showan unchangedcourse. Thusthe sketchmap
of Dr.BenjarninF.Bradshawof 1880(Map l), the Frankenberg
Map of 1912 (Map 7), and Captain Eason's sketch map of
1912accompanying his survey (Annex 15), al1showthe sarne
configuration of the island (with a broader northem and
westem channel) as that shown in the Joint Survey of 1985

(Annex 48). 243. In this connection the comments of Captain
Eason made in his Report of 5 August 1912are of relevance.
CaptainEason suppliedhisown sketchmapof the Chobeinthe

vicinityofKasikWSeduduIsland, at a scaieof 1:50,000 (Sheet
2, Annex 15), and attached a tracing ofa map by Streitwolf,
who was the German Residentin the Caprivi Zipfelprior to
Von Frankenberg. In Eason'sown words:

"Most maps do not show the river as going to that lake
(Liambesi)notably Seiner's Map published by E.S.Mittel

and Sohn, Berlin. Thisis the map at present used by the
German authoritiesin the Caprivi Strip. They do not in
conversation attemptto defenditsaccuracy(or inaccuracy)
with regard to the eastem section of the Linyanti and 1
attach a tracing of a map madeby HauptmannStreitwolf,
the First Imperia1 resident in the Caprivi Strip of the
Linyanti River from Kazungula to the Liambesi which

though inaccuratein detail is, in essentiaipoints sirnilarto
mine. 1 do not believe that either us or the subsequent
residents have attempted to map the river West of
Liambesi" .
(Annex 15).
Earlier Captain Eason reported:-

"It seemsto me that most maps have been constructedon
the principle of mapping the edges of the swamps and
drawing a line down the middle to show where the river
may possibly run". (ibid.)

244. The Chobeis subjectto floodsboth in winterby

reason of the rains and in summer by the Zambezi floods.
These floods, though they vary in volume from year to year,
have beena regular feature of the morphologyof the river for
centuries, probably from about 2000 years ago when the
present climatic conditions began. As a result of the floods
caused by the JanuaryIMarch rains allthe land lying between
the upper Zambeziand the lower stretchesof the Chobeup tothe first rapids and beyondforms one vast lake with only here
and there a tree or small island appearing above the water

level.

245. Flood hydrographs of the annual discharges
1924-1958of the ZambeziRiver at Kariba indicatethe annual
variationinfloodwaters;withhighdischargesestimatedforthe
years 192516, 193112,193819, 1943/44; and recorded high
flows in the years 194718, 195112,195213and 195516,with

exceptionalhigh flow in theyear 195718.SeetheHydrological
Yearbook 1957-1958,Southern Rhodesia,HydrologicalBranch
Division of Irrigation, 1958-9 (Annex 60). These flows are
codïmed by flood hydrographs 1937-1947 of the Zambezi
River at Livingstone Pump House provided by the Lusaka
Officeof the SADC, set out at Annex 61.

246. As isapparent fromthe aerialphotographstaken
during the period over which these flood hydrographs were
compileci,despitethehighandexceptionalhighflowsrecorded,
nonewas sufficient to changethe configurationof the islandor
the course of the main channel.

247. In his Opinion(Appendix1) Dr.Sefe States:

"Hydrologically for a single year's floodsto cause the
level of siltingequired to block a river of the size under
considerationwould imply an extremely activeriver, able
to erode, transport, and deposit tons of sediment. "

(Opinion, p.5).

There is no observable or scientific evidence to supportthe
occurrenceof floodsof such unusualdimensions from 1890 to
the present day.(ii) The Presumptionof Continuity

248. The evidenceofthe stateofaffairsrelatingto the
statusof the northern and western channelas the main channel
which obtainedin the period 1890to 1985is reinforcedby the
presumptionofcontinuity recognisedby internationaltribunals.
This consistsof the common sense assumption thatthere is a
probability that certain conditions or relations continue: see

Halsbury'sLaws of England, 4th Ed., Vol.17, London 1976,
p.26, para.34.

249. In the practice of international tribunals the
presumptionhas been appliedin the form thatlater events were
evidenceof an earlier stateof fact.The p~ciple was applied
in this formby JudgeBasdevantin his distinguishedindividual

opinion in the Minquiers and Ecrehos Case, I.C.J. Reports
1953, p.74 at pp.82-3. This aspect of the individualOpinion
received the approval of Sir Gerald Fitzmaurice in his well-
knowncommentariesonthejurisprudenceof the Court:British
YearBookof International Law, Vo1.30,(1953),pp.55-6; also

Sir Gerald Fitzmaurice, The Law and Procedure of the
International Court, Cambridge, 1986, Vol.I,pp.185-6.

(iii) Absenceof scientificevidenceof change

250. The Government has sought the advice of

scientificexpertsinexaminingthehypothesisofa change inthe
main channel since 1890, and in particular refers to the report
of Dr.Sefe at Appendix1of this Chapter. Dr.Sefe as an expert
in hydrologyand geomorphology giveshis opinion that:-

the present status of the various channelsof the Chobe
(i)
was established about10,000to 26,000 years ago (see
para. 227 above);

(ii) the present gradedprofile could not have ben altered
from the current stable dry phase of the climate about 2000 yearsago (seepara. 228 above);

(iii) there is no evidence of tectonic activity affecting the
Chobe in recent centuries;

(iv) the geomorphological characteristicsof the Chobe are
incompatiblewith the hypothesisof floods of unusual
magnitude;

the Chobe River has been a system in equilibrium for
(v)
a long time;

(vi) the northernand western channelis the mainchannelof
the Chobe River and has beenfor thousandsof years.

251. The results ofthis expert opinion evidence are
very cogent. In particular, the Opinion of Dr. Sefe

complements andreinforces theconclusionsof the Join tuwey
Reportof 1985. The conclusionmust bethat, if a change of
course since 1890was scientificallyimpossible,the conclusion
of the Joint Survey was directly relevant to establishing the
identityof the main channel in 1890.

(e) General Conclusion

252. The facts set out above relating to the
morphologyand hydrology of the Chobe River supported by
direct observation,historicalevidence, aerial photographyand
scientificmeasurementof sedimentarydeposits,all indicatethat

the northern and western channelof the Chobe River in the
vicinityofKasikili/SeduduIsland isthe mainchannel. It isthe
main channel today and hasbeen since 1890, whenthe Anglo-
German Agreementwas concluded. APPENDM TO
CHAPTERVI1

THEHYDROGEOMORPHOLOGICAL BASISFOR
BOTSWANA'SCLAIMON THE SIDUDUIKASIKILI
ISLAND

Opinionpreparedby

F.T.K.Sefe, Universityof Botswana,PrivateBag 0022,
Gaborone

March 1994This Opinion isseparatelypaginated.

1. INTRODUCTION

Siduduor SeduduIsland (or KasikiliIsland to Namibians)isa
small island, about 3.8km2(1.5 sq. milesin area, in the Chobe
River, located withinthe area boundedby, approximately, the
25'07' and 25'08'E longitude and the 17'47' and 17'50's
latitude. The island is envelopedwithin two unequal arms, in
terms of size, of the ChobeRiver.

