4'
THE JOLJ"ALOF THE UGANDAS'

VOLUME 20, No. 2 SEPTEMS'Ekk

C 0 Atenta-
_S
The KiYu MWon, 1909-10 M., C'mm
4" Aweats of Mount ElgOn
T46"mm &F. 1. LMSFAi 414"1
ne Fisheg of Uganda-11 P. H.'GUENW600 'j
Studies in Uganda Swarh* DR, .. E. M. L"
Historlo Sites in Ankole R F. Moms,
Two Sureys of Kasilang Eronyi, Tiso

Muhindi or G4ain of Arabia Da.W. D. W. PBFF"ys "'M
The Third Pan-African Congress on Prehistory,.- E. Q. LAiSm4f

NOTES
The Wilsops- of Early Uganda R; B. TiiomA 21
The Teso and the Karamojong Cluster DR, P, H, GULLtVER

CORRESPONDENCE
hafts in )BpSanda. 4nd TOP

OBITUARY 40,
MAUD 'S, F"j) 21$L

A "aapd Cyrammar (by A. N. Twker and J. Tompo ole Mpaayei)
W. K WIXITIP3
Gxafogie du Congo Relge (by L.Cithen), I.. W1.1 PAUjS'r]P-k jZj'"

Index to Volume .20: of The Uganda: kumd ';'v

PLuNbr4ed by
THE U G A N,'D A S 0 C
XAM'PALA
price AS. 10 (100:
114

Pawn
Ms ftaAo4CYSar Audrew'Cohtn, Y-cm.G,, K.c.vo;. 0AIL

X
Mr.,15AL marphaLua M -MarkDarringtonc Ward

C D. Lawrance
The HOU. Wfttary Mr.M. N,.Majai
The Ron. IreAawer Mn R 1. 1 Mehta
ThoHon., Eltors
Mr. C,N. Mukuyc,
Th-c Ifon.' Librarian Mr. L'.F. SaIdanha
M. L. Fafts
Mr. W. L A. Harris Mr. A K. Sempa
r I r. H. C. Trowell, o.B.B.
;T'I Mr, E W. Nigundu Dr. K. P-Wachsm"A
T'T
Hon. Secretary; Mr. H. F Morris
Ron. Trea=rer: Xrs M. M.WaHis,
Hon. Mitors., IDr W. %Othaff,
Mr. J, W. Paluster
Hon. Librarian: Mr.' N. D. Gram
Ron. Auditor Hon. Legal A&isei.
Mr. A. H. Stump, A.A.C.C.A. : I.. I
Mr. C, L. Holconx
Coreeoonding Secretaryat Jinja, Mr. T. W F. Cox, cmia,
Corresponding Secretary at Mbale Mr. F. Lpkn WMiam
'Corresonding Secretary atMwaka: Mr. F" C. Jawu-rlg
corrsvindfng, Secretmy at serero: Mr. 1), T.,Parsons
Hon. Vkr-Proldtns:
Mut"a 11; K'abaka of Uugaada Sir J-Ohq Wtr Gray
A. Sir' Tito WIA Gifibusa 'IV, Mr. E. B. Ha4ddon
7", C.LIML, Ornuiam of Toro Mr. H 'H, Thomas; 6.n_&
''T'Sir,& E Twiging, Gx-M-G., na-f-
Payt Presidents:
Sir A. R-6ook, cm.q., o.wp_ 1944-45 Dr. K., A. Davies, cKo,,,o.n.E.
,404-35 Mr-E. 1. Wayland, i%Ba 1 45--46 -Mr. G.' K F_ Hop"m, mw&
736 Dr. K W HuntM,0,B.15, '114"7 Mrs. K,
7, Mr. H. lowitt, C., 4. 19448 Or. W;J, Eggcling
'38 Sit-I& F_ Hone, 1948-'50 Or"G. 4p CyTirath
195041 Rov. Fr E 13. GLaffney, wx.
MT-39 Mr. J. $Ykes, OX.P. 1951-52 Dr. A.W. Williams-
Mr. N. Y., Waimea binson,
I Captain, C, fL S, Pitman, 1952-53 Sir J. B.L Hutc
tpl% C.KQ,
"', '4 C.R;B" %S.O".
IN' 1953-54 Mr. J. D. Jamegon, oaiN,
,4441-42 Mr. S. W. Kiftbya, cBx.
6't_ 5 Dr. Aufty 1. Richards, cBs.
iLkj-43 Mr, IR. A. TemploPerkins
;T19 -44 MTLR. A. $noxalt '195-56 Dr4 HX. TroweJL o.-B.E.

Mrs. K.M. TroweU. m-sx Mr., B. K. Mtflyan Mr. G. P,
Secre;ary: MA. M. M, Waffis

The

Uganda Journal

THE JOURNAL OF THE UGANDA SOCIETY

VOLUME 20
SEPTEMBER 1956

No. 2

EDITORIAL COMMITTEE
DR. A. W. SOUTHALL
J. W. PALLISTER
MAJOR B. G. KINLOCH, M.C.
DR. K. P. WACHSMANN
(assisted in England by H. B. THOMAS, O.B.E.)

Published by
THE UGANDA SOCIETY
KAMPALA

PRINTED IN GREAT BRITAIN BY HEADLEY BROTHERS LTD
109 KINGSWAY LONDON WC2 AND ASHFORD KENT

Contents

ARTICLES
The Kivu Mission, 1909-10
Early Ascents of Mount Elgon

H. B. THOMAS

J. M. COOTE

AND R. F. J. LINDSELL

The Fishes of Uganda-II P. H. GREENWOOD
Studies in Uganda Swamps DR. E. M. LIND
Historic Sites in Ankole - H. F. MORRIS
Two Surveys of Kasilang Erony, Teso
P. N. WILSON AND J. M. WATSON
Muhindi or Grain of Arabia DR. M. D. W. JEFFREYS
The Third Pan-African Congress on Prehistory E. C. LANNING

NOTES
The Wilsons of Early Uganda H. B. THOMAS
The Teso and the Karamojong Cluster DR. P. H. GULLIVER

CORRESPONDENCE
Shafts in Buganda and Toro E. C. LANNING

OBITUARY
Mr. K. Hay Dale, O.B.E. MAUD S. BUDD

REVIEWS
A Maasai Grammar (by A. N. Tucker and J. Tompo ole Mpaayei)
W. H. WHITELEY
Gdologie du Congo Belge (by L. Cahen) J. W. PALLISTER

Index to Volume 20 of The Uganda Journal -

Page
105

113
129
166
177

182
198
202

NOTES ON CONTRIBUTORS

R. F. J. Lindsell joined the Uganda Administration in 1935. Since 1951 he has been
District Commissioner at Mbale, in charge of the Bugisu District, and has become
well acquainted with Mount Elgon, which he has climbed four times.
P. H. Greenwood, Senior Scientific Officer of the East African Fisheries Research
Organization, has, since 1951, studied the fishes of Lake Victoria and Uganda.
Before joining the E.A.F.R.O. he held a Colonial Office Fisheries Research Student-
ship and studied at the British Museum (Nat. Hist.) for a year.
P. N. Wilson, Lecturer in Animal Husbandry at Makerere College, has been resi-
dent at Serere in Teso from 1952 till 1956 and carried out original research on goat
physiology.
J. M. Watson, formerly of the Uganda Agricultural Department and Director of
Agriculture in British Somaliland from 1953 till 1956, is already a notable con-
tributor to the Uganda Journal.
H. F. Morris joined the Uganda Administration in 1948. From 1950 till 1953 he
served as a District Officer in Ankole and he is engaged on a study of the heroic
recitations (ebyevugo) of the Banyankole.
M. D. W. Jeffreys, until recently Senior Lecturer in Social Anthropology at the
University of the Witwatersrand, is a well known contributor to many journals of
African interest.
E. C. Lanning joined the Uganda Administration in 1948, after war service and
varied experience in both East and West Africa. While stationed in Masaka and
Mubende, he was able to develop a longstanding interest in archaeology by investi-
gating and recording many of the sites in which these Districts are so rich. Some of
his findings have been presented in previous issues of the Uganda Journal.

THE KIVU MISSION, 1909-10

By J. M. COOTE1

I GOT back to Uganda from European leave towards the end of December
1908 and was posted for a second tour to the old Bukedi District. Almost at
once my District Commissioner, Ormsby, died, and I had to take charge at
Mbale in his place.2
It was at Mbale at the end of May 1909 that I received an urgent summons
by runner (no telegraph lines or telephone in those blissful days!) to report at
Entebbe. I bicycled through to Jinja at speed (2-4 June), was picked up there by
the Government steamer,3 and taken to Entebbe where Boyle4 was at the time
Chief Secretary, Stanley Tomkins5 having just become Acting Governor on Sir
Hesketh Bell's departure. I was told that I was to go as Political Officer in charge
of a mixed force of Sikhs, K.A.R. and Police whom I would find awaiting me at
Mbarara, and was to establish a post on Lake Kivu and administer the surround-
ing district.
I am not surprised that so little is known about this Kivu trip, as the whole
expedition was organized and controlled by the folk in London from whom
came my instructions and to whom my reports were sent.
Boyle handed me a map of a strip of territory given us by the Germans which
he understood had been especially negotiated in order to give us access to Lake
Kivu without trespassing on Belgian territory west of the 30th meridian.
I went on by the steamer to Bukakata and made Mbarara in two days arriving
on 10 June. There I found the escort consisting of Captain de Courcy Ireland
with a detachment of Sikhs, and a company of King's African Rifles under
Captain Hall with two subalterns one of whom was Campbell.6
We were a motley crowd! I was in charge unless and until hostilities broke
out: and though they never did, it was at one time a very near thing.
No one had thought of collecting porters and posho, so everything was in
confusion. Watson7 was District Officer and thought we were assembled to
intimidate the Belgians from infringing a neutral strip twenty miles wide which
had been agreed as a temporary boundary to the west of the Ankole District
pending discussions in Europe.
However somehow we got going next day (11th) and made for the south-west
corner of Ankole; Watson, by the way, was not allowed to know our destination
presumably so that the Belgians would not get wind of it. The War Office
explained all this hush-hush to me subsequently.
The country through which we had to trek between Ankole and Kigezi (the
latter not known by that name until we occupied it) was a series of steep hills
divided by swamps. It was uninhabited, and no Ankole native would guide us,
or in fact ever ventured into the area. We looked at the War Office map and
then at the jumble of hills, and gave it up. We spotted in the dim distance the
snow-capped Mufumbiro mountains, took a bearing on them and steered
accordingly. I have forgotten how many days we took before we emerged in

106 UGANDA JOURNAL
what became the Kigezi District, where I was delighted to find friendly natives
and plenty of posho. I had come armed with beads and had no difficulty in
buying all the food we needed. We reached the neighbourhood of the future
Kigezi station on 19 June, where we found a Swahili speaking headman and
learned about the Belgian post at Rutshuru of which I had not been warned.
Since it was obvious that news of our incursion would get through to Rutshuru
pretty soon we trekked on all night and reached the top of a track across a spur
running down from one of the Mufumbiro mountains. Here we looked down on
Lake Kivu about twenty miles away: and here also I was astonished to learn
that there was a German officer8 right on the shore of the Lake more or less on
the spot where I had been told to stick up the Union Jack!
This was an unlooked-for complication, but I decided to go and call on the
German. I sent Ireland on to the Lake about ten miles west of the German post.
We had dropped Hall at the pass, and I went on alone with an askari to the
German post. The officer nearly had a fit! He was far too cut off from Tabora to
have been told anything about us, even if Tabora had heard, which I doubt.
I asked him what he was doing there as I had been told it was Belgian. He
said the Belgians had had the same idea but he had pushed them back to the
west side of the Lake where he was quite prepared to keep them! I told him
this was fine, as I had no instructions to invade German territory, and if he could
undertake to prevent the Belgians from occupying the bit we wanted, I would
report back to Entebbe and await further instructions. I joined Ireland late that
evening (24 June 1909), spent one night on Kivu and trekked back next day into
the Kigezi District. There was no sign of any Belgians, so I sent a letter in to the
Commandant at Rutshuru9 and told him that I had been sent from Uganda to
take over the district and suggested a meeting to arrange boundaries. I mentioned
for good measure that Rutshuru appeared to be in British territory! I should
love to have been present when he got the letter. The effect was instantaneous.
Next morning (28 June) at dawn an enormous safari of Belgian askaris and
officers was seen approaching. My soldier escort was immensely elated: they
polished up their machine guns, placed outposts and prepared for a scrap! The
senior Belgian officer shortly emerged escorted by an askari with an enormous
white flag. He demanded our immediate withdrawal, etc., etc. I asked him up to
lunch to discuss it over a bottle! He astonished me by accepting.
Hardly had he left before a runner arrived from Mbarara with two urgent
telegrams from London;
(1) ordered us not to start,
(2) addressed to the D.C. Mbarara told him if we had started to send post-
haste to recall us!
There seemed nothing for it but to obey orders and go back. I assumed that they
had learned in London of the existence of the German post and of the Belgians
at Rutshuru.
That evening at dinner in the Belgian camp I came to an arrangement with
the Belgian officer, whereby if he withdrew next day to Rutshuru I would go
back the following day to Uganda, and we could both then await instructions
from our respective headquarters. The Belgian army duly withdrew next morn-

THE KIVU MISSION, 1909-10 107
ing. We had to await the arrival of an outpost we had left at the pass, but started
back the day after. The Belgians cannot have been far away as I left half-a-dozen
askaris behind with some porters collecting posho but they were taken prisoner
almost before we were out of sight.
Well, back we marched, and hardly had we reached Mbarara (14 July) when
there was yet another telegram from London telling us to re-occupy! This time
anyhow we knew the road. Hall and Campbell did not come with us on this
second trip. Nearly half the Nubi company was unfit, so 50 police under
McCombie10 were substituted. The orders this time were a bit more explicit, we
were to get as near Kivu as we could and were to dig in if attacked; but were
not to start an offensive. By this time the Belgian army had completely occupied
the district and we took up a position (31 July) just out of rifle range on an old
crater which became known as Kigezi.
The Belgians" sent me an ultimatum before we had downed loads and our
soldiery felt sure of a scrap! My reply to the ultimatum pointed out that we were
not the sort of people to launch filibustering expeditions, and that he had better
wait for orders before he began shooting! He did not start shooting, but every
morning for a week I received an ultimatum at dawn telling me to leave by
mid-day or-! They marched round our hill daily just out of range. But after a
bit things settled down and we merely sat and glared at one another: and when
they got used to my unwelcome presence I used to hunt the pigmy elephants
that frequented the slopes of the Mufumbiro mountains.
During our absence the Belgians had taken it out of the natives who had
helped us with food. Their askaris had shot up one village wounding several
of the natives and taking away some of the girls. I reported this to London and
weeks later a Belgian judge was sent out from Leopoldville to prove how wrong
I was. But I was not asked to attend the enquiry.
I remember that whilst I was at our headquarters at Kigezi both Leopold II
(17 December 1909) and Edward VII (6 May 1910) died, and the Belgian Com-
mandant and I exchanged official condolences.
This was about the end of the interesting part of our operations. London,
Brussels and Berlin got together and agreed that a joint boundary commission
should meet on the spot and fix up a boundary. Major Jack who had been with
the 1907-08 Commission, was being sent out from London, but would not be
able to leave England until the end of the year. Entebbe wanted me to stay and
accompany the Boundary Commission: but, by this time, I had had enough of
Kigezi and asked to be relieved.12 An Assistant District Commissioner" was sent
to take over and was later with Jack's Commission as political officer. I left
Kigezi on 3 November 1910 and returned to Mbale.
I asked the General or whoever it was I saw at the War Office when I went
home in 1911, why they were so set on having this particular piece of Africa,
and he said that for one thing Speke had been the first to discover it: and
secondly it would make a nice summer resort, a sort of Simla, for Uganda
officials. Having trekked four times over the intervening hills and swamps, I
didn't at the time think much of the secondly!
Well here you have the true and probably only today-known account of the
Kivu Expedition of 1909.