One of the issues pertinent to the resolution of the border
disputebetween theRepublicof Botswana andthe Republicof
Namibia in the vicinity of the Sidudu/Kasikili is the
determination of which of these two channels is the main
channel, theNarnibianspreferring the south channel, thereby
making the Island Narnibian territory. This paper presents
evidence relative to thegnificanceof the channels, whichcan

help resolvethe border dispute.

2. THEPRESENTMORPHOLOGYOF THECHOBE
RIVER

For the purpose of the exercise, the description of the
morphology of the Chobe River will be restricted to that

portion of the river containeci on the Kasane 1:50,000
topographic sheet (SheetNo. 1725C3 and Part C4; published
by the Department of Surveys and Lands, Gaborone 1984 as
Edition 3). (See Fig. 1 for extract from this sheet).

'Present' as used in the context of this report follows the
normal geomorphological connotation in which 'present' is

defmed as "that tirne period over which input and output
conditions have remained reasonably constant on average"
(Knighton, 1984).The main present morphologicalfeatures of the sectionof the
Chobe River of interesthere are:

-
the meander loops which are generally wide although
some are narrow;
- the numerous backwater lagoons, associated with the
outside bendsof the meander loops;
- Sidudu/Kasikiliand Xakuma Islands and smaller sand

- bars elsewherein the river;
the rapids in the vicinity of Commissioner's Kop.

All these features are produced by rivers with an imbalance
between energy and sedimentload, and whichhave also been
subjectto tectonism.Suchrivers often shifttheir coursesacross
the flood plain as they continuously aggradeand degradetheir

channels in a bid to maintain a profile that is adjusted to at
least localbase level (inthis case, the top of Victoria Falls).

In the particular case of the Chobe River, it is one of three
parallel linkedrivers whichrise in the AngolanHighlands. On
reaching the southwest-northeast fault zonesof the Gwembe

Troughand Okavango Grabben, (featureswhichare thoughtto
represent an incipient arm of the East African Rift system
(Scholz, 1975),the Chobe followsa rectilinearfault-controlled
course northeastwards to join the Zarnbezi River at Kasane.
The landscapesuggestsa river system which hasshifted many
times across the floodplain; but the south channel appears to

havebeen stopped from shiftingfurther southwardsby the sand
ridges which,generally , rise steeplyfrom 930mto 1000m.Al1
dong theseridges, thereis no evidenceof recenterosion bythe
Chobe River. Furthermore, the low energy status of the river
at present which is reflectedby the existenceof the meanders,
sandbars and subsequent braiding (subdivisionof the chamel

intotwoor more branchesflowingin the samedirection)would
be the result not only of low bedslopes, but also drastic
reductionsin flow volumesas a result of past clirnaticchange.It is worth noting that Sidudu/KasikiliIsland was initially a
sandbar deposit,createdby methodsindicated above, thatlater

becamestabilisedby vegetation.Its soi1depthwasthickenedas
layersof clayand siltwere deposited as successiveflood waters
sweptover it. Much of the siltand clay became trappedby the
colonising vegetationcover. It is likely thatat thehe of the
establishmentof the present Zambezi course about 26,000 to
40,000 yearsago (Coates etal, 1979;ThomasandShaw, 1992)
the south channel was the main channel, its establishment

having been aidedby the downwarpingand downthrows which
accompanied theintense tectonic activitiesat that time. It was
later abandoned asa result of the erosion of the sand ridges
which deliveredan excessivelyhigh sedimentload, and also as
a result of the downthrows along the Zambezi axis (about
10,000-26,000yearsago)whichchanged slopeconfigurations.

Someof the backwater channels werelikelyprevious channels
(tributaries or distributaries) of the Chobe River as someof
them showthe samelineationsas the regular channelssystems.
Sincethe establishmentof the north channel,the southchannel
has become the largest of the backwater channels,although it
stillarries someflowtowardsthe ZambeziRiver.Manyof the

backwater channels are associated with abandoned meander
loops and lagoonsresulting from spit deposition. While it is
probable that further shiftswill occur in the geologicfuture, it
is certain that the present alignment of the Chobe River
represents a mature, low energy system that is graded in the
classic sense of the concept of grade as describedby Mackin

(Schumm, 1972). Considering the fact that the last major
climatic perturbation which led to the final desiccation of
Palaeo-Lake Makgadikgadi and Lake Ngami occurred over
10,000 years ago (Shaw, 1988), it is most probable that the
present gradedprofile of the Chobe River system was attained
at the sarnetirne.3. DECIPHERING CHANNEL CHANGES IN THE

CHOBERIVER SYSTEM

The 1985HydrographicSurvey

In 1985, following an intergovemmentalmeeting held
in Pretoria on 19 December 1984, a hydrographic

survey was undertaken by a joint team cornprising
representatives from the Department of Surveys and
Mappingand the Departmentof Water Affairs, both of
the Republic of South Africa, on one hand, and
representatives from the Departmentof Surveys and
Lands and the Department of Water Affairs of the
Republicof Botswana, on the other hand.

Inall27 cross sections weretaken as shown in Fig.2.
The resultsclearly indicate the north channelas the
main channel:itsmeandepthof 5.7m exceedsthemean
depth of the south channel by 2.13m. ï?zeconclusion
from this suwey is that the northchunnel is thmain

channelof the ChobeRiver.

As riverscarry clues to their history, the manner in
which the north channelevolved to become the main
channeltoday can be deciphered from themorphology
of the river. The methods usually employed for

decipheringthis historywerelistedbyKnighton(1984).
They are listed in Table 1. The available data are
summarisedin Table 2.Table 1: Method used in studying river channel
changes

Direct observations - Instmment records
Photographic records
Ground survey

Historical records - Maps and photographs of
different dates
Historical documents

Sedimentary evidence- Surface forms
Internal structures

Dating Techniques - (1) Relative methods
Relative height
Organic remains
Artifacts
- (2) Absolute methods
Radioactive isotopes
Dendrochronology

Thermoluminence (TL)

The word 'main' has both hydrological and

geomorphological connotations relating tothe velocity
of flow, hence the discharge. So it connotes theability
of the river to transportdebris in terms of particle size
(i.e. itscompetence[Mackin, 19481)'andthemaximum

load (sediment of particular grain size) a stream can
cany (i.e. its capacity as defineclby Gilbert [1914]).
Both the competence and capacity of a river are
functions of energy of the river. As energy in a river

reach is proportional to the product of the mass (Le.
size)andthe bedslope, 'main'is synonymouswith size.
Thus of two tributaries of the same river, the larger is

consideredthe main channel.3.2 The hypothesisof floods in 1899

The documents available contain references to the
possibilityof floodsoccurringin 1899, nine years after
the Anglo-German Treaty that fixed the boundary
between the Caprivi Strip and Bechuanaland
Protectoratedong the mainchannelofthe ChobeRiver.
It is claimed by the Govemmentof Namibia that until
these floods, the main channelof the Chobe Riverwas

the South Channel.