108 UGANDA JOURNAL

EDITORIAL NOTES
I John Methuen Coote. Uganda Administration from 24 November 1905: District
Officer 1909: resigned 2 January 1912. For a while in 1917 during the German East Africa
campaign he was Liaison Officer with the Belgians following the occupation of Tabora.
2 S. Ormsby died Mumias 21 January 1909.
3 This was the Railway Marine s.s. Sybil.
4 Later Sir Alexander G. Boyle, K.C.M.G. served in Uganda 1895-1910: Provincial
Commissioner Eastern Province 1905-10. Died 18 April 1943.
5 Stanley C. Tomkins, C.M.G. Served in Uganda 1896-1911: Chief Secretary 1908-11.
Died 17 December 1946. There was no substantive Governor in Uganda from April 1909
to April 1911.
6 Captain C. R. Hall, Royal Munster Fusiliers: Lieutenant N. A. H. Campbell,
Somerset Light Infantry. For the other subaltern Mr. Coote is, perhaps, thinking of
Lieutenant E. G. D. Lardner, Liverpool Regiment, who, as is explained later, did not join
until the Mission was returning from its original dash to Lake Kivu.
7 A. H. Watson. Administrative Service: in Uganda 1902-23. Provincial Commissioner
1919.
8 Herr Kraus.
9 Captain Wangermie; with him was Lieut. A. T. Brochard. Wangermee had recently
been in friendly association with the British, when (April 1908-February 1909) he was a
member of the Belgian section of the Measurement of the Uganda portion of the Arc of
Meridian 30" east of Greenwich. Mr. Coote's MS. journal records that he and Captain
Ireland had their chairs carried 300 yards outside the perimeter: on these they sat to await'
the approaching Belgians.
10 L. H. D. McCombie. Administrative Service: in Uganda 1906-15. Killed on active
service at Dwagandu, Kagera Military District, 3 July 1915, while on duty as Assistant
Intelligence Officer.
11 Wangerm&e's superior officer, Frederick Olsen (a Swede by birth), who was Com-
mandant-Supdrieur du Russissi-Kivu, had now taken command, and proved much more
intractable in negotiations.
12 It should be added that the real reason for Mr. Coote's withdrawal was that his
health was seriously affected by the great hardships which he had undergone; there was
no medical officer with the original force which advanced to Kivu. Moreover by this time
it was clear that the Belgians would be retiring from Kigezi, and that a period of inaction
lay ahead. There are on record glowing official tributes to Coote's conduct of the affairs
of the Mission. He had experienced over twelve months of ceaseless wrangling with the
Belgians who maintained armed posts on Muhavura and at Kisoro.
13 T. V. Fox. Administrative Service: in Uganda 1908-18: died 31 May 1918.

POSTSCRIPT
Mr. Coote's narrative draws the curtain from one of the more obscure inci-
dents of the Protectorate's recent history, for published records contain hardly
a mention of the Kivu Mission of 1909-10.14
In the background is the tangled story of the Anglo-German-Belgian bound-
14 No notices of appointment appear in the Uganda Oficial Gazette. The most alert
intelligence officer of a foreign power would be unlikely to draw any deduction from a
notification (5 July 1909) that H. Pellew Wright, 3rd Class Magistrate, was granted powers
of a 1st Class Magistrate "during the absence of the District Commissioner from Mbale".
The earliest public reference to Kigezi seems to be a Notice, 27 September 1911, appoint-
ing "the officer-in-charge Kigezi to exercise jurisdiction within the area in the Uganda
Protectorate administered from Kigezi".
A reference will be found in Lieutenant-Colonel E. V. Jenkins' A History of the 4th
The King's African Rifles, published by the Government Press, Uganda, in 1911, page 21:
"the Battalion was called upon in March 1909 to supply troops for the Lake Kivu Mission,
undertaken with a view to effectively occupying territories claimed by our Government
in the Mfumbiro, Lake Kivu districts.
"The British Mission consisted of a Political Officer, 4 Military Officers, 70 Sikhs,
2 Sudanese companies, and some Police. The objects of the Mission were successfully
carried through, but not without our friendly relations with the Belgians being strained
almost to breaking point."
The Sikhs were from the 5th (Indian) Battalion, K.A.R. which was disbanded in 1913.

THE KIVU MISSION, 1909-10 109
aries in south-western Uganda which were settled by an agreement signed by
delegates of the three Powers in Brussels on 14 May 1910, and were demarcated
on the ground by the Anglo-German-Belgian Boundary Commission of 1911.
This is no place to attempt a detailed account of the diplomatic manoeuvres
which produced this settlement. They spread over many years and the docu-
mentation is voluminous.1 This has been cited as the classic instance of a
boundary difficulty caused by unsuitable definition-definition geometrically
clear, but adopted in almost complete ignorance of the geography of the terri-
tory.
Until the inter-departmental correspondence of the period 1906-11 is
accessible, it will not be possible to follow closely the ratiocinative processes by
which the Foreign Office came to the conclusion that it was legitimate or
expedient to launch the Kivu Mission. But some account of the current political
situation as seen from the British angle is given by Sir Charles Arden-Close
(who as Lieutenant-Colonel C. F. Close was then Chief of the Geographical
Section of the General Staff, War Office) in "A Fifty Years Retrospect" (Empire
Survey Review, 2 (1935), 135).
"In the early part of this century not much was known about Mfumbiro either at home
or in Uganda. A Foreign Office paper bears a marginal note by Lord Salisbury, 'Mfumbiro
is a myth'. In April 1906 Lieut., later Lieut.-Col. T. T. Behrens (who had reached the
Kazinga Channel with the Anglo-German Boundary Commission in 1903) wrote a
memorandum for the G.S.G.S. on 'Mount Mfumbiro in relation to the Treaties between
Great Britain, Germany and the Congo Free State' showing that Great Britain could
undoubtedly claim possession of the range. Major R. G. T. Bright (who had commanded
the British Section of the Uganda-Congo Boundary Commission 1907-08) came home in
August 1908, and he also stressed the point that we should not let possession of this
interesting region escape us by reason of not claiming it. So it appeared to the Geographi-
cal Section that the authorities should be moved to make the claim, and the Director of
Military Operations approved. The claim was thereupon made by the Foreign Office in a
communication to the Belgian Government. During the life-time of King Leopold II it
was impossible to come to any agreement on the matter or even discuss it. The Germans
also had their difficulties with the Congo authorities. I went with Captain Behrens to
Berlin in April 1909, and we were there able to agree (19 May 1909) with the Germans
about boundary matters affecting both countries in the region, Count de Salis, our Charge
d'Affaires in Berlin, taking charge of the discussions, with Herr Ebermaier, head of the
Africa department of the German Colonial Office, leading for the Germans. In the circum-
stances it seemed desirable to bring matters to a head and this was done very effectively
by sending a military force from Uganda into the disputed territory. This, perhaps
naturally, produced a howl of indignation. But we only wanted to regularize matters and
this occupation forced a settlement. Leopold II died at Laeken on 17 December 1909.
Almost with his last breath he adjured his ministers not to have any discussion with Great
Britain on the subject of Mfumbiro: but after the accession of King Albert things rapidly
straightened themselves out and it was soon arranged that a tripartite conference should
be held at Brussels to discuss Mfumbiro and other boundary questions."
The tap-root of all these misunderstandings had been planted some twenty-
five years before with a definition of boundaries shown in the map annexed to a
Convention, 8 November 1884, between the International Association of the
Congo and the German Empire. Later came a unilateral Declaration of
Neutrality, 1 August 1885, in which the newly created Independent State of the
Congo defined its eastern boundary between 40 north and 1 20' south latitude
as the 30th meridian east of Greenwich. From the point of intersection of long.
s1 Some idea of its ramifications can be found in A Manual of the Belgian Congo,
published by the Admiralty in 1919 (pp. 17-22); see also Thomas and Spencer, A History
of Uganda Land and Surveys, 1938, Chapter X.

110 UGANDA JOURNAL
300 E. with lat. 1 20' S. the boundary was a straight line to the northern
extremity of Lake Tanganyika.16
There followed on 1 July 1890 the agreement between Great Britain and
Germany which laid down that the German sphere to the west of Lake Victoria
should be bounded by parallel 1 south latitude to the frontier of the Congo
Free State "where it terminates". It was however provided that if Mount
Mufumbiro proved to lie to the south of latitude 1 S. the boundary was to be
deflected to exclude the mountain from the German sphere. Kilimanjaro was by
the same agreement included in the German sphere. Report has it that this latter
was a concession to the wishes of Kaiser Wilhelm II, since Germans had played
a major part in its exploration; and the reservation of Mufumbiro to Great
Britain was perhaps by way of recognition that an Englishman, Speke, was the
first European to catch sight of the mountain from Karagwe.
At this time Ruwenzori and Lake Edward had only just been discovered by
Stanley's Emin Pasha Relief Expedition (1889). It was supposed that the greater
part of the mountain and much of the lake lay to the east of the 30th meridian,
and for sometime (but with increasing uncertainty) they continued to be so shown
on maps. Lake Kivu17 was only a name until its position was determined by
Count von G6tzen in 1894, whereupon Germany claimed a right of access to it
as being east of the Congo State's frontier as defined by the 1884 Convention
(though west of the boundary unilaterally declared on 1 August 1885). An
agreement between Great Britain and the Congo State on 12 May 1894 rashly
continued the use of a geometrical definition of boundaries: the sphere of the
Congo State "shall be limited to the north of the German sphere in East Africa
by a frontier following the 30th meridian east of Greenwich up to its intersection
by the watershed between the Nile and the Congo and thence following this
watershed in a northerly and north-westerly direction".
Much more was becoming known of the Mufumbiro area at the turn of the
century.18 Mufumbiro which proved to be not a single peak but a whole group of
16 The basis of the German case was seemingly that the boundary shown in the map
annexed to the Agreement of 8 November 1884 lay westwards of that of the declaration
of 1 August 1885 (despite A. R. Hinks' statement that they agree-see Geographical
Journal, 58 (1921). 417). An informative map showing the debated lines, "Die neue Grenze
zwischen Belgisch Kongo u Deutsch-Ostafrika nach dem Vertrage vom 11 August 1910"
will be found in Petermanns Mitteilungen, 1911/1, Tafel 46. The accord of 11 August 1910
would be a preliminary to the triparite Anglo-German-Belgian instructions signed in
Berlin on 26 August 1910 (see Thomas and Spencer. op. cit. p. 34).
17 Speke's map (Journal of the Discovery of the Source of the Nile, 1863) shows "Mt.
Mfumbiro" with a "Rusizi L." in a position corresponding with that of Lake Kivu.
Stanley records the name L. Kivu as given to him by King Rumanika of Karagwe in
1876 (Through the Dark Continent, 1878, chapter XVII); and it may be found thereafter
in a number of maps.
18 The German explorer Dr. Richard Kandt, in the course of a prolonged stay, 1898-
1901, added much to knowledge of the features of Lake Kivu. Grogan and Sharp (From
the Cape to Cairo, 1900) traversed it from south to north early in 1899: and J. E. S. Moore
and-M. Fergusson (To the Mountains of the Moon, 1901) followed much the same route
early in the following year. On 10 April 1900, Germany and the Congo State made an
agreement for delimiting a common frontier in the neighbourhood of Lake Kivu: and a
German boundary expedition under Captain Herrmann met Belgian Commissioners in
the area before the end of that year. Dr. Lamp the Germans' geodesist died 21 June 1901,
and the commissioners seem not to have reached any mutually agreeable settlement:
certainly Germany remained unsatisfied. More recently the Duke of Mecklenburg's
Central African Expedition of 1907-08 had provided further glowing accounts of the area.

THE KIVU MISSION, 1909-10 111
mountains was found to be not only south of latitude 1 S. but west of longitude
30 E. At length the surveys of the Uganda-Congo Boundary Commission
1907-08 and of the 30th Meridian party which was at work in Western Uganda
later in 1908 provided incontrovertible proof that the 30th meridian was situated
some 11 to 12 miles east of its supposed position.19 The whole of Lake Edward,
much of southern Toro with the Katwe Salt Lake, and a greater part of
Ruwenzori were now seen to fall on the Belgian side of the meridian. One of the
cards remaining in British hands was that, by reason of the geographical position
of the Nile-Congo watershed, Belgium continued to be cut off from access to
Lake Albert, other than by way of the leased Mahagi Strip.
The British Government doubtless realized that tough bargaining would be
necessary if it was not to be excluded from vital features in western Uganda, and
it cast about to strengthen its hand where it could. Germany was at the same
time intent on obtaining recognition of its claims in the direction of Lake Kivu.
With this community of interests the "Big Brothers", by their delegates, got
together for a "Mfumbiro Conference" in Berlin: where "convinced that the
Government of the Congo State cannot but recognize the justice of these
views ... and will evacuate the territories claimed ... respectively by Germany
and Great Britain", they concluded an agreement (with a supplement) on 19
May 1909. Within a matter of days the Kivu Mission was on its way.
This agreement has not been published, for within twelve months it had lost
any significance. By it Britain acquiesced in the regulation of Germany's frontier
"from the north end of Lake Tanganyika along the River Russissi to Lake Kivu
and from there to Cape Limboga on the northern shore of that lake". From Cape
Limboga a generally north-easterly line was agreed, north and west of which
Germany conceded to Britain a free hand. It followed the summits of various of
the volcanoes-Ninagongo, Karisimbi, Visoke, Sabinio, Mgahinga to Muhavura:
thence with numerous twists, which it is profitless to attempt to identify on
modern maps, it moved generally eastwards to link up with the Anglo-German
Boundary Commission, 1902-04, boundary pillar Bl1 a few miles north-east of
the present Kakitumba Bridge at Merama.20
Today such a boundary dispute as was brought to a head by the Kivu Mission
would be submitted to the adjudication of the International Court of Justice at
the Hague. Fifty years ago the recognized method of establishing a claim was to
get there first and hoist a flag. As a tactical surprise-an unexpected move on
the chess-board-the Kivu Mission was perhaps an unqualified success.
The only account hitherto in print of any of the Mission's proceedings seems
to be in Captain E. G. D. Lardner's Soldiering and Sport in Uganda 1909-10
(London: 1912). But this has been almost valueless as a contribution to an
19 The 1907-08 Commission mapped the frontier zone northwards from latitude 1" S.
It did not demarcate a boundary: that was left for negotiation in Europe and was carried
out on the ground some years later (Thomas and Spencer. op. cit. pp. 38-9).
20 This line is described in some detail by physical features culled from the Grosser
Kolonial Atlas 1902-04 (Map No. 16 Usambara, 1: lm.); and map sheets 1, 2, 3, 4 of the
Kiwusee-Grenz Expedition (1: 100,000), ultimately joining the map of the Anglo-German
Boundary Commission, 1902-04 (T.S.G.S. 2118 (a), 1:100,000). Mr. Coote records his
view of the merits of the map with which he was supplied to illustrate these arrangements.
For some account of the proceedings of the A.G.B.C. 1903-04, see 'Extracts from Lt.-Col.
C. Delmd-Radcliffe's Diary', Uganda Journal, 11 (1947), 9.