Significantly, there is noestimate of the magnitudeof
the alleged floods. However, for such a claim to be
sustained, the floods would have to be at such a
magnitude and spectacular in inflicting flood darnage
that it wouldpass into folldore-perhaps, a flood of the

magnitude rivalling the biblical floods during Noah's
time! The issue of recollectionof these floods will be
taken up again later.

Hydrologically, for a single year's floods to causethe
level of siltingrequired to block a river of the size
under consideration would imply an extremely active

river, able to erode, transport and deposit tons of
sediment. With a flood season lasting from March to
May, the amountof sedimentmovedin this way would
leave an indelibleimpressionon the morphologyof the
river. There are several reasons why such a scenario,
one season'sfloods siltingup a river of the size of the

Chobe and causing a shift in its alignment, is highly
implausible .These are:

(i) The gwmorphology of the Chobe River system
was influenced by the hydrology of the two
major palaeo lakes of Northern Botswana,
namely , Paiaw-Lake Makgadikgadi and Lake

Ngami (Shaw, 1988).Of greater relevancehere are the topographie elevations of these lakes -

940 to 945mand 936mrespectively. (Fig. 3)

Table 2: Data available for deciphering the
morphology of the Chobe River in the vicinity of

SiduduIKasikiliIsland

TABLE1 NA'IVREOF SOURCE
DATA
CATEGORY

Historical
records

Documents: Reporton survey Botswana
by Capt. H.V. National
Eason - 1912 Archive. BNA
Airphotographs S 91311
of different 1925 scaie-

dates: 1:16000 approx Photos
(G) included in
1943 scaie- folder obtained
unstated (F) from DWA.
1962scaie - Letters in
1:16000 approx bracketsare

(El codes
1972scaie - previously
1:16000 approx assigned.
@>
1972 scaie-
unstated (C)

1981scaie - 16000
approx (BI
1985 scaie-
1:17500(A)

-Direct
observation: Hydrographic In Report on
survey at 27 cross Chobe River
Ground survey: sections(see Fig. Boundary
Survey
2)
Sidudu/Kasikili
Island July
1985, Survey
undertaken by
-ajoint South
Africa and
Botswana
Tearn. Report
contained in

folder obtained
from DWA.

Field visit Undertaken by
author on 10th
and
11th March
1994

The 940-945m level of Paleao-Lake
Makgadikgadi extended into the alluvial flats
surroundingthe Chobeand Zambezirivers. The

fullextent of this lake would require inflow
from localand distant valleys.It is thought that
now fossil valleys like the Okwa and Groote
Laagte contributed, as did the Zambezi (Bond,
1963). Several authors have suggestedthat the

Zambezi transferred water into the early
Makgadikgadi system, with the water level
regulatedby overflowto the ZambeziValley at
VictoriaFalls. Subsequenttectonism thenaltered
the configuration of the system, by

downwarping dong the Zambezi axis, andtor
isolationof the Makgadikgadiby movement ofthe lower OkavangoFaults. Thomas and Shaw
(1988) noted that fossil alluvium is extensivein

the Caprivi Strip of Namibia, between the
Chobe and Zambezi rivers, reaching 1050mat
the Kafue watershed to the north of the
Zambezi. However no landforms at the 940-
945m level are encountered at the Chobe-
Zambeziconfluence,suggestingthatdeformation
has been active in this area. Carbon-14 dates

from shellsat the40-945mlevel suggestanage
of c.40,000-35,000 B.P. (Cooke and
Verstappen, 1984).The history of Palaeo-Lake
Makgadikgadihasbeen one of altemate dry and
wet periods which eventually culrninatedin the
last major dry perioduring which Palaeo-Lake
Makgadikgadidried up.

Later at the 936m level Palaeo-Lake
Thamdakane overflows reached the Chobe via
the Savuti channel irnmediatelyto the west of
the Magikwe Ridge. It is thought that Palaeo-
Lake Tharnalakane, via the Boteti River, was

instrumental in the maintenance of standing
water at the 920 and 912m levels in
Makgadikgadionnumerousoccasionsin the last
20,000 years. The 936m level lake was
controlled by the Mambova Falls where the
Chobe and Zambeziconverge.

The eventual dryingup of these lakes caused a
significantchangein the hydrologyof the entire
Okavango System, including the Chobe. The
present morphology of the Chobe could have
been established about 2,000 years ago
coincidingwiththe startof the current stabledry

phase of the climate.(ii) Tectonicactivity appears not to be in evidence,
at anytirne in recent centuries. If an earthquake

hadoccurred, the mostlikelyresponsewould be
fault-controlledhannelalignmentswhichwould
significantlydisturbthe present meanderingand
braiding obvious with the Chobe River. The
mechanismof braiding, and the accumulationof
sandbars has been exhaustively studiedin field
laboratory experimentsby the US Anny Corps
of EngineersandtheUS GeologicalSurvey. The

topical papers on these research activities have
been compiled into a book entitled "River
Morphology " (Schumm , 1972). The research
clearly showsthat erosion and sedirnentationin
rivers are slow and, often episodic, processes.
Channelerosion and sedimentationare essential
for a river to adjustto base level.

(iii) The stratigraphy of such islands, like
floodplains, usuallyconsistsof severallayers of
clay and silt, overlying fluviatile sands.
The deposition of clay and silt occur in layers
correspondingto depositionalepisodes; a single
depositional episodewould last several years. It
would take thousandsof years, even under ideal

conditions, to build a silt and clay deposit in
excess of 2m. Fieldwork around the
Sidudu/KasikiliIsland indicatedthat the silt and
clay depositswould exceed 1Om.

From the geomorphologicaland hydrological point of
view, the conclusion can be drawn that the present

equilibnumprofile of the Chobe Rivepredates the
assumed1899floodsby severalthousandsof years.To
claimthatthesefloods causedthepresentalignmentis
to turntheestablished science o gfeomorphologyon its
head! Further doubt is cast on the alleged floods in 1899 by
Captain H.V. Eason of the BechuanalandProtectorate

Police who undertook a survey of the Chobe River in
1912 -that is only some 13yearsafter the 1899floods.
CaptainEason, in his report, displayeda very keen eye
for the landscape. If, indeed, the 1899 floods were
cataclysmic,he would have noticedand commentedon
tell-tale signs on the river landscape. But there is no
such comment in his otherwise detailed description of

the landscape. CaptainEason unequivocally identified
the northem channelas the main channel. He wrote:
"Two miles above the rapids lies KasikiliIsland. Here
1 considerthat undoubtedlytheNorth should beclaimed
as the main channel ....the South Channelis merely a
backwater, what current there goes round the North. "

His report wasaccompaniedby a map which showsthe
Island accurately@NA S 35/12].

3.3 Evidencefrom Airphotographs

If floodsin 1899causedthe shiftof the mainchannelof
the Chobe River, it could be expected that the river

wouldbe very active,in geologicterms, in recent years
as it strives to re-establisha newradedprofile. In that
case, all features in the river, whether depositionalor
erosional, would be highly unstable. Such instability
would applyto size, shape, location, and topographic
height of those features. Tirne series airphotographs

would easilyrecordvariationsin theseparametersfrom
one period to theother.

With the above in mind, the airphotographs listed in
Table 2 were analysed. The analysis showed that the
shape, location and size of Sidudu Island has not

changed in the 60-yearperiod between 1925and 1985.
In fact featureson the Island are so stable thatasmall
cluster of trees on the water'sedge on the northem spit appeared on allthe airphotographs and was easily
identified by this author during the fieldwork of 10th
and 11th March 1994. There was also no discernible

change in topographic levels.

Theconclusionis reachedthatthepresent Chobe River
isa systemin equilibrium(gradedprofile)andhasbeen
sofor a longtime, certainlypriorto1899.

4. CONCLUSION

The only logical conclusion that can be drawn from the
scientificdata is ththenorthchannelis the mainchannelof

theChobeRiverinthetrue hydrogeomorphic senseof the word,
and hasbeensofor thousands ofyears.

Bond, G .(1963):PleistoceneenvironmentsofSouthemAfnca .
In: F.C. Howell and F. Bourliere (Editors) African Ecology
and Human Evolution; Aldine, Chicago;pp. 308-335.
Coates, J.N.M., Davies, J., Gould, D., Hutchins, D.G. and

Jones, C.R. (1979):The Kalatraverse OnR eeport,Geological
SurveyDepartment, Bulletin 21; Governmentof Botswana.
Cooke, H.J. and Verstappen,H. Th. (1984):TheLandfomzsof
Westem Makgadikgadi Basin in northem Botswana with a
consideration of the chronology o he evolutionof Lake
Palaeo-Makgadikgadi Z,. Geomorphol., NF, 28: 1-19.
Gilbert, G.K. (1914): Transportationof debris by running
water; US Geol. Survey Prof. Paper 86.
Knighton, D. (1984): FluvialFomzsand Processes; Edward

Arnold, UK.
Schumm, S.A. (1972) edt. : River Morphology; Dowden,
Hutchinson andRoss, Inc., Stroudsburg, Pennsylvania.
Shaw, Paul (1988): After thejloods: The Fluvio-Lacustrine
LandfomzsofNorthem Botswana; Earth Science Reviews,25,
pp.449-456.Shaw, P. and Thomas, D.S.G. (1988): Lake Caprivi: a late
Quatemary link benveen the Zambezi and middle Kalahari
Drainage systems. Z. Geomorphol., NF 32: 329-337.

Thomas, D.S.G. and Shaw, P.A. (1992): ne Kalahari
Environment, UniversityPress, Cambridge.

Dr. F.T.K. Sefe Date 17
March 1994CV OF DR. F.T.K. SEFE CURRICULUMVITAE

Francis Tetteh Kwaku
Sefe

DATE OF BIRTH: 2 June 1949

EDUCATIONALQUALIF: B.A. Hons.(Ed. C)ape
Coast; Ph.D.,
Queensland
EDUCATION HISTORY:

Studied at the

University of Cape
Coast for the B.A.
Hons. degree in
Geography
(Geomorphology) .
Other subjects studied
are Economics

(including Economics
Mathematics) and
Education 1 passed
with Second Class,
UpperDivision.

1 studied at the

University of
Queensland, Australia
for the Ph.D degree in
Geography
(Hydrology) by
research. My thesis
was in the area of

Mathematical
Modelling of River
Flow.(Degreeawarded
in August, 1983). 1alsotook examination
courses in
Sedimentology
(Geology Department)
and Photogrammetry
(Surveying
Department).

1 audited courses in
EngineeringHydrology
(Civil Engineering
Department) and in
Computer Science. 1
acquired sufficient

proficiency and wrote
the computer
programme for my
research.

WORK EXPERIENCE:

Employed by the
Architectural and
Engineering Services
Corporation
(Hydrological

Division) as Assistant
Scientific Officer
(HYdrology -

Employed as Lecturer
in Hydrology (but also
Geomorphology)at the

University of Nigeria,
Nsuka.

Employed as Lecturer
in Hydrology (but also
taught Geomorphology CHAPTERViIi

The Relevanceof Map Evidence

(a) The GeneralContext

253. The issue to be addressed is the meaning of the
phrase "the main chamel" of the River Chobein ArticleIII of
the Anglo-German Agreement of 1890. The position of
Botswanais that the main channelof the Chobe in the vicinity

of KasikiliISeduduIsland is the northern and western channel
and that the principal criterion of the assessrnentof theain
channel (German"the Thalweg ")is navigabilit.

(b) Role of Maps as evidence

254. Maps provide some assistance in the
interpretationand applicationof the terms of ArticleIII of the
Anglo-German Agreementto determinethe boundary between
Botswana and Narnibia in the vicinity of KasikiliISedudu
Island. However, their relevance is entirelydependent on the
circumstances. As the Chamber of the Court statedthe matter
in the Case concerning the Fronner Dispute (Burkina

Faso/Republicof Mali):

"Whether in frontier delimitations or in international
territorialonflicts, maps merely constitute information
whichvaries in accuracy fromcase to case; of themselves,
and by virtue solely of their existence, they cannot

constituteaterritorialtitle, that is, a document endowedby
internationallaw withintrinsiclegal force for the purpose
of establishingtemtorial rights."

(ICJ Reports 1986, p. 554at p.582 para. 54.) 255. Atbestmapsprovidesecondaryevidencerelating
to the factson the ground atany specifictime. As Sandiferhas

pointed out:
"...Tribunals ..have, however, probably appliedseverer
tests in evaluating maps than almost any other evidence.
This is due to the fact thatmaps are in most instances, at
best, secondary evidence, and frequently hearsay in
character. Consequently although the best evidence rule is
not strictly applied to maps,theyconstituteone of the best

instances of a distinctionbetween primary and secondary
evidence and of the consequences attaching to such a
distinction".

Evidencebefore International Tribunals, Revised edition,
1975, p.229.

256. The danger of reliance on maps has been
consistently reiterated by international tribunals, arbitrations
and municipal courts. Thus, in the Island of Palmas
Arbitration, Judge Huber observed:

"If the arbitrator is satisfiedas to the existenceof legally
relevant facts which contradict the statements of

cartographerswhosesourcesofinformationare notknown,
he can attach no weight tothe maps, however numerous
and generallyappreciatedthey may be".

ReportsofInternationalArbitral Awards, Vol. II, p.831at
p.853.

(c) Maps available relatingto the Caprivi Strip

257. The maps available relating to the Caprivi,
whether they come from an official or private source, are,
generally speaking, of no relevance to the interpretation and

application of the phraseumainchannel" of the River Chobe.This is for two independent but mutually reinforcing reasons.
First, the scale of the maps is usually so small as to present no

indication of theiverine topography. Secondly, when the maps
indicate the political boundary, the mode is very
impressionistic, more a general indication of direction than an
accurate demarcation by reference to geographical or
topographical features. In consequence it is unlikely that the
maker of the map wiü have adverted to the question of the
identification of the main channel.