112 UGANDA JOURNAL
understanding of events, for it contains hardly a date and not one personal name
by which participants can be identified.
It is now possible to fit Lardner's experiences into the picture. He was at
Hoima in June 1909 when he received his orders to join. He marched by way of
Mubende and the Makole crossing of the Katonga to Mbarara. The Mission's
force had already left, but he found orders to follow with rations.
The route taken by the Mission was not that of the present road to Kabale by
way of Ntungamo and Lutobo Hill. It led to the headquarters of the Chief of
Kazara (now Lwasamaire), and, bearing southwards, entered the present Kigezi
District to the north of Ihunga Hill to reach Nyalusanje. From here a track across
the hills led to Kumba. Approximately the line of the present road was then
followed to the north end of Lake Bunyonyi, Muko and the Kanaba Gap; then
across country over the eastern spurs of Muhavura and Karisimbi to Lake Kivu
at Kisenyi. Only those fortunate enough to have explored the Kigezi country on
foot will appreciate-with understanding, admiration and perhaps nostalgia-
the Mission's marching achievements. The Kivu force returned by a route still
further west, and from 26-30 June made a post on Rubona Hill, some 18 miles
south of Rutshuru. Here they hoisted the Union Jack, and had their first meeting
with the Belgians: and from here they began their withdrawal into Uganda.
Lardner met the retiring force near Lake Bunyonyi on 4 July; and he returned
with it to Mbarara. He set off for a few days shooting, only to be recalled and to
find that the Mission had already started once more for Kigezi. Again he
followed and overtook it to the west of Lake Bunyonyi, where he found it
digging in to form a standing camp. This, the original Kigezi post, lay to the
north of the present Kabale-Kisoro road not far from the present Nyakabande
rest camp and the site may be reached by a track leading off the main road.21
It remained the headquarters of the Kigezi District until a move was made to
Kumba in November 1912. The transfer of the Station to its present site at
Kabale took place in 1914. H.B.T.

21 See M. J. Bessell, Nyabingi, Uganda Journal, 6 (1938), 77, note 14.

EARLY ASCENTS OF MOUNT ELGON
By H. B. THOMAS AND R. F. J. LINDSELL

N a letter in a recent issue of the Journal (Uganda J., 19 (1955), 102), it was
pointed out that JACKSON'S SUMMIT of Mount Elgon, which, as has long been
recognized, is by no means the highest point around the crater, could hardly
be the peak climbed by Jackson and Gedge in 1890.
This conclusion is unquestionably correct. Further investigation points to
another inaccurate identification in Thomas and Dale's article on Uganda Place
Names in U.J., 17 (1953), 110. FRANZ JOSEF SPITZE is not the peak now known
as Mubiyi: much more probably the name, when introduced, was given to the
peak Koitobbos, on the eastern rim of the crater which is not in Uganda but in
Kenya.

The story of the early ascents of Mount Elgon calls for elucidation. But it may
be of interest first to trace the derivation of the name Elgon, which is no more
than an anglicized Masai name. The Masai themselves were little concerned
with the mountain beyond raiding occasionally to the lower slopes of the eastern
and northern sectors. The Bagisu however are intimately connected with it by
their traditions, which recognize it as their original home and personify our
Jackson's Summit as Masaba, the father of the tribe (whose wife by the same
traditions is Wagagai). Hence they come to refer to the whole mountain as
Masaba. Stanley records this as Marsawa which he saw from Lake Victoria in
1875, and it is unfortunate that the appropriateness of this name, so familiar to
the Bantu tribes to the west, did not become known to Europeans until the name
Elgon was fully established. Joseph Thomson, the first European to visit the
mountain, published the name Elgon in his book Through Masai Land in 1885,
and described how he first saw it from far to east, on the Uasin Gishu plains
then dominated by the Masai.
It appears that El Kony was in fact the name used by the Masai both for the
mountain, and for the scattered pastoral groups who inhabited the upper slopes
and who were at that time, owing to Masai raids, the only people to be found
on the eastern flank and the plains below. Hence they came to apply it in a
vague way to the whole mountain area. 'Elgony' was early used in English
writings as the name of these pastoral groups. They are in fact practically
indistinguishable from the Sebei, who recognize them as a group of their tribe,
and call them Kony-Jek-people of the mountain. Thomson himself was quite
confident in the use of the name, and it may be surmised that it was in common
use among the trading caravans which were already passing to and from
Karamoja along the eastern side of the mountain.
Thomson approached the mountain from Kwa Sundu (Mumias) and reached
the country of the Chief Mangichi of the Lako people some 20-30 miles east of-
the present Uganda border. On 30 December 1883 he examined "numerous
artificial caves inhabited" on the southern slopes. It was his farthest west at this

114 UGANDA JOURNAL
point of his journey. His health and resources were at a low ebb. Without delay
he turned right about and started his return to the Coast. He could record little
of the mountain but estimated its height with remarkable accuracy as 14,000 feet,
and he saw clear evidence of its "volcanic origin".
The first ascent of Elgon in February 1890 was but an incident in the journey
of the Imperial British East Africa Company's caravan under the command
of F. J. (later Sir Frederick) Jackson. The caravan had halted at Mumias as
his instructions did not permit him to go on to Buganda in the political situa-
tion as then known to him. Instead he went north to Turkana hoping to obtain
ivory to pay the Company's costs. Unable to proceed beyond the Turkwel (that
is the lower reaches of the present Suam River), the caravan returned over the
mountain to Mumias. With him were Ernest Gedge, Dr. A. D. Mackinnon and
James Martin. It is worth remembering that the caravan mustered over 500 men.
Gedge was the surveyor: he was an exceptionally neat draughtsman, and, having
spent two years on the training-ship Worcester, he had a useful knowledge of
map-making. Allowing for the limitations of the boiling-point thermometer
method, his observations for altitude merit a large measure of acceptance.
The expedition ultimately reached Mengo in April 1890. Gedge remained
until the following year, while Jackson returned to the Coast where he arrived
on 3 September 1890. His report and Gedge's maps must have been sent at once
to the Company's Directors in London. Jackson himself did not proceed to
England, being posted elsewhere on the Coast. The Directors were preoccupied
with financial and political problems-Lugard was already on his way to
Uganda. They thought the expedition had cost a lot of money with no great
results, and showed little appreciation of the geographical aspects of what was a
very fine performance. For not only did the journey down the Turkwel River
link up Mount Elgon and Uganda to Ngaboto where Count Teleki, returning
from his discovery of Lake Rudolf, had arrived in June 1888; but Jackson's
was the very first British caravan to reach Mengo by the Masai route, and his
conciliatory firmness achieved this result with a notable absence of that fright-
fulness which marked the progress of so much early exploration.
No copy of Jackson's report is so far as is known extant. It is a sad fact that,
since the liquidation of the Chartered Company in 1895, all records of the Court
of Directors, which must have contained the reports of dozens of pioneer
journeys in the East African interior, have disappeared. Gedge's maps, how-
ever, soon came into the hands of the Intelligence Division of the War Office,
who drew a map "Imperial British East Africa Company's Uganda Caravan
1889-90 under command of F. J. Jackson. Sketch of Route by Ernest Gedge,
F.R.G.S. November 1890".1 On this every camp site is numbered and plotted.
A copy is in the Royal Geographical Society's Library, but it is absent from the
British Museum's map catalogues. There is no explanation of the fact that no
use of this route across Mount Elgon was made by Lieut.-Col. Macdonald when
compiling his Map of Uganda, 1899-1900,2 a most comprehensive work incor-
porating almost every available authority. Much subsequent misunderstanding
may perhaps be traced to this omission.
I Scale. 6-41 miles to one inch. I.D.W.O. No. 814, in two sheets.
2 Scale, 10 miles to one inch. T.S.G.S. No. 1429, in four sheets.

FIG. 1

Mount Elgon, interior of crater: looking east and north-east from the approximate site of Jackson and Gedge's camp in 1890. Their route
into the crater was probably over the saddle to the left. The entrance to the gorge of the Suam River can be seen to the right. Mubiyi is
off the photograph to the left.

'ID PY nV f na Ll I:'[II

EARLY ASCENTS OF MOUNT ELGON 115
On 9 February 1891, E. G. Ravenstein read a paper to the Royal Geographical
Society which he had prepared from Jackson's and Gedge's reports and maps.3
The ascent of Mount Elgon occupies only a few paragraphs. Jackson's report is
likely to have been as much political as geographical: that a copy is not among
the records of the R.G.S. is therefore understandable.
Ravenstein could get no help from either of the explorers, for both were
absent from England. This may explain an error which has crept into his text.
Gedge recorded and mapped the altitude of the highest point on the crater rim
climbed by him and Jackson as 14,094 feet: Jackson's report seems to have
added that it was estimated that none of the other peaks was more than 50 feet
higher. Ravenstein, however, prints in his text 14,044 feet as the observed
height,4 and this is repeated on his map where he inserts nearby to the east an
estimated maximum height of 14,094 feet.5
Jackson had a life-long detestation for paperassie, excepting only when he was
recording the bird-life of Africa. He was at the same time so indifferent to any
public recognition, that he made no effort to give a more detailed account of his
and Gedge's exploits. So, for over twenty years, Ravenstein's was the only
available description of the summit of Mount Elgon and its discovery: and it is
quite remarkable that virtually no reflection of Jackson and Gedge's route-
march through the crater of Elgon ever found its way onto the maps of Uganda.
Reluctantly, in retirement, Sir Frederick set about writing the story of his life in
East Africa. The result is his typically generous and urbane Early Days in East
Africa, 1930. But this lacked his final guiding hand, for it was still unfinished at
his death, and had to be seen through the press by Lady Jackson. A careful
examination reveals some confusion of sequence. His printed recollections of
the Elgon trip were thus separated from the event by nearly forty years. Such
papers as he had to refer to have not been preserved.
The disappearance of Jackson's original report or any diary he may have kept
of the 1889-90 journey to Uganda has meant that there was no contemporary
record by which to verify subsequent accounts. Providentially this gap has now
been bridged. Ernest Gedge's son, Mr. Cuthbert Gedge of H.M. Foreign Service,
has recently been so good as to afford access to his father's original journals
which contain a day-by-day account of the whole of his stay in East Africa from
1888 to 1891.6 This is a record of considerable historical importance: and since
Gedge and Jackson maintained throughout an almost affectionate relationship,
any divergence of view is unlikely. Gedge's contemporary account of the ascent

3 This paper, with a map on a reduced scale (1: 1,000,000 or 16 miles to one inch)
compiled by Ravenstein, is printed in Proceedings R.G.S. New Series 13 (1891): see also
for text only U.J. 12 (1948), 129.
4 Ravenstein, P.R.G.S. ibid. p. 203.
5 The I.D.W.O. map correctly follows Gedge's journal and shows only one observed
height, 14,094 feet. In thus locating the higher peak to the east of Gedge's observation
point, Ravenstein fell into error. Jackson (Early Days, p. 245) could well be referring to
Wagagai, some two miles to the north-west when he remarks that "none of the other peaks
was more than 50 feet higher". Jackson himself adds to confusion by stating ibidd. p. 244)
that Gedge's height was Thomson's 14,000 plus 192 feet, evidently the source of the figure
quoted by Sir Albert Cook (p. 126, post).
6 Also preserved are Gedge's journals of his visit to Uganda in 1892-3 as correspondent
of The Times, and many contemporary letters.

116 UGANDA JOURNAL
to the summit of Mount Elgon is clearly the most authoritative record extant.
His journals at this period leave the impression that, despite circumstances
which would dismay most modern travellers, Jackson and he at all times had
the situation 'under control'. There is no complaining but rather an unquestion-
ing confidence in the outcome of their journey, and among all its vicissitudes
they found time to observe and record the natural history of these unknown
regions and to add to their collection of bird-skins.
The journal shows that on 4 February 1890 the caravan arrived at Save
(Sebei), the present Kapchorwa, headquarters of Sebei County, and camped on
a hill above the village (Camp xcviii) with an altitude of 6,364 feet.7 The Masai
and Nandi had raided recently. "I wish they would come while we are here.
They'd catch it hot." In fact on the 9th a party of about 1,000 Nandi made a
tip-and-run raid on the next valley, but the caravan's presence saved the Save
people from any follow-up. On 11 February some people came in from Elgon
"with a report that a large caravan is at Kwa Sundu with three Europeans".
This was remarkably quick news of the arrival of Carl Peters who had reached
there on 2 February. But Jackson had no suspicion that this was a German
expedition with political objectives and assumed that it was some hunting
party.
Jackson in his Early Days, p. 241, says that they had intended to return to
Kwa Sundu by the known route round the east of the mountain but their guides
led them for some hours upwards before the mistake was discovered; and they
then decided to take the risk and go on. But Gedge's diary makes it clear they
started out with the definite intention of crossing the mountain.
Having managed to get an old man to act as guide they struck camp on the
13th. "Whether we shall ever succeed in getting to the summit is very doubtful,
but we mean to have a shot for it."
They set off in a south-easterly direction, first up a long valley and then,
ascending a long well-wooded spur, camped in the forest by a small stream-a
short march (Camp xcix, altitude 7,881 feet). "The gorges and slopes of the
mountain are a mass of forest, with crags jutting out boldly among the trees,
and here and there patches of grass and bamboos with bracken and grey-beard
moss." Next morning (14th) they broke camp at 6.15 a.m. and pounded away up
the mountain through the forest. Getting higher and higher they had a grand
view. "Finally we struck on to the track of the Wa-Nandi raiders and eventually
after messing about, and getting more or less lost, we descended a gorge and
camped again in forest. This camp (c) is at an elevation of 9,528 feet by B.P....
We go still higher tomorrow." The night turned very cold but "not like Kikuyu
or Mau, for it is ever so much drier here".
Next day (15th) still bearing generally south-east they continued up-hill
through "the same sort of going as yesterday, patches of forest and scrub, water-
courses, etc. Left the forest towards the end of the march and emerged into the
open coarse tufty grass and evergreen bushes. A magnificent view was
obtained from one place. The plain below appeared like a vast sea stretching

7 6,364 feet is the height in Gedge's journal. Both I.D.W.O. and Ravenstein's maps
print 6,346 feet.

EARLY ASCENTS OF MOUNT ELGON

From

1 o ) 2 3 4
Scole I I I I I a miles

1m-u4t4 w Approximate position of crater rim
--------- Present tracks
..*... Jackson and Oedges probable route
S "* ,, campsite

[Drawn by G. Webster, District Forest Officer, Busgsu
FIG. 2
Crater Area of Mount Elgon.