(d) The Main Channel at the present time

258. There can be no dispute that at the present tirne,
that is at the filing on 29 May 1996 with the Court of the
parties' Special Agreement, that the location of the main
channel in the vicinity of Kasikili/Sedudu Island is the northern

and western channel of the River Chobe. The 1:50,000 scale
map 3rd edition 1984 (Map 23), prepared by the Department
of Surveys and Lands, Republic of Botswana, clearly represents
the northern and western channel asthe main navigable channel
of the Chobe, and, dong with the earlier sketch map of the
1985 Joint Survey (Annex 48) can be accepted as good

corroborative evidence that the northern and western channel
is the present main channel of the Chobe. In the absence of
evidence to the contrary, the presumption must be that this was
also the main channel at the time of the conclusion of the
Anglo-German Agreement.

259. The alternative position adopted by the
Govemment of Botswana is that, in accordance with the object
and purpose of the Agreement, the main channel is constituted
by the navigable channel at any given time, and that at present
the northern and western channel is the main channel on this
basis. The maps referred to above in paragraph 6 also support

this alternative position.(e) Evaluationof the Maps available

260.
As a practical way forward it is useful to
evaluate the maps available. Broadly they are to be
distinguished into twotypes-

those which record the facts in respect of the
main channelof the River Chobein the vicinity
of KasikiliISedudu Island at the the of the

making of the Anglo-German Agreement in
1890; and

those which indicate the views of the maker
subsequentto 1890as to where the boundary in
that area shouldrun.

(i) Officia1and privatemaps

261. A classificationof mapsinto official andprivate
can at tirnesbe of assistancein that a makerof a mappublished
officiallyby or underthe auspicesof a Statemay be in a better
positionthan a private individual to secure al1the information
on the region availableat the the, but "his trustworthinessas

a witness must depend upon the impartiality with which he
paints his picture": see Hyde, "Maps as Evidence in
InternationalBoundaryDisputes" (1933),AmencanJournalof
InternationalLaw, Vol. 27, p.311 at p.314.

262. The position has been formulated helpfully by

Judge Huber in the PalmasIslandArbitration:

"A cornparison of the information suppliedby the two
Parties shows that only with the greatest caution can
account be taken of maps in deciding a question of
sovereignty, at any rate in the case of an islanduch as
Palmas (or Miangas). Any maps which do not precisely

indicate the political distribution ofenitories, and in particular the Island of Palmas (Miangas) clearly marked
as such, must be rejected forthwith, unless they contribute
- supposing that they are accurate - to the location of
geographical names. Moreover, indications of such a
nature are only of value when there is reason to think that

the cartographer has notmerely referred to aireadyexisting
maps - as seems very often to be the case -but that he has
based his decision on information carefully collected for
the purpose. Above dl, then, official or semi-official maps
seem capable of fuKillingthese conditions, and they would
be of special interest in cases where they do not assert the
sovereignty of the country of which the Governrnent has

caused them to be issued."
(Reportsof International ArbitraA lwards, Vol. II, p.831
at p.852).

(f) Categorisationof the availablemaps

263. The available maps appear to fail into six

categories:

(i) MapscontemporarywiththeAnglo-German Agreement
(ii) Maps publishedby the German coloniaa ldministration
(iii) Britishoflcial maps
SouthAfncan oflcial rnapspublishedin 1949 or later
(iv)
(v) Botswanaoflcial maps
(vi) ThirdStatemaps. The maps available to theGovernmentof Botswanaare
as follows:-

(i) Maps contemporary with the Anglo-German
Agreement

264. British War OBce Map of 1889 (Map 3)

Thisrnap waspreparedby the IntelligenceDivisionof the War
Office and it is referred to expressly in the provisions of
Article III of the Anglo-German Agreementof 1890.The rnap
is reproduced in Hertslet'sMap of Afnca by Treaty, 3rd ed.,

Vol.IiI, facingp.902. The boundarydong the Chobeis shown
by means of a broad red band. The rnap thus provides no
evidencewhatsoeverof theprecisepositionof the mainchannel
of the Chobe either in the vicinity of KasWSedudu or

elsewhere7.

265. Map of Matibililand andAdjoining Temtones of

1889 (Map 2)

Thisrnap wascompiledin the IntelligenceDivisionof the War
Office in 1889and appears to have been a predecessor of the

rnap referred to above also dated 1889. However, this rnap
does not relateto the 1890 Agreementas finally concluded.In
any event, given its scale, the rnap conveys no information
relating to the identificationof the main channelof the Chobe.

7~ later version of this Map exists in the Botswana National Archives
(BP-139), entitled 1891 Intelligence Division No.846b, with a note stating that
it was compiled and lithographedin the IntelligencDivision by
Lt.Col.J.C.Dalton DAAG 1891:Scale 1:1,584,000, 1 inch to 25 miles.

A brokendash dot line follows the Chobe; and Sedudu Island is shown but not
named, noris it clear which side of the island the boundary passes. This is
hardly surprising as 1 millimetre on the rnap equals 1 mile.(ii) Maps publihed by the German Colonial
Administration

266. Kriegskarte von Deutsch-Sudwestafrika 1904
(Andara sheet) (Map 4)

This is on a scale of 1:800,000 and indicates "Solumbu's
Island". In doing so it appearsto showthe northern channelby

a thick black line and the southem channel is barely visible
except as the edge to the shaded area which represents the
island. There can be no doubt that on German maps
Kasikilil/Seduduwasnormally indicatedas "Solumbu'sIsland".

267. The sarne(Linjantisheet) 1904 (Map5)

This sheet in the same series and to the same scale is
reproduced in the original colouring. The island indicatedas
Solumbu's Islandis indicated as a part of the Bechuanaland

Protectorate.

268. Von Frankenberg, Karte des Caprivi Zipfels,
Blatt 1, 1912(Map 7)

Thismapwasproducedwiththeauthorityof VonFrankenberg,

the District Chiefand Residentin the Caprivi Zipfel,and is on
a scale of 1:100,000. It is dated May 1912, and exists in at
least two editions. The map clearly indicatesKasikWSedudu
and the southem channel bears the legend "Kassikhi
Fluss-Arm". This term translates (see annexure to Map7) as

"branch or tribut. of a main river". The translation also
appears in standard worksof reference. Thereis no boundary
markedas suchbut the map is importantin its identificationof
the northern and western channelas the main channel.'

ch ra nener'sMappublishedin 19In1909themappingworkcarried
out by Franz Seiner in 1905-6 was used as the basis of an impressive mapof
the Okavango andZambezi regions. The map was produced under the(iii) British Official Maps

269. Sketch Map of Bechuanaland Protectorate, to
illustrateilitary Report, 1906
CM~ 6)

This British official map shows the boundary (by means of a
pecked line)on the northernsideofthe Chobe.It was compiled
in the office of the G.O.C. in Chief, SouthAfrica, December
1905 (reference: T.S.F.S. 2199). The island is not visible.

270. SketchMap of BechuanalandProtectorate, 1912
(M~P8)

A 'Sketch Map of BechuanalandProtectorate', on a smaller
scale, appears in the Colonial Report series:Bechuanaland

Protectorate:Reportfor 1911-12 (Cd. 6007-28), London,
H.M.S.O., 1912. The map, printed at the Ordnance Survey
Office, 1912 showsa boundary to the southof the Chobe.

271. Sketch Mapof BechuanalandProtectorate, 1913

(M~P 9)

The same Sketch Map, showing a boundary to the South of
Chobe, also appears in the next volume: Bechuanaland

Protectorate:Report 1912-1913 (Cd. 7050- 14), London
H.M.S.O., 1913.

editorship of Paul Spngade on a scale of 1:500,000. The names of previous
cartographers with dates are wntten in the map Schulz and Hammer,
1884, and Reid 1896; "Bradshaw 080"ntten dong the north bank of the
Chobe immediately to the Westof KasikiliISedudu. The boundary is not shown
but Kasikili/Sedudu is marked as Solumbu's Island. It is significant that the
configuration of the island is identical to that shown on the 1880 Bradshaw map
and similar to the configuration visible on aerial photographs taken inthe penod
1925 to 1985. See further CW,ptparagraphs 237-2.1 272. SketchMapofBechuanalandProtectorate, 19 13-
14 (Map 10)

The same 'Sketch Map', showing a boundary south of the
Chobe, also appears in the next volume: Bechuanaland
Protectorate: Report for 1913-1914 (Cd. 7622-6), London
H.M.S.O., 1914.

273. SketchMap of BechuanalandProtectorate, 1915
(M~P11)

The 'Sketch Map' included in the next volume indicates a
boundary, very clearly, to the north of the Chobe: see
Bechuanaland Protectorate: Report for 1914-1915 (Cd.
7622-48), London, H.M.S.O., 1915. The map was printed at

the Ordnance Survey Office, 1915.

274. British War Office Map of Bechuanaland
Protectorate, 1933; Geographical Section, General Staff
No.3915. (Map 13)

In 1933 or thereabouts a set of eight sheets covering the
Bechuanaland Protectorate was published at a scale of
1:500,000 by the British War Office. Sheet No.2 in the series
was published in 1933. The map shows an intercolonial
boundary betweenthe Protectorate and the Caprivi southof the
River Chobe both in the vicinity of KasikiliISedudu and

elsewhere dong the Chobe.

275. The evidential value of this indication of the
boundary is substantially reduced by the following
considerations:-

(a) The boundary is shownas an "intercolonial" not as an
"international" boundary;(b) the rnap is a compiled rnap produced by technical
expertsfrom previous mapsin the absenceof reference
to legal questions.

(c) the fact that thentirealignmentof the boundarydong
the Chobeis indicated southof the river establishesthat
the rnap maker had no interest in a precise

representationof aboundary following themainchannel
of the river.

(d) In the penod 1915 to 1929 the Caprivi had been
administeredas part of BechuanalandProtectorate and
the maps relied on in the production of the 1933 rnap
relate predominantly to this period. In a period when

the United Kingdom was in functional terms the
sovereign administrator on both sides of the Chobe it
would be unlikely that the precise depiction of the
boundary on a rnap drawn on a scale of 1:500,000
would be a matter of particular concern.

276. BechuanalandProtectorate, 1960:Directorateof
Overseas Surveys(D.O.S. (Misc.) 282) (Map 16)

This rnapis statedto have been: "Drawnand photographedby
D.O.S. from material suppliedby the High Commissionerfor
Basutoland, the Bechuanaland Protectorate and Swaziland".
The sheet indicates the boundary dong the northem and

western channelof the Chobe. In view of the small scaleno
detail is visible in the vicinity of KasikiliISedudu.

277. Bechuanaland, 1965 : Directorate of Overseas
Surveys (D.O.S. 847) (2462)) (Map 17)

This is sheet 2 of a series of sheets at a scale of 1:500,000

published by the D.O.S. in a first edition in 1965. The
boundary is marked actually along the Chobe and it clearly
follows the northern and western channel in the vicinity ofKasikiliISedudu. The sheet carries a disclaimer: "This rnap
must not be considered an authority on the delimitation of
international boundaries"

(iv) South African OfficialMaps publishedsince 1949

278. South AfricaOfficiaiMap1:250,000, 1949(Map
14)

In 1949 thefirst edition of a South Africanofficia1rnap on a
scaleof 1:250,000 was published (ina seriesof five) with the
title'KatimaMulilo' .The rnapwascompiledand drawnbythe

UnionDefence Forces, Survey Depot (Technical)in 1945 and
printed by the Govemment Printer, Pretoria, in 1949. The
boundaryinthe vicinityofKasikiliISedudufollowsthe southem
and eastem channel.

279. The evidence available indicates that this rnap
reproduces,without any furtheror independentverification,the

boundary indicatedon theWar Officernapof 1933(see above,
paragraphs274 and275). Thusthe authoritieslisted includethe
War Office rnap of 1933.

280. In any case, even if it be assumedthat the map
of 1949was published after consultationwith ExternaiAffairs

in Pretoria, the evidentialsignificanceof the sheet is removed
by the fact that by 1949 the South African Government was
well aware of the dispute relating to the boundary as a
consequence of the correspondenceat a high level resulting
from the dispute between Trollope and Redmanin 1948: see
Chapter V, paragraphs 153to 159.In terms ofthe admissibility
of evidencethisrnap is self-serving,not havingbeen produced

ante litemmotam, that is, before the parties wereaware of the
dispute. It is a general principle of the law of evidence that
declarationsasto publicor general rightsmusthavebeenmade
before a dispute has arisen over the right in question: see
Hoffmann, The South Afncan Law Of Evidence, Durham,1970, p.117; Halsbury 'sLaws of England, 4th ed., Vol 17,
1976, p.59 (para.80); Phipson on Evidence, 4th ed. by

Howard, Craneand Hochberg, 1990,pp.738-9 (paras. 30-32).

281. The sarne general principlesare to be found in
the sources of general international law in the form of the
principle that the parties to a dispute cannotby their conduct
subsequent to the critical date (when the dispute crystallised)

improve their position: see Fitzmaurice, British Year Bookof
International Law, Vo1.32 (1955-6), pp.20-44. In the present
case it is very probable that thecritical date falls within the
period 1948-51.

282. The 1949rnapis a compiledrnapand one of the
authoritieslistedisthe War Officernapof 1933.The alignment

indicatedtherefore results fromthe factors indicated above in
paragraph 275. In particular, the alignmentshownis unrelated
to the provisionsof the Anglo-GermanAgreementand cannot
be taken as an interpretationthereof.

283. South African Official Map, 1:250,000; 1967

(M~P19)

The 1967 editionof the same rnap does not list the authorities
usedbut it isobviouslybaseduponthe 1949rnapand therefore
shares its deficiencies.