118 UGANDA JOURNAL
away as far as the eyes could see broken by the isolated mountains of Lekakisera8
and Debasien, with one other further away north standing out like islands . .
To the S. by W. rose the crater of Elgon with a wreath of cloud about it... We
camped by some curious trees, like dragon palms, and seen for the first time
today. Found altitude of camp (ci) to be 11,110 feet."
16 February. They now turned westwards. "Got under way again at 6.20 a.m.
taking a S.W. course ... Descended first into a gorge and crossed a small torrent
and then reascended. As we got higher the cold increased ... Dragon palms but
no other trees plentiful; coarse grass, white heather, and everlasting flowers,
another plant growing much like thyme, some small marigolds and various
other mosses and lichens made up the vegetation of the high slopes ... After a
long stiff pull we approached the rim of the N. crater and topping it went down
inside and camped. In the N. side of the rim there is a deep gorge which gives
passage to a stream rising from the interior of the crater. The sides slope
gradually down from the broken rim at a depth of about 800 feet. The interior
is covered with coarse grass, and big boulders are scattered about. The diameter
-will be about I to 2 of a mile and the circumference about 3 to 31 miles. The S.
side rises considerably higher than the N. The swamp from which the stream
rises is in the centre and is of a peaty character. Whether this is the N. crater
proper remains to be seen. This can only be done from higher up on the summit.
The winds are excessively cold up here. How the porters can stand it beats me...
By B.P. the altitude here (Camp cii) is 12,802 feet. It is exceedingly interesting
that a caravan should have gone so high, as they cannot stand 10,000 feet on
Kilimanjaro ... fortunately the dead palms give plenty of good fuel. We have to
go still higher tomorrow and now cannot be far short of the actual summit. As I
write this my nose is frozen, so are my fingers. This is truly 'English' . God
keep the rain away, a storm would be certain death to half our men ... Could
get no sights as a thick haze overhung the plains and the magnetic attraction is
tremendous varying the compass from 15" to 20 ... All well."9
17 February. "At 6 p.m.... 3 of frost. Ice in the buckets, etc. Left camp
about 7.15 a.m. as soon as sun got into the hollow to warm us up a bit. On
topping the edge of the small crater we found ourselves on the edge of a vast
crater of almost circular form about 8 miles in diameter. The slope was gradual
down to the bottom, principally grass and various heaths and mosses, with
dragon palms. At the bottom, which is from 1,500 to 1,800 feet in depth, is a
swamp the source of the R. Angalul which makes its exit on the E. side through
a stupendous gorge. The serrated edge of the crater though broken in places
stood out well against the sky. There is then evidently only one crater to the

8 Lekakisera and Lemboto appear on Thomson's map of 1884, a copy of which Gedge
had with him. The actual copy with Gedge's annotations is still among his papers. But
Thomson (who must have got the names from Swahili traders for he probably never caught
sight of them), puts "Lekakisera (snow-clad)" roughly where is today Nepak or Kamalinga,
while "Lemboto" is in the position of the present Debasien. Gedge adds to the confusion
by using "Debasien" for the more distant Nepak, and "Lekakisera" for the nearer Debasien.
9 Jackson, Early Days, p. 243 relates that during this day a party searching for
straggling porters came upon one who, having coiled himself up knees to chin, was so stiff
from the cold that an askari hoisted him on his shoulder feet upwards and so carried him
to camp amid general merriment. Two others could not be found: they had evidently
crawled into the scrub and doubtless died from the cold.

EARLY ASCENTS OF MOUNT ELGON 119
mountain. Climbed with J. up to the top on the crater edge on the N. side,
giving us an altitude of 13,602 feet. The S. side has a large flat crag jutting up
which is about 700 feet higher than this one which we stood on. This will give
the mountain an altitude of over 14,000 feet. Took photos of various aspects the
crater presented, and then continued our march keeping along the W. side. Met
a party of Elgon warriors in their war paint. They had seen the smoke from our
fires and thought we were Save people coming to attack them ... One donkey
left behind to die. Got to camp on S.W. side of crater at 12 noon. Exposed
situation. This is truly a mighty mountain and moves one by its grandeur.
Stayed in camp and did up maps, etc. all afternoon." (Camp ciii, 13,328
feet.)
18 February. "This morning everything white with a rime frost ... Made a
start as soon as the sun showed up over the edge of the crater. The road was
good round the side in a S.E. direction. The caravan halted on reaching the top
and I photographed them before we commenced the descent. J. and I went up
the highest crag on the edge and there we were virtually on the summit of mighty
Elgon. Boiled the thermometer, altitude 14,094 feet. View obscured by thick
haze. Descent over the same kind of country as on the other side we ascended
and after a sharp walk of 5 miles reached camp (civ) on a mountain stream at
12 noon at an elevation of 11,044 feet ... Doing up map, etc. in afternoon. Am
very fit and well. J's birthday yesterday which was drunk in champagne, and
may he live to see many more. Queer place to spend one in."
19 February. "Continued our journey down the mountain. The slope is more
gradual on the N. side and there is none of that disagreeable scrub to go through,
the forest line extending higher up than on the N .... Entered the forest and
descended a long steep gorge with a stream of good water at the bottom and
camped." (Camp cv, altitude 8,218 feet.)
From here, examining on the way the inhabited caves, some of which would
probably be those seen by Thomson in December 1883, they passed on to reach
Kwa Sundu on 4 March 1890.
A study of Gedge's photographs-or rather of the inconsequent titles which
have been attached to them particularly in Sir Harry Johnston's The Uganda
Protectorate, 1902-explains some of the misunderstandings regarding the
highest peaks around the crater which have become current. Four only of the
top of Mount Elgon, all of which are reproduced by Johnston, are extant. A set
of Gedge's own prints, titled by himself and used by Jackson's publishers
includes no others.
On p. 58, Johnston, ibid. is "Edge of crater wall, Elgon, from the East", while
Gedge's own title is "Upper slopes of Mount Elgon looking North approaching
rim of crater". It is reasonably certain from the description that this is a view of
the outer north-facing slopes of the main crater taken doubtless soon after start-
ing on the morning of 17 February. Gedge's wording "looking north" is confus-
ing as, of course, the photographer must have been looking south-westerly and
this seems to be confirmed by the shadows which are those to be expected at that
time of day and year.
On p. 59, Johnston, ibid. "The highest point of the crater rim (14,200 feet)" is
more than misleading: it is gravely inaccurate. For Gedge records this as "View

120 UGANDA JOURNAL
within the northern rim of crater Mount Elgon".'1 This picture seems to show
the same rock that appears in the previous picture; and is probably on the very
lip of the crater somewhere on the northern side. It appears to be looking towards
the east and the sweep of hill seen in the background would be the eastern floor
of the crater dropping down to the bed of the Suam River. At much the same
time "The descent to the central crater of Elgon" photograph at p. 61 ibid. was
doubtless taken.
"The Jackson-Gedge Expedition at the top of Mount Elgon" at Johnston,
p. 60 (Jackson, Early Days, p. 230), taken on 18 February calls for minute
inspection. Three Europeans and about a dozen Africans, probably headmen,
with a flag are grouped on top of a rocky hummock which cannot be more than
forty or fifty feet high. The lower slopes are crowded with squatting porters
(who seem to have little interest in the proceedings) while to the right
lare at least two donkeys. This is almost certainly taken at the point on the
southern crater rim where the caravan crossed it, and behind is possibly the
peak which Jackson and Gedge quickly climbed. Anyone who has experienced
the searing northerly wind which prevails in the January-March dry-season will
realize from the position of the flag that the photograph is looking eastward,
hence to the left of the track, and this accords with Gedge's description of the
summit they climbed. Jackson (p. 245) is explicit: "We wished to climb to the
highest peak on the south side (sc. of the crater), and only a short way to the left
of the track, to ascertain its height, while the whole safari, after reaching the
rim, was to collect together in order to be photographed. The wind on the top of
that.peak was rather violent, and gave Gedge much trouble with the boiling-
point thermometer. The height, when worked out, gave 14,094 feet, and we
estimated that none of the other peaks was more than 50 feet higher.""
From this evidence there is little difficulty in retracing with reasonable
precision Jackson and Gedge's route across the top of Mount Elgon. They
approached the northern rim from the north-eastern slopes. On 16 February
they camped in a depression which they mistook for the true crater. This so-
called 'north crater' has not been identified with any certainty. The description
does not fit what is now the usual approach from the north, crossing the saddle
immediately to the east of Mubiyi, but an examination of air photographs does
suggest that the next valley to the east is surrounded by fairly steep cliffs except
for the gap in the north where a tributary of the Siti flows out, and this might
well have appeared to be a crater.
Next morning (17th) the caravan clambered up the south side of this depres-
sion and found that at last they were on the northern rim of the main crater.
Jackson and Gedge went up the nearest peak: while the caravan keeping to the
western side of the crater descended in a south-westerly direction and "after
traversing rather more than three parts of the basin" (Early Days, p. 244) camped
on the south-western inner slopes.
A well worn path still runs around the inner western slope of the crater, and

10 Jackson, Early Days, p. 242 gives "Crater of Mount Elgon" which is merely a
publisher's contraction.
11 Gedge's journal does not mention that "none of the other peaks was more than 50
feet higher".

EARLY ASCENTS OF MOUNT ELGON 121
it may be assumed that their route joined the track inside the crater, and
followed it next day into the southern sector. Here it rises over the rim within
what is now the Kenya side near a peak called Kiongo, which seems to be the
second highest on the mountain. It is likely therefore, that this was the peak
actually climbed on the 18th when its altitude was determined as 14,094 feet.
This is calculated today as 14,115, being thus 63 feet lower than the 14,178 feet
now accepted as the height of Wagagai. It is not likely that they went near
Wagagai which lies rather set back at the extreme south-western edge of the
crater and because of this fact and various perspective illusions, does not stand
out as noticeably the highest peak. Jackson and Gedge apparently climbed from
the track to this peak that lay on their left and completed their observations as
quickly as possible, thereupon rejoining the caravan waiting to be photographed
near the track. The course of their descent would continue down the ridge east
of the Lwakaka valley (the present Kenya-Uganda border) and from this they
deviated eastwards in order to visit the inhabited caves.
From all this emerges the indisputable fact that they did not go near, and
probably did not notice Jackson's Summit which, lying a mile or so to the west
of the crater rim and at about the same average height, would only just be visible
from Kiongo peak.

For long after 1890 no one seems to have gone near the summit of Elgon.
C. W. Hobley, later Provincial Commissioner, Kenya, made a complete circuit
of the mountain in January 1896.12 He rounded the Namisindwa Bluff to Bupoto,
crossed the pass to the east of Nkokonjeru and over the Siroko plain to Mbai
and Sebei.13 His map shows that he was aware of the route followed by Jackson
and Gedge across the crater.
The Macdonald Expedition of 1898 maintained a base camp at Save (Sebei)
under the command of Lieut. Hanbury-Tracy: but no member of the expedition
climbed to the upper levels.14
Early in 1901 Sir Harry Johnston, homeward bound but anxious to visit the
north-east of the Protectorate, reached the south end of Mount Elgon. Here he
saw the cave discovered by Joseph Thomson: it was masked by "a splendid
waterfall", which, he says, was not on his map (sc. Macdonald's). "It is the
descent of the Sasuru River, which I call Thomson Falls.""
No itinerary or details of Johnston's circuit of Elgon seem to be extant. But
he followed a road recently made by Kakunguru round the western foot-hills
and reached Sebei. "From the Sebei country I was obliged to travel for 16 days
without road or guide to Ravine Station";16 and on the way he seems to have
12 C. W. Hobley. Notes on a journey round Mount Masawa or Elgon. Geographical J. 9
(1897), 178.
13 To Hobley is due the naming of JACKSON'S FALLS (U.J. 17 (1953), 114). Having
reached the Simu River he records "Several waterfalls of great beauty were discovered:
the first, which has a sheer drop of 60 feet, I have taken the liberty to name after F. J.
Jackson, the first European to ascend the mountain" (G.J. loc. cit. p. 182).
14 H. H. Austin, With Macdonald in Uganda. 1903.
15 H. H. Johnston, The Uganda Protectorate. Geog. J. 19 (1902), 14 ff. The Thomson
Falls are illustrated in Johnston, The Uganda Protectorate, i, 49.
16 H. H. Johnston, Geog. J. ibid.

122 UGANDA JOURNAL
visited Lake Baringo.17 It was doubtless on this occasion that he found F. J.
Jackson and James Martin awaiting him at the Ravine.'8
Sir Harry's brother Alex. Johnston in his Life and Letters of Sir Harry
Johnston, 1929, p. 216 remarks, "After crossing Elgon in a snowstorm our
expedition descended" to the Uasin Gishu Plateau. In a recent letter Mr. Alex.
Johnston confirms, however, that none of the expedition ascended the mountain.
At one point on the north slopes they skirted the bamboo forest, and here they
experienced a fall of snow which did not lie.
Mbale was opened as a district station in 1904, and the first attempts to bring
Bugisu under administration followed. At the end of 1907 a small French party,
Maxime de Bary and Jean Lefebvre arrived in East Africa. In December they
called on Jackson (tr&s aimable) then Lieutenant-Governor in Nairobi, who
told them that his expedition of 1890 was the only one to succeed in making an
ascent to the summit of Elgon. By way of Kisumu and Mumias, they reached
Mbale in February 1908, and there met S. Ormsby, the Collector, and his
assistant J. M. Coote. Ormsby thought the ascent would be very difficult, and
advised them to attempt it from the south side and to descend at the north-east
of the mountain. From this it may be inferred that at this date no European,
since 1890, was known to have reached the crater. De Bary's main safari was
directed to circle the south-eastern foothills, while he and his friend with four
Somali hunters and six porters set off. On 1 March they visited the southern
caves and may then have followed upwards much the line by which Jackson
descended in 1890. On 5 March they found "une breche permittant l'entr6e du
cratere". They certainly entered the crater (two good photographs are repro-
duced in de Bary's book), but visited only the eastern sector and did not perhaps
camp within. On 6 March they left the summit, following a valley (probably the
Suam) which led straight to the Uasin Gishu, and rejoined their main safari on
13 March. They left Mombasa on 27 April 1908.19
In November-December 1909 an American party, consisting of Carl Akeley
and his wife, J. T. McCutcheon and F. M. Stevenson, made its way from the
Uasin Gishu (where they encountered Theodore Roosevelt and his hunting
party), up the southern face of the mountain, visiting the cave dwellings and
perhaps following much the same route as de Bary, one of whose guides joined
their party. They climbed "the highest peak" on the rim where their aneroid
registered 14,375 feet, and camped for one night in the crater. They returned
over the eastern rim and through the forests to the Nzoia river. McCutcheon
was the cartoonist of the Chicago Tribune. His book, In Africa. Hunting Adven-
tures in Big Game Country (Indianapolis: The Bobbs-Merrill Company, 1910),
into which he collected his newspaper articles, is enlivened by his drawings: but
this, which seems to be the only record of this ascent, is primarily 'bright'
journalism and gives little information.
It was during these years that government officers began to find their way to
the higher levels of the mountain, hitherto unadministered and unknown.
Shortly after de Bary's visit, say about April-May 1908, the first ascent from the
17 H. H. Johnston, The Story of My Life. 1923. pp. 377-8.
18 F. J. Jackson, Early Days, p. 72.
19 Maxime de Bary, Grand Gibier et Terres Inconnues. Paris: 1910.