284. South African Map Compiled by JARIC,
1:100,000 (c.1974) Wap 22)

This is an official South African rnap clearly indicating the
boundarydong the centre of the northernand western channel
of the Chobe. Indeed, Kasikili/Sedudu is not shown as an

island andthe southern channelis not shown.The provenance
of the rnap is the Southfrican DefenceForces. The rnap was
compiledby a unit knownas JARIC (Joint Air Reconnaissance
Intelligence Centre) (HLVS is the Afrikaans equivalent). Themapbearsno datebutall the indications(forexample,the table
for magnetic variation gives the correction of 1974) suggest
thatit was produced in 1974or 1975. The map forms part of
a seriesprimarily of the eastem Caprivi Strip.

(v) Botswana Official Maps

285. Republic of Botswana, 1:50,000 (Dept. of
Surveysand Lands, 1984) (Map 23)

Maps produced by the Republic of Botswana since
independence consistentlyindicate the boundary as following
the northern and western channel. An example is the sheet in
the 1:50,000 seriesof whichthe third editionwaspublished in
1984 (Sheet 1725 C3 and Part C4). These sheets all carry a
disclaimer. No doubt these maps were prepared after the
controversy had surfaced, as in the case of the SouthAfrican

official maps, but their existence demonstratesthe absence of
any subsequent practice (with reference to rnaps) which
establishes the agreement of the parties' regarding the
interpretationof the Anglo-GermanAgreement inaccordance
with Article31 (3)(b)of the Vienna Conventionon the Law of
Treaties.

(vi) Third State Maps

286. Republic of Zambia, 1:250,000 (Sesheke),
Surveyor-General,Lusaka, 1971 (Map 20)

This sheetin the series(Sesheke, SE-35-5)showsthe boundary

along the northem and western channel. The sheet carries a
disclaimerconcerninginternationalboundaries. 287. Zimbabwe. 1:250,000 (Kazungula),

Surveyor-General,Zimbabwe, 1981 (Map 21)

This sheet in the series (Kazungula, SE-35-519) showsthe
boundaryalongthenorthemand westernchannel.The southem
channelof the Chobeis not shown.

General analysis of the availablemaps
(g)

(i) No rnap as an integral part of the Anglo-German
Agreement

288. After setting out the course of the line which
should bound the sphere of influence reserved toGermany,

ArticleIII of the 1890 Agreementprovides:

"The course of the above boundary is traced in general
accordance with a rnap officially prepared for the British
Govemmentin 1889".

No rnap was annexed to the Anglo-German Agreement1890.
Hertslet in hisMap of Afnca attached a rnap to illustrate the
line specifiedin the Agreement butthe scalewas too small to
indicate its exact demarcationinconnectionwith a channel of
the Chobe; the rnap whichit was based on, Map 84623of the
Intelligence Division of the War Office, was equally

uninformative.It is thereforeclearthat no rnapwas in any way
agreed or accepted by the parties as an integral part of the
agreement.

289. As a Chamber of the Court explained the
position:

"Of course, in some cases maps may acquire such legal
force, but where this is so the legal force does not arise
solely fromtheir intrinsic merits; it is because such maps
fall into thecategoryof physical expressionsof the will of the State or States concemed. This is the case, for
example, when maps are annexed to an official text of
which they form an integral part. Except in this clearly
defmed case, maps are only extrinsic evidence of varying

reliability orunreliability which may be used dong with
other evidence of a circumstantial kind, to establish or
reconstitutethe real facts"

Case Conceming the Frontier Dispute (Burkina
Faso/Republicof Mali), ICJ Reports, 1986, p.582, para.

54).

(ii) Too smallscale of maps

290. The dimensionsof the island andthe channelsof
the river are to be kept in mind whenassessing the relevance

of maps. The greatest width of the river at any point in the
vicinity of KasikiliISedudu island does not exceed 250m.
Having regard to the fact that, at its widestpoint, the distance
north to south acrossKasikilUSeduduIsland from the northern
bank of the north channel to the south bank of the southem
channel is 2.85 km., and the distance east-west across the
island from the east bank of the eastem channel to the west

bank of the western channel is 2 km., it wiil readily be
appreciatedthat a map of smaller scalethan 1 100,000or 1cm
to lkm (1' inch to 1 mile) is likely to be of little use in
representing accurately the configuration and size of the
channels andisland.

291. Very few of the maps, sketches and other
surveys referred to by the parties, are drawn to such a large
scale. They include:-

the sketch map of Dr. Benjamin B.Bradshaw of 1880
drawn to a scale of 1.lcm to 1 mile, See Chapter VII,

paragraph 232 and Map 1;
the Frankenberg map, the Karte des Caprivi Zipfels,of 1912 (1 :100,000), Map 7;
the rnapprepared by the South African Defence Force

in 1974 (the 1974 JARIC map) (1 :100,000), Map 22;
the rnap prepared by the Joint Survey 1985, (approx
1:15,000), Annex48;
and the seriesof maps ata scale of 1:50,000 prepared
by the Dept of Surveys and Lands, Republic of
Botswana from1974,Map 23.

292. It is to be notedthat athese maps clearly show
the northem and western channel as the main channel of the
Chobe.

293. In this connectionit is usefulto bar in mindthe

TabaCase, where, confrontedwith the task of identifyingthe
exact locationof boundarypillars by referenceto a rnapdrawn
to the scale of 1100,000,the Tribunal said:-

"The Tribunal does not consider these map-based
indications to be conclusive since the scale of the rnap

(1:100,000) is too smallto demonstratea location on the
ground exactly as required in these instances where the
distancesbetween disputedpillar locations are sometimes
only a few metres. By way of illustrationit is sufficientto
recall thata rnapof the scaleof 1:100,0001 millimetreon
the rnap represents 100 metreson the ground",

Case concerningthe Location of Boundary Markersat
Taba (Egypt/Israel),Reports of International Arbitral
Awards, Vol. XX, p.4 at p.48, para. 184.

(iii) Unreliabilityof boundariesdrawn on maps

294. Except in very large scale maps the drawing of
a riverine boundary by means of a superimposed line on
geographical features(whethercoloured or broken or in some
other way differentiated fromthe line indicatingthe course of
the riveritself), inevitablyproduces inaccuracyand distortion.To achieve visibility and a representation of the general
direction of the boundary the superimposed line may first
extend across the full width of the river, even in some cases
encompassing bothbanksas well as the river itself, or second,

be placed along one or other of thebanks of the river.

295. Examplesof the first are to be foundin the maps
of 1889 and 1890 prepared in the Intelligence Divisionof the
War Office which were subsequently used to illustrate the
terms of the 1890 Agreement. The line used to indicate the

southem boundary of the German and British spheres of
interest in the lower reaches of the River Chobe extends well
beyond the banks of the river on eitherside; given the maps'
scale of 1;584,000 or 1inch to25 miles, the boundary line of
1132 of an inch is equivalent to nearly half a mile on the
ground.