EARLY ASCENTS OF MOUNT ELGON 123
Uganda side was made by Ormsby and Coote starting from Bulucheke. They
camped at about 12,000 feet where Ormsby, feeling the altitude, had to give up
while Coote continued to the top. For the first time they learnt that the bamboo
zone was inhabited, as it is today, by small groups of pastoral Sebei, known
variously as Elgonyi, Nandi, Wanderobo, Basibolo or Lako.20 Later in the same
year Ormsby, coming up by the northern slopes from Kelim, crossed the summit:
for Kmunke (see infra) narrates that in December 1911 his party found near
Franz Josef-Spitze and within the crater, an empty whisky bottle containing
Ormsby's card. He may well have descended past Jackson's Summit by way of
the Butandiga Ridge, and have then discovered the ORMSBY FALLS.21 Ormsby
died at Mumias on 21 January 1909 from an attack of blackwater fever brought
on, it was thought, by the strain of this experience (teste Dr. R. Stigler).
In April 1911 (5th-6th) Coote with Captain A. E. Newland, R.A., of the
4th K.A.R., made what may be the, first ascent from the west. They went up
from Buginyanya. It was an entirely unpremeditated excursion, while on a
patrol of the lower slopes. They spent one miserable night below the summit.
Next morning they scrambled to the crater rim to have a view for half an hour
across the crater, and returned at once to Buginyanya.22
Sir Frederick Jackson who had been serving for some years in the East Africa
Protectorate returned to Uganda as Governor in April 1911. He had not been
near Mount Elgon since 1890. In July-August 1911 he was on tour through
Mbale, and crossed Lake Salisbury to visit the site of the future Soroti station.
He must then for the first time have seen the western aspect of Mount Elgon, so
magnificently photographed by A. D. Combe in the frontispiece to Dr. K. A.
Davies' The Building of Mount Elgon (Entebbe: 1952). On this the Mubiyi peak,
the gorge south of it (which is the source of the River Simu), Jackson's Summit
and parts of Wagagai, are all distinguishable on the sky line. As the crater
rim extends for some 8 miles from north to south and peaks of approximately
equal height are dotted along it, it is not unnatural that perspective gives very
different impressions from different angles; in fact, there are few places from
which Wagagai appears to be the highest point, or, indeed, is even visible,
whereas, from close on the western side, Jackson's Summit, standing high and
sheer as a knob above the upper Siroko valley, appears to be the highest point.
Three months later, in November 1911, there passed through Entebbe an
Austrian party under Rudolf Kmunke. Jackson received them with his customary
kindness and, discussing their project of climbing Mount Elgon, he may have
given them to understand that the 'knob' which he had so lately seen, had not
been climbed by him and Gedge in 1890. Kmunke, by profession an architect,
had with him three companions, Dr. Robert Stigler, R. Storch, naturalist, and
H. Schwarzer, photographer. The party crossed Busoga and the Terinyi ferry to
Mbale. P. W. Perryman, then District Commissioner, Mbale, accompanied them
for the first few stages of their ascent of Elgon. They were also joined by Dr. L.
Sells, the Government Medical Officer at Mbale, who continued with them to
the crater, where however he was incapacitated by mountain sickness. Since
20 Recent communication by Mr. J. M. Coote.
21 U.J. 17 (1953), 117.
22 Mr. Coote's MS. journal.

124 UGANDA JOURNAL
they passed through Buyobo and Buhugu they doubtless followed the Butandiga
Ridge and were thus perhaps the first party to ascend by this route. Leaving
Jackson's 'knob' to their south they entered the crater by the gorge of the Simu
River, and camped within for three nights, 13-16 December 1911. From there
camp excursions were made to climb Franz Josef-Spitze and Jackson-Spitze.
The selection of these particular peaks for climbing and naming, seems to have
had no more significance than that they were prominent and were thought not
hitherto to have been climbed. Kmunke makes no claim that they are the highest
points on the mountain, though he regarded Franz Josef-Spitze as the highest
virgin peak, and he records higher altitudes on his map. The honour of naming
new peaks seems to have been uppermost in his mind. None of his party visited
the southern rim of the crater. On his return to Mbale Kmunke telegraphed for
and obtained permission from Emperor Franz Josef and from Governor Jackson
to affix their names to these peaks.23 This was a common enough practice at this
period, though it was coming to be frowned on in geographical circles. Jackson
certainly is unlikely to have given the matter a second thought. Nevertheless
these two names had become common currency in Uganda before World
War I.
From Mbale Kmunke and his party followed the normal route by way of
Kumi and Lake Salisbury to northern Teso, whence after exploring Akisim and
the Nyakwai Hills they struck directly across country through Adilang to the
Uganda station at Nimule where they arrived in mid-February 1912 just as the
Lamogi Rebellion was coming to an end.24 They passed through territory which,
though unadministered was not really unknown, for the Northern Patrol had for
some time been active in Acholiland and northern Teso. In fact Governor
Jackson provided them with an escort of 15 askaris under a Sudanese sergeant.
On his return to Europe Kmunke published, in lavish format, his book Quer
durch Uganda. At the same time he claimed for his journey a significance and
a pioneer character which could hardly be justified. He thus opened himself to
attack in the Vienna press, where he was accused as a boastful imposter repre-
senting a holiday trip as major scientific exploration: particularly was he taken
to task for having bestowed names on two of the peaks of Mount Elgon.
It would not now be profitable, even were it possible, to probe the personal
animosities which nourished this debate. Kmunke had an influential backing
and had secured favourable notice of his book from the venerable explorer and
geographer Dr. Oscar Lenz in the proceedings of the Vienna Geographical
Society. The attack was led by Rudolf Grauer, a zoologist and professional
collector of wild animals, who had been in Uganda some years before. In
January 1906 he had climbed in Ruwenzori-where his name is perpetuated in
the Grauer Rock at the head of the Mubuku Valley:25 and he was thus able to
recognize the weaknesses of Kmunke's assertions. Dr. Stigler, the physician of
the expedition, supported his chief with advice to be cautious in his claims. But
Kmunke had been on bad terms with the other two members of his party who
provided ammunition for the attack. The Royal Geographical Society possesses
23 R. Kmunke. Quer durch Uganda. Berlin: Dietrich Riemer, 1913, p. 44.
24 U.J. 18 (1954), 166.
25 U.J. 17 (1953), 112.

EARLY ASCENTS OF MOUNT ELGON 125
a copy of a 25-page pamphlet26 by Rudolf Grauer, couched in uncompromising
terms, in which is the disclosure by Storch that Kmunke, who was no longer a
young man, climbed neither of his peaks; Stigler and Schwartzer having gone up
Franz Josef-Spitze: and Stigler and he (Storch) Jackson-Spitze. A long letter
from Gedge to Grauer (3 March 1914) is also printed in which Gedge sets out the
salient facts of the 1890 ascent.
Less than two months after Grauer's 'Opinions' had reached the R.G.S.,
World War I had broken out and communication with Vienna ceased. Kmunke's
book received a review by Sir Harry Johnston in Geographical Journal, 44
(November 1914)-not unfriendly but 'judgement reserved': and Johnston
approved of the name 'Jackson' for one of the peaks. But virtually nothing was
available beyond the meagre information of Ravenstein's paper of 1891 by
which to weigh Kmunke's claims, and they received no further attention.
Kmunke died in Vienna in 1918, but the name Jackson's Peak or Summit
remained, and came unthinkingly to be equated with the highest peak climbed
by Jackson and Gedge in 1890.
It must be confessed that a close study of Kmunke's book does little to
establish confidence in his record. It would be wearisome to traverse critically
the details of his account. He seems to have confused Jackson with Sir Harry
Johnston. Jackson and Gedge's achievements are almost completely ignored,
and instead Kmunke makes grandiose claims to hazardous exploits.2
More careful scrutiny can however be directed to Kmunke's somewhat pre-
tentious map of the crater. At first glance this seems to bear little resemblance
to the results of later surveys: but it becomes comprehensible when it is
realized that it is plotted not less than 45 out of true bearing, and that its north
point should be placed somewhere between north-east and east. It is referred to
as a photogrammetric map-a number of photographs and some theodolite
measurements were taken; but the methods of its compilation are not stated
and all its altitudes are clearly too high.28
This wrong orientation of Kmunke's map may carry some of the responsi-
bility for the incorrect identification of Franz Josef-Spitze with Mubiyi which
became current in some quarters in Uganda: for Franz Josef-Spitze appears
thereon to lie only a little east of north from Jackson-Spitze, in the position
actually occupied by Mubiyi. When Kmunke's map is correctly orientated

26 Urteile iiber Rudolf Kmunke's "Quer durch Uganda". Wien: 1914.
27 One statement in Kmunke's book is worth straightening out. On p. 43 he remarks
that five expeditions had reached the Elgon crater before his: by the English Colonel
Macdonald (who never climbed the mountain); by the present Governor of Uganda Harry
Jackson (referred to elsewhere as "the distinguished scholar", a description which would
have amused Sir Frederick); by Captains Ormsby, Cook and Newland (Cook is unquestion-
ably J. M. Coote); and finally by a French party (sc. de Bary).
It may be added that Kapitan Taughner (Kmunke, p. 59) is Captain H. M. Tufnell,
4th K.A.R., who had lately transferred to the Provincial Administration.
28 There is some indication of its shortcomings in the fact that editions exist of the
same map on which the scale appears as 1:40,000 (in Kmunke's book, and accompanying
an article by Kmunke in Petermanns Mitteilungen, 52/2 (1913), 75-6), and as 1:35,000 (in
R.G.S. map collection); while the two 1:40,000 scale maps have differing scales of metres.
There are similar discrepancies in Kmunke's route-map. Both 1:760,000 and 1:750,000
are printed on differing editions. Probably Petermann is the best guide to select: but the
crater map is clearly of questionable value to the geographer.

126 UGANDA JOURNAL
Franz Josef-Spitze may be equated.with Koitobbos on the far eastern rim of the
crater and in Kenya.29

From 1912 onwards the Butandiga Ridge became the normal route to the
summit. None of the officers stationed at Mbale at this period had any special
zest for mountaineering. They climbed in the course of their duties, and it was
normally accepted that honour was satisfied and the due reward earned by
attaining the knob, Jackson's Summit, as a convenient goal.
The record of one other early ascent which has contributed to confusion calls
for study. As this links up with the experience of one of the writers of this
article (H. B. Thomas) he continues in the first person.
In his Uganda Memories, 1945, pp. 299-301, Sir Albert Cook recounts his
ascent of Elgon in January 1914. The party included also his wife, Archdeacon
T. R. Buckley and the Rev. H. B. Ladbury. Governor Jackson had lent a map
"compiled by three Austrian explorers who had recently ascended the moun-
tain", which suggests that a copy of Kmunke's book had already reached
Uganda. Starting from Buwalasi they struck up the Butandiga ridge, camped in
the bamboo forest, and on the following day reached in open moorland what
Cook names 'Summit Camp'. From here next morning they "commenced the
final ascent... At length we reached the rim of the crater and saw that one of the
peaks that rise from it was covered with fresh fallen snow". Baffled by the wind
in an effort to boil tea on a Primus stove they "set about the climb up the final
peak or hump". Buckley was nearly exhausted but managed to follow. "Seated
on the cairn of piled up stones ... we let our gaze wander round the horizon...
The opposite rim of the crater shut out something of the view to the north-east,
but west and north and south, the panorama was superb ... A search among
the stones of the cairn brought to light a broad-mouthed pickle bottle, from
which we drew out a sheet of paper on which were inscribed nine names. The
first arrival at the top, Sir Frederick Jackson with his companions Martin and
Gedge, were alas not on it... Gedge found the altitude of 'Jackson's Summit'
by his boiling point thermometer to be 14,192 feet only eight feet less than
the accepted height today30 ... Among the nine recorded names was one of a
woman, who had gone up shortly before us ... on mule back; she had evidently
lacked a pencil, but with true feminine ingenuity had pricked her name with a
29 In a recent letter (1954), Dr. Stigler, writing from Going b. St. Johann, Tirol, confirms
explicitly that, in order to reach Franz Josef-Spitze from Kmunke's camp (which was on
the inner western slope of the crater probably not far from Jackson and Gedge's camping
place) it was necessary to cross the whole crater, as is indeed indicated on Kmunke's map.
Kmunke records Franz Josef-Spitze as having an altitude of 4,382 metres (14,377 feet)
and Jackson-Spitze as 4,311 metres (14,144 feet). He also shows a maximum height on the
unvisited south side of the rim of 4,486 metres (14,718 feet).
Two important papers by Dr. Stigler are: Ethnographische und anthropologische Mit-
teilungen fiber einge wenig bekannte Volkssttimme Ugandas (Mitteilungen der Anthro-
pologischen Gesellschaft in Wien, 52 (1922), 197-261, and 53 (1923), 113-189, 5 plates and
map): and Ein Beitrag zur Geschichte der Erforschung Ugandas (Mitteilungen der
Geographischen Gesellschaft, Wien, 94 (1952), 230-42). These deal primarily with his
scientific work during the Kmunke expedition, but they merit consultation as providing a
succinct and balanced account of the achievements of the expedition.
30 14,200 feet is the height of the top of Mount Elgon given on Macdonald's map and
by Sir Harry Johnston.

EARLY ASCENTS OF MOUNT ELGON 127
pin! ... We remained ... until the spectacle of great clouds forming below and
threatening to leave us isolated in dense mist, warned us thence. We ... passing
our last camp, reached the bamboo forest before nightfall."
Cook prints two photographs, at p. 174 is "Near the Summit of Elgon, 1914"
which is the quite unmistakable rocky ridge lying to the west of Jackson's
Summit: there is no claim that this is the peak his party climbed. "On the
Summit of Elgon, 1914" at p. 300 shows the party grouped on the cairn of
.stones which they had found on the 'peak or hump' which they had climbed.
Just two years later, in January 1916, I followed closely in Sir Albert's foot-
steps. My companion was the Rev. W. S. Syson of the C.M.S. Ngora Mission.
He was well aware of the climb of the party of missionaries led by Cook. We
too started from Buwalasi, up the Butandiga ridge to 'Summit Camp'. From
here, next morning, we made for the 'knob', Jackson's Summit. In the cairn at
the top we found the pickle bottle, which we had been told to look for, and we
saw two pricked names-they were those of Margaret Robertson (Government
Protozoologist, in Uganda 1911-1914) and A. H. Cox (who became Assistant
District Commissioner, Mbale as his first appointment in January 1912). Our
names when added were Nos. 29 and 30, so that nearly twenty people had made
their way to this point in the past two years.3' From here Syson pointed north-
wards to Mubiyi and remarked that there would be Franz Josef-Spitze; and my
inference is that Syson had learnt this from Cook or one of his party, who,
looking at Kmunke's map had, very understandably, made this identification2
We too returned on the same afternoon, through 'Summit Camp' which had
been evacuated by our safari during the morning, to camp once more at the
bamboo camp then known as Bulambuli. This round trip had become estab-
lished practice, for the temporary Bagisu porters engaged at Butandiga would
consent to no deviation from the routine which involved any delay at Summit
Camp after a shivering night.
Thus with the evidence of the pickle bottle it can hardly be doubted that Sir
Albert Cook's ascent was of Jackson's Summit, and that he erred in identifying
this with the high point of 14,094 feet which Jackson and Gedge had attained in
1890. All surmise would be set at rest, if it were possible to refer to the list of
names in the bottle. Regrettably this has disappeared, destroyed perhaps by the
Dini ya Misambwa,33 who are thought to have made away with a similar bottle
which had in later years been placed on Wagagai.
Cook records the beautiful weather on the morning of his visit; yet it may be
remarked that, despite his vivid descriptions of the scene, he makes no mention
of the almost breath-taking view across the main crater which breaks upon any-
one who reaches the crest of the true crater rim on a fine day. It may well be
31 It was no doubt as a result of this activity that the Provincial Commissioner, Eastern
Province, published, in December 1916, some notes on the ascent of Elgon in which the
normal marches referred to only go to Jackson's Summit; thus confirming that this was
regarded as the main ascent.
32 Mr. Cox has recently informed me (H.B.T.) that his understanding at the time was
that Franz Josef-Spitze was on the far eastern rim of the crater, which being in Kenya
was regarded as forbidden territory. Its incorrect identification with Mubiyi may therefore
have been current in the first instance only in missionary circles.
33 U.J., 16 (1952), 125. The bottle seems to have been intact as lately as the end of 1933:
see Loveridge, Forest Safari, 1956, p. 75.

128 UGANDA JOURNAL
inferred that he did not in fact look over into the main crater. My own recollec-
tion is of disappointment with the view eastwards from Jackson's Summit-
that the crater was so much less impressive than expected. Today I contemplate
with shame our lack of enterprise in not going the 'second mile'-we were in
fact only looking across to the outer western rim of the main crater. Sir Albert's
descriptions so clearly accord with my memory that I have little doubt that he
also went no further.
Since writing this I have been in correspondence with A. H. Cox, who is quite
satisfied that Sir Albert went only to Jackson's Summit. He confirms that he
(A.H.C.) made the ascent with Miss Robertson in 1913; they went up from
Bulambuli to Jackson's Summit, she riding his mule, and returned through the
forest to Butandiga in one exhausting day. On the summit they found that they
had no pencil, and she produced a safety-pin with which their names were
pricked. He recalls a pleasant story that was later current-that Cook's party
sang a hymn on the summit. Certainly they left a text suitable for the occasion
in the pickle bottle.

It is not the purpose of these notes to consider more recent surveys and maps
of Mount Elgon. It is unfortunate that the mountain lies across the Kenya-
Uganda boundary, so that the mapping of the whole seems never to have been
undertaken as a single operation; the survey from the Uganda side by G. M.
Gibson of the Uganda Survey Department in 1930 stopped abruptly at the
boundary. As a result bewildering discord can be noted among the recorded
altitudes of the various summits.
Some of the best work on the Kenya side of the mountain was done after
World War I by Swedish scientists working latterly from the Swedish Biological
Station built in 1924 on the eastern slopes of the mountain by the interest of
the Swedish Consul-General Sven A. Loven.34 There is room for a co-ordinated
effort by high authority to produce a good large-scale map of the whole moun-
tain comparable with those obtainable for the Swiss Alps. Such a task could be
facilitated by modern photographic survey techniques: and to its execution the
Mountain Club of Uganda should be able to contribute valuable local
knowledge.

34 Reference may particularly be made to a 1: 50,000 scale map of the crater contained
in Erik Nilsson, Quaternary Glaciations and Pluvial Lakes in British East Africa. Stock-
holm: 1932. This map, made by O. Odman in 1927-8 as a member of the Swedish Geo-
graphical Expedition to East Africa, seems to be the only comprehensive map of the
crater. A composite map of Mount Elgon (scale 1:250,000) is contained in Dr. Davies'
The Building of Mount Elgon. Entebbe: 1952. An inadequate assessment of the Swedish
contribution to the mapping of Mount Elgon is at U.J. 2 (1934-5), 250.

THE FISHES OF UGANDA-II
By P. H. GREENWOOD

CHAPTER II
FAMILIES MORMYRIDAE, CHARACIDAE AND CITHARINIDAE
D ESCRIPTIONS of families and genera given in this and succeeding chapters
are based on external or other readily-determined characters exhibited by
species occurring in East Africa. In this respect they cannot be considered fully
diagnostic for all species referable to any particular family over the full range
of its distribution.
Likewise, data on habitats, food, breeding habits and abundance refer
specifically (unless stated otherwise) to investigations made in the waters of
Uganda and neighboring territories. A word of warning must be sounded here.
Although much of the ecological information may seem to be complete and
precise, that is rarely so. Indeed, it is the very lack of exhaustive studies which
enables such apparently succinct accounts to be given. At present our knowledge
is too imperfect for true generalizations to be made.
Table IV summarizes the distribution in Uganda of species dealt with below.
When considering this distribution, it is important to differentiate between
various sections of the Nile, especially those reaches above and below the
Murchison Falls. As mentioned earlier, these Falls provide a sharp demarcation
between fish faunas of so-called Nilotic and Victorian types. Throughout this
account, the river above the Murchison Falls is referred to as the 'Victoria Nile',
whilst that section between the falls and its delta in Lake Albert is called the
'Murchison Nile'. The term 'Albert Nile' is retained for the stretch between
Lake Albert and Nimule.

Order: ISOSPONDYLI
Family: Mormyridae

Body somewhat elongate (31-61 times longer than deep); caudal peduncle
narrow and distinct. Snout variable, from short and rounded to elongate and
trunk-like. Opercular bones hidden beneath the skin, the opening of the
branchial cavity reduced. Eyes small and covered by skin. Head naked, its skin
thick and well-supplied with mucus glands. Mouth non-protractile; teeth
invariably present. Scales small and cycloid; the lateral line complete. Fins
without spines. Generally only the left gonad is developed. Muscles of the
caudal peduncle modified to form an electric organ.
Some preliminary studies have been carried out on the nature and function
of the electric organ in Mormyrus kannume. These show that a continuous
stream of electrical impulses is discharged from the organ, at a variable fre-
quency. With the fish at rest discharge frequency is at its lowest, but rapidly
increases to between 80 and 100 impulses per minute when the fish is disturbed.
Any electrical conductor entering the electro-magnetic field set up around

130 UGANDA JOURNAL
the fish evokes an immediate response. In this way, the organ may be likened to-
a radar warning device, which would be of considerable value to a fish whose
life is spent in muddy waters with poor visibility.
The means whereby disturbances in the electrical field are detected by the
fish is still undetermined. Certain neuro-glandular epidermal cells, the so-called
'mormyromasts', are possibly associated with the perception of such changes.
Besides the electric organ, members of this family show many anatomical,
osteological and histological peculiarities. The brain is especially noteworthy
for its great size. Boulenger (1904) states that in these fishes brain weight equals
A-- of the total weight "a thing unparalleled among lower vertebrates". Since
greatest hypertrophy is found in the cerebellum and acoustico-lateralis areas of
the hind brain, it is not unreasonable to suppose that its functional significance
may be related to the co-ordination of sensory impulses detected through
organs other than the eyes, and therefore indirectly, to the presence of an electric
organ.
The Mormyridae are entirely African in distribution; six genera occur in
Uganda.

KEY TO THE GENERA
Anal fin five or more times longer than the dorsal fin Hyperopisus
Anal fin from three-fifths to twice as long as the dorsal fin 1
Anal fin less than half the length of the dorsal fin Mormyrus

1 10-36 teeth implanted along the entire edge of both jaws
3-10 teeth, confined to the middle of each jaw (b)

(a) (i) Body elongate (5-6 times longer than deep); mouth terminal,
anterior to the eyes; nostrils of each side separated from one
another and distant from the eye Mormyrops
(ii) Body short (3 times longer than deep); mouth inferior and
situated below the eyes; nostrils of each side close together and
placed near the eye Petrocephalus

(b) (i) Mouth inferior, situated well below the horizontal level of the
eye; snout rounded Marcusenius
(ii) Mouth terminal, not, or only slightly, below the horizontal level
of the eye; snout not rounded Gnathonemus

Hyperopisus Gill 1862
Body elongate and compressed. Dorsal fin very short, anal long; pelvic fins
much nearer the pectoral than the anal fin. Nostrils nearer the tip of the snout
than the eye. Mouth terminal, situated below eye level. Teeth in both jaws small
and notched; never more than six in either jaw. Tongue and roof of the mouth
with large, rounded and contiguous teeth, which form a crushing mill.
Only one species is known, but at least two geographically restricted sub-
species are recognized (Daget 1954; see below).

THE FISHES OF UGANDA 131
Hyperopisus bebe (Lac6pede) 1803. (Fig. 4)
Native name: Ngai (Jonam).
Description: Depth of body contained 31-51 times in standard length, length
of head 44-51 times. Dorsal head profile strongly curved. Eye small, its diameter
J (in young fishes) to of head length in adults. Mouth small and terminal;
teeth notched, 3-5 and 4-6 in the upper and lower jaws. Dorsal fin with 13-16
rays, of which the first two or three are unbranched. Anal fin with 54-64 rays.
Lateral line with 90-120 scales.
On the basis of its possessing fewer dorsal and more anal fin rays, and on
other counts, Daget (1954) recognizes a West African sub-species, Hyperopisus
b. occidentalis (Giinther). The local sub-species is accordingly H. b. bebe
(Lac6pbde).

FIG. 4
Hyperopisus bebe. (Drawn by Barbara Williams after Boulenger.)

Coloration: Silver, the dorsal surface iridescent grey-olive or purple.
Size: The largest recorded specimen was 47 cm. long (ex Niger; Daget, 1954).
Habitat: Lake Albert: Fluviatile; not recorded from the lake itself (Worthing-
ton 1929a). Daget (op. cit.) mentions Hyperopisus b. occidentalis as being
common in swamps and other areas inundated during seasonal flooding of the
Niger.
Food: There is no information on the food of H. bebe in Uganda. For the
West African sub-species, molluscs (Marcusen, 1864), and insects (Daget,
op. cit.) are recorded.
Breeding: No data from Uganda. In west Africa, Hyperopisus spawns in
swamps during the rainy season. The nest attributed to H. bebe by Budgett
(1901), has subsequently been shown to be that of Tilapia melanopleura
(Svensson, 1933).
Abundance: Apparently rare.
Distribution: Uganda: Albert and Murchison Niles. Elsewhere: Nile basin,
Senegal, Niger, Volta, Gambia and Lake Chad.

Mormyrus Linn. 1758
Body relatively elongate (31-5 times longer than deep). Snout form very
variable, from short and rounded to elongate and trunk-like. Dorsal fin long, its

132 UGANDA JOURNAL
basal length at least twice that of the anal fin; pelvic fins equidistant from the
pectoral and anal fins or slightly nearer the former. Mouth terminal. Teeth
small and notched; from 3-16 and 6-22 in the upper and lower jaws respectively.
Teeth occurring on the tongue and roof of the mouth are fine and pointed.
Certain species of Mormyrus were venerated by the ancient Egyptians and
were frequently represented on hieroglyphs and murals. To quote from
Giinther's An Introduction to the Study of Fishes-"They (the ancient Egyp-
tians) abstained from eating it because it was one of the three different kinds of
fishes accused of having devoured a member of the body of Osiris, which there-
fore, Isis was unable to recover when she collected the rest of the scattered
members of her husband."
In Uganda there is a widespread belief in the pre-natal influence exerted by
Mormyrus flesh. It is held that if eaten by women, embryonic development will
be so disturbed as to produce infants with a Mormyrus-like cast of countenance.
Another present-day belief is that eating Mormyrus induces sterility in both
men and women.

KEY TO THE SPECIES
Snout elongate and trunk-like (its length about equal to the post-ocular
part of the head) 1
Snout short and stout (about half the post-ocular part of the head)
M. macrocephalus

1 (i) Dorsal fin originating above a point about midway between the
pectoral and pelvic fins; more than 75 rays in the dorsal fin (a)
(ii) Dorsal fin originating above a point only slightly in advance of
the pelvic fins; less than 75 rays in the dorsal fin M. kannume

(a) (i) Snout inclined downwards M. caschive
(ii) Snout straight M. niloticus

Mormyrus macrocephalus Worthington 1929 (Fig. 5)
Native names: Ngolobo (Lunyoro and Luruli); Kimenge (Lunyoro); Nsulusu
(Lukenyi); Kiswala (Ludope); Menya (Lango).
Description: Depth of body contained 4-5 times in standard length, length of
head 31-4 times. Head with straight or slightly convex upper profile; snout about
half as long as the post-ocular part of the head. Eye diameter -1j the interorbital
width. Mouth terminal; teeth notched, 8-12 and 12-16 in the upper and lower
jaws respectively. Dorsal fin with 64-70 branched rays, its origin in advance of
the pelvic fins; 4-5 times longer than the anal fin which has from 17-20 branched
rays. Lateral line with 85-98 scales; 30-34 scales round the caudal peduncle.1
This species is closely related to Mormyrus hasselquisti C. and V. from the
Lower Nile (Worthington 1929 b). When more material of the two species is

1 The caudal peduncle is that region of the body between the origin of the caudal fin
(determined by bending this fin until it is at right angles to the fish's long axis) and the last
dorsal or anal ray, whichever fin is situated more posteriorly.

THE FISHES OF UGANDA 133
available for comparison, it is possible that M. macrocephalus may have to be
regarded as only sub-specifically distinct from M. hasselquisti.

8cm.

FIG. 5
Mormyrus macrocephalus. (Drawn by Barbara Williams.)

Coloration: Dull bronze above, paler below.
Size: The largest recorded specimen was 42 cm. long.
Habitat: Found in most areas of the lakes and rivers from which it has been
collected (Worthington 1929 a).
Food: Predominantly insect larvae, with prawns, oligochaet worms and fishes
occurring less frequently (Worthington 1929 a).
Breeding: No data available.
Abundance: Rare.
Distribution: Known only from Uganda, occurring in Lakes Kyoga and
Kwania, the Victoria Nile and dams in the Teso district (Worthington 1929 a,
and E.A.F.R.O. unpublished reports). A single specimen 42 cm. long was caught
recently at Bugungu, near Jinja. This is the first record of M. macrocephalus
from Lake Victoria. The fish had been feeding almost exclusively on small fishes.

Mormyrus kannume, Forsk. 1775 (Fig. 6)
English: Elephant-snout fish. Native names: Kasulubana, Kasamene
(Lunyoro); Kobia (Jonam); Ntanant (Lunyankole); Kasulubana (Uganda waters
of Lake Victoria).
Description: Depth of body contained 31-5 times in standard length, length
of head 4-5 times. Dorsal head profile straight or somewhat curved, sloping
steeply. Snout produced into a stout trunk whose thickness and angle to the face
are variable; its length equal to, or slightly shorter than, the post-ocular part of
the head. Eye small, its diameter contained 1i-21 times in the interorbital width.
Mouth small and terminal, lips thick. Teeth notched, 4-7 in the upper, and 8-10
in the lower jaw. Dorsal fin with 55-75 rays, originating above or in front of the
pelvic fins. Anal fin short, with 17-22 rays. Caudal fin almost entirely covered
with small scales. Lateral line with 80-116 scales; 26-32 scales round the caudal
peduncle, which is from 1I to twice as long as deep.

UGANDA JOURNAL

Scma
FIG. 6
Mormyrus kannume. (Drawn by Barbara Williams.)

Coloration: Dull bronze above, lighter below.
Size: Largest specimen recorded was 80 cm. long (E.A.F.R.O.); the size range
of adult fishes caught by fishermen in Lake Victoria is usually between 40-60 cm.
Habitat: Lake Victoria: Widespread in both coastal and off-shore waters at
depths up to 120 feet; since little exploratory work has been carried out in
deeper water this figure should not at present be considered as indicative of a
depth limit for the species. (Graham 1929 and E.A.F.R.O., unpublished investi-
gations.)
Lake Kyoga: Open off-shore waters, and less frequently in areas of dense
plant growth (Worthington 1929a).
Lake Albert: Very little information is available from this lake, as previous
surveys have not distinguished between the three Mormyrus species occurring
there. I am told by Mr. A. Anderson, Fisheries Officer, Uganda Game Depart-
ment, that Mormyrus are relatively common in the delta areas of Lake Albert.
Lakes Edward and George: Mormyrus kannume is apparently restricted to
the Kazinga Channel and to the northern, papyrus-fringed shores of Lake
George (Worthington 1932b; Poll and Damas 1939). It is interesting to note
that these authors have found that a slight difference exists between scale counts
of M. kannume occurring in Lake Edward and those from other localities. There
are 32 instead of 26-30 scales round the caudal peduncle of fishes from this
region.
Food: Predominantly insect larvae, particularly larval lake-flies (Chirono-
midae). (Worthington 1929a and 1932b; Graham 1929; E.A.F.R.O. 1954/55
and unpublished data.)
Breeding: Neither the spawning sites nor larvae of M. kannume have been
found. In Lake Victoria it seems probable that any one population of the species
may have an annual reproductive cycle, but that spawning times vary in different
geographical areas of the lake. (E.A.F.R.O. records.)
Abundance: In Lakes Victoria and Kyoga, the species may be considered
common; it contributes substantially to commercial fisheries. On the other hand,
the rather confused records indicate its rarity in Lakes Albert, Edward and
George.