296. Examples of the second are to be found in the
1905 SketchMap to illustratethe Military Report, producedby
the Officeof the General OfficerCommandingin South Africa
which showstheboundaryfollowingthe northem bank;and the
maps produced by the Ordnance Survey Department for the
annual Colonial Reportson the British Protectorate presented

to the British parliament; the maps attached to the Colonial
Reports for 1911112, 1912113,and 1913114(Maps 8 to 10)
show a boundary line following the south bank whilst that for
1914115(Map 1 l), showsthe line following the northbank of
the Chobe.

(h) Conclusionon the Map Evidence

297. Themapevidenceis contradictoryandconfused.
It is not simplythe casethat the officialmaps of Botswana and
South Africa (with one important exception)do not constitute
subsequent practice 'which establishes the agreement of the

parties' regarding interpretation of the Anglo-German
Agreement in accordancewith Article 31 (3)(b) of the ViennaConventionon the Law ofTreaties. In addition, the different

seriesof officia1maps are intemally inconsistent.

298. ThefourBritishmapsof 1912to 1915contradict
each other. Thus the 1915 rnap shows the boundary on the
northem side of the Chobe. The 1933 War Office Map is
contradictedby the SketchMap of 1915and also by the maps
of 1960 and 1965. Likewise the South African maps of 1949

and 1967are contradictedby the SADF rnapof circa 1974(the
JARIC Map).

299. In any case the value of the rnap evidence
subsequentto the years 1948-51is reduced by the fact that the
existence of the disputebecame known to the parties in that

penod and the parties are not permitted to improve their
positions subsequentto the critical date.

300. In this contexttheJARICrnapis of considerable
importance. It was produced whilst South Africa was still
illegallyinpossessionof Namibiaand it was produced (it may

be presumed) for interna1 use. It thus constitutes the
authoritativeprofessionalopinionofthe South AfricanDefence
Forces and this opinionwas expressedin a context (patrolling
and other military activities) in which the location of the
boundary was of great relevance. It is perfectly normal for
intemational tribunals to rely uponmilitary mapping: see the

Frontier Land Case, I.C.J. Reports, 1959, p.209 at p.227
(reference toBelgian military staff maps). Moreover, in the
TempleCase (Merits)the InternationalCourt referredto a Thai
officia1rnap of 1937 and observed: 'That thisrnap may have
been intended for intemal military use does not seem to the
Courtto makeit anylessevidenceof Thailand'sstateof mind'.

I.C.J. Reports, 1962, p.6 at p.28. CONCLUSIONS

The Governent of Botswananow presents its conclusionsin
summaryform.

First:
The centralquestionis the interpretationand applicationof the
words 'main channel' of the River Chobe. These words
involve a reference to a question of fact and, in so far as may

be necessary, a questionof scientificfact, callingfor expertise
in hydrology, geology and hydrogeomorphology.

Second:
The assessment of the facts should reflect the objects and
purposes of the Anglo-German Agreement. For present

purposes the relevant object and purpose was that the riparian
States should have accessto watenvays and, in particular, to
the navigableparts of the majoroundary rivers in the region.
It must follow that in case of any doubt the channel which
qualifies as the 'main channel' isthat which is navigable by
cornparisonwith anyother candidate channel.

Third:
The scientific evidence available shows that there is no
possibilitythat the riversystemhasundergoneanymodification
since 1890 of a kind which would change the status of the
channelsin the vicinity of KasikiliISeduduIsland.

Fourth:
It follows that the evidenceavailable as to the location of the
main channelin the penod 1912to 1985is relevantto the issue
of the locationof the mainchannelin July 1890. The evidence
includesthe following:-

(a) The two maps produced under the authority of Von
Frankenberg, the GermanDistrict Chief in the Caprivi, in May 1912, showing the southem channel as the
'fluss-am' that is, the branchor tributaryof the Chobe.

(b) The assessmentby CaptainEasonin his Report dated 5
August 1912.

(c) A remarkable senes of aerialphotographsfor the years
1925, 1943, 1947, 1962, 1972, 1981and 1985.

The conclusionsof the JointSurveycarriedout in 1985.
(d)

AU the evidenceavailableindicatesunequivocallythat in 1890
the northem and westernchannelwasthe 'main channel'of the
Chobe.

Fifth:
There is a complete absenceof evidenceto the effect that in
1890, or thereabouts, the southemchannelwas identifiableas

the main channel.

Sixth:
The Pretoria Agreement of 1984 is an intergovernmental
agreementwhichconstitutes'anagreement betweenthe parties
regarding ...the application'of the provisions of the Anglo-
German Agreement. This agreementwasduly irnplementedby

means of the Joint Survey and the resulting Joint Suwey
Report,whichisitselfanintergovernmentalinstrumentderiving
from the Pretoria Agreement and having a legally binding
character.

Seventh:
Independently of its status as a binding intergovernmental

instrument the Joint Suwey Report constitutesexpert opinion
evidence on the key question, the identificationof the 'main
channel'of the Chobe.Eighth:
The subsequent practice of the parties (and their successors)
can only be taken into account if it establishesthe agreementof

the States concemed regarding the interpretation of the treaty.
The only subsequent practice which is legally pertinent in the
present case is the Joint Suwey Reportof 1985. However, it
is more appropriate to classify the transactions of 1984 leading
to the Joint Survey Report in 1985 as a 'subsequentagreement
between the parties regarding the interpretation of the treaty on
the application of its provisions' in accordance with the Vienna

Convention on the Law of Treaties.

Ninth:
The evidence of official maps available shows no consistency
and therefore does not evidence either 'the agreement between
the parties' regarding the interpretation of the Anglo-German

Agreement in accordance with Article 31(3)(a) of the Vienna
Convention, or a subsequent practice 'which establishes the
agreement of the parties' regarding interpretation of the Anglo-
German Agreement in accordance with Article 31(3)(b) of the
Vienna Convention.

Tenth:

Notwithstanding the general status of the oflcial maps
indicated in the previous conclusion, individual maps have
evidential value on their own terms. Individual maps with
discrete probative value include the Von Frankenberg map of
1912 andthe South African military map compiled by JARIC.

Finally:
The northem and western channel of the Chobe in the vicinity
of KasikiliISeduduIsland is the 'mainchannel' of the Chobe in
accordance with the provisions of Article III(2) of the Anglo-
German Agreement and the intergovernmental Joint Suwey
Report of 1985. SUBMISSIONS

Haviag regardto the considerations setforth in the Memonal
presentedon behalf ofthe Republicof Botswana,

May it please the Courtto adjudge and declare that:

(1) The northem and western channelof the ChoboRiver
in thevicinity of KasikiliISeduduIsland constitutesthe
'mainchannel'of the Chobe River in accordancewith
the provisions of Articlem(2) of the Anglo-German
Agreementof 1890;and that:

(2) Consequently, sovereigntyin respectofKasikili/Sedudu
Islandinheres exclusivelyin the Republicof Botswana.

(signed) Abednego Batshani

Tafa,
Deputy Attorney-
General.

Agent of the Republicof Botswana