THE FISHES OF UGANDA 135
Distribution: Uganda: Lakes Victoria, Kyoga, Albert, Edward and George;
the Victoria and Murchison Niles. Elsewhere: The whole Nile system as far as
the delta. Probably also in the Athi River system, Kenya (Boulenger 1909).

Mormyrus caschive Linn. 1757
Native names: Apparently no distinction is made between this species and
M. kannume; the same dialect names are therefore used.
Description: Similar to M. kannume except that the dorsal fin has more rays
(75-90) and that its origin is above a point midway between the pectoral and
pelvic fins. Also, the scales are smaller and more numerous in M. caschive, there
being 100-130 in the lateral line series. In many individuals, the snout gives the
appearance of being more slender than in M. kannume.
Coloration: Dark bronze above, lighter below.
Size: Largest individual recorded from Uganda was 100 cm. long.
Habitat: Lakes Edward and George: Similar to M. kannume. Lake Albert:
Probably like that of M. kannume. Worthington (1929a) is of the opinion that,
to a large extent, M. caschive replaces M. kannume in this lake.
Food: Insects, particularly Chironomid larvae (Worthington, op. cit.).
Breeding: No satisfactory data available.
Abundance: Not common in any Uganda lakes.
Distribution: Uganda: Lakes Edward, George and Albert, Albert and Mur-
chison Niles near the delta. Elsewhere: Lower Nile and Bahr-el-Jebel.

Mormyrus niloticus (Bloch-Schneid.) 1801 (Fig. 7)
Native names: Unknown.
Description: Similar to M. caschive, except that the snout is straighter and
prolonged in the axis of the body, not curved ventrally.

8cm.
FlG. 7
Mormyrus niloticus. (Drawn by Barbara Williams after Boulenger.)

Coloration: Probably similar to M. caschive.
There is little ecological information for this inadequately studied species.
Distribution: Uganda: Lake Albert basin. Elsewhere: Lower Nile and Bahr-
el-Jebel.

136 UGANDA JOURNAL
Morhzyrops Miiller 1843
Body elongate (5-6 times longer than deep), not strongly compressed.
Dorsal fin much shorter than the anal; pelvic fins equidistant from the pectoral
and anal fins, or slightly nearer the former. Nostrils distant from one another;
situated nearer the eye than the snout. Mouth terminal. Teeth conical or slender
and notched; 10-36, implanted in a single series along the entire edge of both
jaws; teeth on the tongue and roof of the mouth minute and conical.
A single species occurs in Uganda.

Mormyrops anguilloides (Linn.) 1764 (Fig. 8)
Native names: None known.
Depth of body contained 5-6q times in the standard length, length of head
4 times. Upper head profile slightly concave; snout rounded, its length }-I that
of the head, projecting somewhat beyond the mouth. Teeth notched or slender
and conical, 20-24 in each jaw. Eye small, its diameter 21-3 times length of the
snout. Dorsal fin with 25-28 rays; its origin twice as far from the snout as from
the caudal fin origin. Anal fin with 39-42 rays, its origin considerably in advance
of the dorsal. Caudal fin small, the lobes rounded and almost completely
covered with small scales. Lateral line with 87-96 scales; 20 round the caudal
peduncle which is nearly twice as long as deep.

8cm
FIG. 8
Mormyrops anguilloides. (Drawn by Barbara Williams after Boulenger.)

Coloration: Unknown in life; preserved specimens are olive or greyish-brown
above, whitish below.
Size: The maximum length attained is about 60 cm. although one authority
mentions 100 cm. as the maximum videe Sandon, 1950).
Little has been learned of the ecology of this species and nothing is known
regarding it in Uganda.
Distribution: Uganda: Lake Albert basin (a single specimen from Panyamur
(Worthington 1929a) and one other from Pakwach, collected by Miss R. Osborn,
1956). Elsewhere: Lower and Upper White Nile.

Petrocephalus Marcusen 1854
Body short and laterally compressed. Dorsal and anal fins almost equal in
length, or the latter somewhat longer. Nostrils very close to one another and

THE FISHES OF UGANDA 137
situated near the orbit. Mouth inferior, lying almost vertically below the eye.
Teeth bicuspid; 6-26 in the upper, and 13-36 in the lower jaw. Teeth on the
tongue and roof of the mouth minute and conical.
One species occurs in Uganda.

Petrocephalus degeni Blgr. 1906. (Fig. 9)
Native names: Unknown.
Depth of body contained 31-31 times in standard length, length of head 3J-4
times. Snout very short, about J length of head, rounded and projecting beyond
the mouth, which lies immediately below the eye. Eye small, its diameter I head
length. Teeth notched, 10-12 and 20-22 in the upper and lower jaws respectively.
Dorsal fin with 19-20 rays, its origin above the first anal ray. Anal fin with 27
rays; equally distant from the origins of the pelvic and caudal fins. Caudal
fin with pointed lobes, scaled on the basal part only. Lateral line with 40-41
scales; 12 scales round the caudal peduncle, which is 21 times longer than deep
and about I head length.

8cm.
FIG. 9
Petrocephalus degeni. (Drawn by Barbara Williams.)

Coloration: Silver. Preserved specimens dark grey above, silvery below; fins
grey.
Size: Maximum recorded length 88 mm.
We are completely ignorant of the ecology of this small and rare species.
Until recently, P. degeni was only known from Lake Victoria, but its distribu-
tion is now extended to include Lake Nabugabo and the Victoria Nile (collected
by Makerere College and E.A.F.R.O.).

Marcusenius Gill 1862
Body 3 to 4 times longer than deep, and laterally compressed. Anal fin length
variable, from shorter, to longer than the dorsal fin. Nostrils distant from one
another, their position on the snout variable. Mouth situated well below the

138 UGANDA JOURNAL
horizontal level of the lower orbital margin, terminal or sub-terminal in posi--
tion. Teeth in a single series and restricted to the middle of each jaw; conical,
truncated or notched; 3-10 in both jaws.
Three species occur in Uganda.

KEY TO THE SPECIES
Dorsal fin originating well in advance of the first anal ray 1
Dorsal fin originating above or slightly behind the first anal ray
M. nigricans
1 Dorsal fin with 33-36 rays, its basal length equal to or slightly shorter
than its distance from the head M. petherici
Dorsal fin with 28-31 rays, its basal length less than its distance from
the head M. grahami

Marcusenius nigricans Blgr. 1906. (Fig. 10)
Native names: Unknown.
Depth of body contained 3-31 times in standard length, length of head 31-41
times. Snout j-1 head length, rounded and slightly projecting beyond the mouth,
which is situated below the nostrils. Anterior nostril mid-way between the eye
and the end of the snout. Teeth notched; implanted in a single series restricted
to the middle of each jaw; 5-7 in the upper and 5-9 in the lower jaw. Dorsal fin
with 15-18 rays; anal with 21-25 rays. Caudal fin with rounded lobes, the basal
part covered with small scales. Lateral line with 46-53 scales; 16-20 round the
caudal peduncle, which is 2 to 3 times longer than deep.

8cm.
FIG. 10
Marcusenius nigricans. (Drawn by Barbara Williams.)

Coloration: Uniformly dark brown, or lighter ventrally.
Size: The largest known specimen was 93 mm. long; sexually mature indi-
viduals are usually between 80-90 mm. in length.
Habitat: Marginal vegetation and the fringes of papyrus swamps (Worthing-
ton 1929a and E.A.F.R.O. unpublished).

THE FISHES OF UGANDA 139
Food: Insects, particularly Chironomid larvae.
Breeding: Apart from some scattered records of ripe fishes, nothing is known.
Abundance: M. nigricans is rarely encountered. Its small size and restricted
habitat tend to bias any estimate of abundance.
Distribution: Uganda: Lakes Victoria, Nabugabo and Kyoga; the Kiruni
River (Semliki valley). (Worthington 1929a and E.A.F.R.O. collections.)
Graham (1929) lists M. nigricans as also occurring in East Africa, but with the
qualifying remark that there is some doubt about its supposed wider distribu-
tion. Subsequent collections indicate that it may occur in the Malagarasi River
and even Lake Tanganyika. (Personal observations; Worthington and Ricardo
1936b.)

Marcusenius petherici Blgr. 1898. (Fig. 11)
Native names: Unknown.
Depth of body contained 2*-3 times in standard length, length of head 4J-5
times. Snout 1-} length of head, rounded and projecting beyond the mouth,
which is situated below the nostrils. Teeth notched in young fishes but worn and
truncated in adults; 5-6 in the middle of each jaw. Dorsal fin with 34-36 rays;
anal fin with 24-36, its origin below the middle of the dorsal. Lateral line with
55-65 scales; 12 round the caudal peduncle, which is 21-3 times longer than deep.

87m.
FIo. 11
Marcusenius petherici. (Drawn by Barbara Williams after Boulenger.)

Coloration: Dark brown above, silvery-white below.
Size: Largest recorded specimen was 22 cm. long.
No ecological data are available for this species.
Distribution: Uganda: The Murchison Nile. Elsewhere: Lower White Nile
and the Blue Nile.

Marcusenius grahami Norman 1928. (Fig. 12)
Native names: Kamtontoru (Lunyoro); Ndonge (Lukenyi); Adol (Ludope);
Dono (Lango); apparently none for this species in Uganda waters of Lake
Victoria.

140 UGANDA JOURNAL
Similar to M. petherici; differing principally in having fewer dorsal fin rays
(29-31). The basal length of the dorsal fin is clearly shorter than its distance
from the head. Also, the scales are smaller and more numerous, with 60-69
in the lateral line series.

8cm.
FIG. 12
Marcusenius grahami. (Drawn by Barbara Williams.)

Coloration: Golden brown above, lighter, almost yellow, below.
Size: Adult fishes are usually between 15 and 20 cm. long.
Habitat: Lake Victoria: Common in shallow coastal waters, particularly
where the bottom is sandy (E.A.F.R.O., unpublished).
Lake Kyoga: Especially common in areas overgrown by water-lilies. (Worth-
ington 1929a.)
Food: Insects, particularly Chironomid larvae.
Breeding: Unknown.
Abundance: Fairly common.
Distribution: Known only from Lake Victoria, its affluent rivers and the
Kyoga basin.
Gnathonemus Gill 1862
Body moderately elongate (3-41 times longer than deep) and compressed.
Dorsal and anal fins almost equal in length, or the anal fin noticeably longer.
Nostrils distant from one another and from the eye. Mouth terminally situated
slightly below the horizontal level of the lower orbital margin. Teeth restricted
to the middle of each jaw; conical, truncated or notched; not more than ten
teeth in either jaw. Teeth on the tongue and roof of the mouth minute and
conical.
Three species occur in Uganda.

KEY TO THE SPECIES
Chin not produced into a long, tapering, fleshy appendage 1
Chin so produced G. longibarbis

THE FISHES OF UGANDA 141
1 Dorsal fin with 25-30 rays, anal with 30-37. Lateral line with 70-86
scales G. cyprinoides
Dorsal fin with 22-25 rays, anal with 27-31. Lateral line with 60-70
scales G. victoria

Gnathonemus longibarbis Hilgend. 1888. (Fig. 13)
Native names: Nkoiro (Lunyoro, Lunyara, Ludope and Lango); Mpumbi
(Lukenyi); Ekumbi (Bukakata area, Lake Victoria).
Depth of body contained 3J-5 times in standard length, length of head (exclud-
ing lower lip) 4-41 times. Head with gently curved upper profile. Snout 1-1
length of post-ocular part of the head. Chin produced into a long, cylindrical
and fleshy appendage which is as long as, or longer than the snout. Mouth
terminal. Teeth in a single series restricted to the middle of each jaw, small and
notched; 3-5 and 4-6 in the upper and lower jaws respectively. Dorsal fin with
22-25 rays; anal with 28-31, its origin slightly in advance of the dorsal. Caudal
fin almost entirely covered with small scales, the lobes pointed. Lateral line with
58-64 scales; 10-12 round the caudal peduncle, which is from 2-21 times longer
than deep.

8cm.
FIG. 13
Gnathonemus longibarbis. (Drawn by Barbara Williams.)

Coloration: Dark brown, lighter ventrally.
Size: Sexually mature fishes are between 23-35 cm. long.
Habitat: Lake Victoria: Shallow inshore waters over both sandy and rocky
bottoms; the species is nowhere common (Graham 1929, and E.A.F.R.O.
unpublished). Lake Kyoga: Abundant in or near water-lily swamps (Worthing-
ton 1929a).
Food: Insect larvae, especially Chironomidae, Polymitarcidae and Odonata
(Worthington op. cit. and E.A.F.R.O. unpublished).
Breeding: Unknown.
Distribution: Uganda: Lakes Victoria, Kyoga and Nabugabo.

Gnathonemus cyprinoides (Linn.) 1764
Native names: Unknown.
Depth of body contained 3-4 times in standard length, length of head 4-5f

142 UGANDA JOURNAL
times. Upper head profile gently curved. Snout J- length of head. Eye J-j
length of snout. Mouth small; a globular, fleshy swelling on the chin. Teeth
small and conical, arranged in a single series confined to the middle of each jaw;
5-6 in both jaws. Dorsal fin with 25-30 rays, its origin above the 6th-9th anal fin
ray. Anal fin with 30-37 rays, caudal fin partly covered with small scales, its
lobes pointed. Lateral line with 70-86 scales; 16-18 round the caudal peduncle,
which is from 2-3 times longer than deep.
Coloration: Dark bluish-green dorsally, almost silver ventrally.
Size: Maximum length about 30 cm.
Scarcely any information has been published on the ecology of this species.
In the Nile, Stubbs (quoted by Sandon 1950) observed that G. cyprinoides pro-
vides the main catches in fish traps set on the toiches, where spawning takes
place during flood seasons.
Abundance: Apparently rare in Uganda.
Distribution: Uganda: A single specimen was collected by Worthington from
the Albert Nile at Pakwach. Elsewhere: Nile, from the delta to Bahr-el-Jebel; in
west Africa from the Niger, Chad basin and upper Congo.

Gnathonemus victoria Worthington 1929. (Fig. 14)
Native names: Unknown.
Similar in general appearance to G. cyprinoides, from which species it is
distinguished by having fewer rays in the dorsal and anal fins (22-25 and 27-31
in the two fins respectively), and by its smaller scales (60-70 in the lateral line
series).

8cm.
FIG. 14
Gnathonemus victoria. (Drawn by Barbara Williams.)

Coloration: Golden-brown dorsally, yellow or golden ventrally.
Size: Adult fishes are between 13 and 16 cm. long.
Habitat: Lake Victoria: Similar to that of G. longibarbis with which species
it usually occurs (E.A.F.R.O., unpublished). Lake Kyoga: In or near water-lily
swamps (Worthington 1929a).

THE FISHES OF UGANDA 143
Food: Insect larvae, chiefly Chironomidae.
Distribution: Known only from Lakes Victoria and Kyoga, and from the
Victoria Nile.
The taxonomic status of this species is at present under review. Worthington
(1929b) mentions its close affinity with G. macrolepidotus, whose wide distribu-
tion includes the Zambesi, Kafue, Rovuma and Malagarasi rivers, as well as
Lakes Bangweulu and Rukwa.

Order: OSTARIOPHYSI
Family: Characidae
Body slender and fusiform, covered, except for the head, by cycloid scales.
Cheek partly or completely covered by the sub-orbital bones. Lateral line run-
ning below the mid-line of the flank and caudal peduncle. Fins entirely composed
of soft rays; an adipose dorsal fin is present. Mouth not protractile. Teeth
always present and firmly fixed to the underlying bone.
The Characidae are restricted to fresh waters of Africa, Central and South
America. In Uganda, where two genera are represented, absence of this wide-
spread family from Lakes Edward and George is noteworthy and at present
unexplained. One genus, Hydrocyon, was however present in the Lake Edward
basin during the middle Pleistocene period.

KEY TO THE GENERA
(i) Teeth large and fang-like Hydrocyon
(ii) Teeth stout, with three or more cusps Alestes

Hydrocyon Cuvier 1819
Body elongate (4-5 times longer than deep). Scales cycloid. Mouth large, and
laterally expansible. Teeth firmly fixed, large and fang-like, visible when the
mouth is shut. A second series of teeth lies below the exposed and functional
outer teeth; these may be considered as replacements and are sometimes visible
in the course of eruption. Cheek covered by the thin sub-orbital bones. Nostrils
close together, separated by a flap of skin. Dorsal fin placed above the pelvic
fins; adipose dorsal fin small.
An African genus, with two species occurring in Uganda.

KEY TO THE SPECIES (See Fig. 15)
(i) Distance between the origin of the adipose fin and the base of the
last dorsal fin ray contained 11-1l times in the distance between
the snout tip and the first dorsal ray. Adipose fin grey. Lateral line
with 48-53 scales H. forskali
(ii) Distance between adipose and dorsal fins contained 2-21 times in
the distance between snout tip and first dorsal ray. Adipose fin
black. Lateral line with 45-48 scales H. lineatus

UGANDA JOURNAL

FIG. 15
Hydrocyon, showing the two measurements used in determining species:
(a) is the distance between the adipose dorsal fin and the base of the last dorsal
ray; (b) is the distance between the snout tip and the first dorsal ray.

Hydrocyon forskali Cuvier 1819
English: Tiger fish. Native names: Ngassa (Lunyoro and Lugungu); Ngasia
(Alur); Ngala (Jonam).
Depth of body contained 4-51 times in standard length, length of head 4-5
times; the head 1I-2 times as long as broad. Snout J-k head length. Eye with
vertical adipose lids, its diameter i-k of head. Mouth extending to below the
nostrils or as far back as the anterior orbital margin, the maxilla to the centre
or posterior border. Teeth large, 10-12 and 8-12 in the upper and lower jaws.
Dorsal fin with 9-11 rays, the first two unbranched. Caudal fin deeply forked,
the lobes long and pointed. Lateral line with 48-53 scales; 2 scales between the
lateral line and a scale-like process at the base of the pelvic fin.
Coloration: Silver or silvery-yellow; a dark longitudinal line running through
each scale row above (and including) the lateral line and often through the line
immediately below. Dorsal and adipose fins grey; pectoral and pelvic fins
flushed with orange-yellow, aS are the inferior caudal lobe and the anterior part
of the anal fin. Both caudal lobes are outlined in black along their posterior
margins.
Size: Adults are between 26-59 cm. long. The average weight of individuals
caught in Lake Albert during 1953 was 1-6 lb. (Uganda Government, 1953).
Habitat: In- and off-shore waters, including those of considerable depth
(Worthington 1929a). Mr. Anderson tells me that he has often seen large shoals
swimming at the surface some miles from shore.
Food: Chiefly small fishes (Lates fry, Barilius niloticus and Haplochromis
spp.), but also insect larvae and Crustacea (Worthington, 1929a and Anderson,
personal communication).
Breeding: Spawning sites in Lake Albert are still undiscovered. Worthington
(op. cit.) records breeding H. forskali (i.e., fishes ready to spawn) in various in-
and off-shore areas of this lake, but he failed to find any larval fishes.
Abundance: Fairly common.
Distribution: Uganda: Lake Albert, the Albert and Murchison Niles. Else-
where: Nile, Volta, Niger, Senegal, Gambia, Lakes Rudolf and Chad.

THE FISHES OF UGANDA 145
Hydrocyon lineatus Bleeker. 1862. (Fig. 16)
English: Tiger fish. Native names: Wagassa (in general use). Similar in most
respects to H. forskali, except that the body is somewhat deeper and stouter
(3j-41 times longer than deep) and that the distance between dorsal and adipose
fins is less (see key to species). The scales of H. lineatus are also slightly larger
(45-48 in the lateral line).

8cm.
FIG. 16
Hydrocyon lineatus. (Drawn by Barbara Williams.)

Coloration: Like that of H. forskali, but with a tendency towards a less yellow
and more silver ground colour. Also, the longitudinal stripes, especially those
below the lateral line, are generally more intense in this species.
Size: Worthington (op. cit.) says that specimens up to 63 cm. and weighing
2 kg. were taken by his expedition. None of these was sexually mature. Although
fish weighing 35 lb. (length unspecified) are recorded from net catches, the
average weight of individuals caught in Lake Albert is approximately 3-5 lb,
(Uganda Government, 1953).
Habitat: Shallow inshore waters of Lake Albert, and in the Murchison and
Albert Niles. Although H. lineatus occurs in the same habitats as H. forskali, the
numbers of either species at any one place tends to show an inverse relationship,
so that in certain areas H. lineatus may appear relatively scarce (Anderson,
personal communication).
Food: From published stomach analyses, it would seem that fishes are the
principal food of H. lineatus (Worthington 1929a; Poll 1952a; Worthington and
Ricardo 1936a, Ricardo 1939).
Breeding: No data from Uganda. Poll (1952a) is of the opinion that, in Lake
Tanganyika, H. lineatus probably spawns in the delta areas of affluent rivers.
Distribution: Uganda: Lake Albert, the Albert and Murchison Niles. Else-
where: Lakes Tanganyika and Rudolf; the Nile, Niger, Volta, Senegal, Katanga,
the Congo Basin, Zambesi and Limpopo systems.
Hydrocyon lineatus and H. forskali, together with other species of this genus,
are held in high regard by anglers for their sporting qualities. A devotee of
Hydrocyon, Leander J. McCormick, expressed the opinion of many fishermen
when he wrote "I have stated heretofore in print, and am still ready to maintain

146 UGANDA JOURNAL
my pronouncement, that the tiger fish of Africa is the fiercest fish that swims.
Let others hold forth as advocates for the mako shark, the barracuda, the
piranha of the Amazon, or the blue fish of the Atlantic. To them I say, 'Pish and
tush!'" (in "Game Fish of the World").
Indeed, there are few accounts of African sporting fishes which have not their
purplest patches almost breathlessly eulogizing the tiger fish!
Nevertheless, divergent opinions exist on the palatability of these fishes. The
view most commonly expressed is that the flesh "tastes like cotton wool filled
with needles". On the other hand, I have been told by Europeans and Africans
alike, that tiger fishes make very good eating, provided that some care is taken
with their preparation.
Accounts of a giant Hydrocyon-like fish from Lake Alb'ert are not infrequent,
although only one such specimen has been seen by a European. Worthington
(1929a) discusses the available evidence for the identity of these fishes and
fairly conclusively overrules the possibility of their being Lates (Nile perch), the
only fishes known to reach a considerable size in Lake Albert.
In fact, one is forced to conclude that a very large tiger fish does occur in
these waters. Three guesses may be made as to its identity: that it is an undes-
cribed species of Hydrocyon, that it is a large deep water form of H. lineatus,
or finally, that H. goliath Blgr., a giant species at present known only from Lake
Tanganyika and the Congo baasin, may also occur in Lake Albert.

Alestes Miiller and Troschel 1846
Body slender (3-5 times longer than deep), scales cycloid. Snout short. Mouth
small, non-protractile. Teeth firmly fixed, those of the outer series stout and not
compressed basally. Two series of functional, multi-cuspidate teeth in the upper
jaw, the outer series usually tri-cuspid and less massive than the stout, many-
cusped teeth in the inner series (see Fig. 17). The lower jaw has an outer series
of tricuspid teeth and an inner row of two small, unicuspid and recurved teeth.
Nostrils close to the eye, separated by a flap of skin. Cheek covered by sub-
orbital bones. Dorsal fin situated above or behind the pelvic fins.
The genus is restricted to Africa; five species occur in Uganda.

FIG. 17
Upper jaw of Alestes macrolepidotus, showing the
tricuspid outer teeth and the multi-cuspid inner teeth.

THE FISHES OF UGANDA 147
KEY TO THE SPECIES
Lateral line with 40-50 scales 1
Lateral line with 22-34 scales 2
1 20-26 gill rakers on the lower part of the first gill arch A. dentex
30-38 gill rakers on the lower part of the first gill arch A. baremose
2 Origin of the dorsal fin almost equidistant from the snout and the
origin of the caudal fin; jaws equal anteriorly (a)
Origin of dorsal fin nearer the caudal fin than the snout; upper jaw
projecting beyond the lower A. macrolepidotus
(a) (i) Lateral line with 31-34 scales; 61 or 7 scales between the lateral
line and origin of the dorsal fin. A bright silver longitudinal
stripe A. sadleri
(ii) Lateral line with 27-29 scales; 5-51 between the lateral line and
the origin of the dorsal fin. No longitudinal silver stripe A. nurse

Alestes dentex (Linn.) 1757
Native names: Unknown.
Depth of body contained 31-41 times in standard length, length of head 4J
(in young) to 6 times. Head 11 to twice as long as broad; a well-defined fronto-
parietal fontanelle is present. Eye diameter contained 31 (in young) to 4 times in
head length; well-developed adipose lids. Snout as long as the eye, not project-
ing beyond the lower lip. Maxilla not reaching the anterior orbital margin. 6 and
8 teeth in the outer and inner series of the upper jaw, 8 and 2 in these series in
the lower jaw. Gill rakers long and slender, 20-26 on the lower limb of the first
gill-arch. Dorsal fin with 10-11 rays, the first two unbranched; its origin above
the pelvic fin insertions. Anal fin with 22-26 rays, the first three unbranched.
Lateral line with 44-50 scales.
On the basis of slight differences in body proportions, and in its lower modal
fin-ray and scale counts, Daget (1954) distinguishes a west African sub-species
Alestes d. sethente (C. and V.). Uganda fishes should be referred to the sub-
species Alestes d. dentex (Linn.).
Coloration: Silver, bluish-grey above. Dorsal fin grey, lower caudal lobe red,
the posterior margin of both lobes outlined in black. In adult fishes, the pelvic
and anal fins have an orange-red flush.
Size: Adult fishes are from 30-55 cm. long, and weigh about one pound.
No ecological data are available for this species in Uganda, where it is appar-
ently rare. Daget (op. cit.) discusses the biology of Alestes d. sethente at some
length.
Distribution: Uganda: Lake Albert and the Murchison Nile. Elsewhere: Lower
Nile, White Nile, Chad basin, Senegal, Gambia, Niger, and probably Lake
Rudolf.
Alestes baremose (Joannis) 1835. (Fig. 18)
Native names: Ngara (in general use; probably also covers A. dentex).
Very similar to A. dentex. The principal difference lies in the gill rakers,
which are both more numerous (27-38) and more slender in this species.

UGANDA JOURNAL

8cm.
FIG. 18
Alestes baremose. (Drawn by Barbara Williams.)

Coloration: As for A. dentex.
Size: Adult fishes are between 30-55 cm. long, with an average weight of
about one pound (Worthington 1929a and Uganda Government 1953).
Habitat: Probably limited to the inshore regions of Lake Albert. In any one
locality, their numbers are inversely related to those of Hydrocyon spp. occur-
ring in that area (Uganda Government, 1953). H. baremose forms an important
element in the seine net catches of African fishermen.
Food: Apparently omnivorous, with small Crustacea, insects and, less fre-
quently, fishes predominating (Worthington 1929a; Worthington and Ricardo
1936a).
Breeding: Spawning sites and seasons are unknown. Worthington (op. cit.)
states that breeding fishes of both sexes were found in sheltered bays around the
lakeshore, as well as below the Murchison Falls and in the Albert Nile at
Pakwach.

Abundance: Common.
Distribution: Uganda: Lake Albert, the Albert and Murchison Niles. Else-
where: Lower Nile, Blue and White Niles, Lake Rudolf, the Chad basin, Senegal,
Gambia and Niger.

Alestes macrolepidotus (Cuvier and Valenciennes) 1849. (Fig. 19)
Native names: Waraga (Lunyoro and Lugungu); Gowa (Alur); Owaro
(Jonam).
Depth of body contained 31-41 times in standard length. Length of head
31-4J. Head much flattened dorsally, 1- to twice as long as broad; fronto-
parietal fontanelle absent. Snout pointed, clearly projecting beyond the lower
jaw; as long as (in young fishes) to twice the diameter of the eye (in adults). Eye
contained 3 (in young) to 5 times in head length (in adults), the adipose lids
very feebly developed. Maxilla not reaching the anterior orbital margin. 8-10
teeth in each of the upper jaw series, 6-8 and 2 in the outer and inner lower jaw
series. Gill rakers long and slender, 15-22 on the lower limb of the first gill arch.
Dorsal fin with 9 or 10 rays, the first two unbranched; its origin considerably
nearer the caudal fin than the snout. Anal with 15-17 rays, the first three

THE FISHES OF UGANDA 149
unbranched. Lateral line with 22-26 scales; 4- scales between the lateral line
and the origin of the dorsal fin.

8Cm.
FIG. 19
Alestes racrolepidotus. (Drawn by Barbara Williams.)

Coloration: Silver, bluish-grey dorsally; a dark blotch above the pectoral fin
in young fishes, as well as a dark, narrow longitudinal band on the flank. Both
these markings are intensified after death, but are faint in life. Dorsal, caudal
and adipose fins orange-red or pinkish. In adults, the pectoral and pelvic fins,
together with the operculum, are suffused with a pinkish flush.
Size: Adult fishes are from 40-55 cm. long.
Habitat: Inshore regions of Lake Albert; also in rivers.
Food: Insects, bottom debris and small fishes (Worthington 1929a, and
Anderson, personal communication).
Breeding: Spawning sites and seasons unknown. Worthington (op. cit.) records
breeding individuals (i.e. fishes ready to breed) from below the Murchison Falls
and at Pakwach, but not in the lake itself.
Abundance: Common.
Distribution: Uganda: Lake Albert, the Albert and Murchison Niles. Else-
where: The Nile, Omo River, Senegal, Niger, Volta, Chad basin, Congo.

Alestes nurse (Riippell) 1832. (Fig. 20)
Native names: Nsoga (in general use).
Depth of body contained 3-4 times in standard length, length of head 31 (in
young) to 4 times. Head 1i to twice as long as broad; fronto-parietal fontanelle
absent. Snout rounded, equal to or slightly longer than the eye, whose diameter is
contained 21 (in young fishes) to 4 times (in adults) in the head length; adipose
lids barely visible. Maxilla not reaching the anterior orbital margin. 8 teeth in
each series of the upper jaw, 8 and 2 in the outer and inner lower jaw series. Gill
rakers moderately long, 16-20 on the lower limb of the first gill arch. Dorsal fin
with 9-11 rays, the first two unbranched; its origin above the pelvic fin insertion.
Anal fin with 14-19 rays, the first three unbranched. Lateral line with 27-29
scales; 5-51 between the lateral line and the origin of the dorsal fin.

150 UGANDA JOURNAL
An interesting example of sexual dimorphism is seen in the anal fin of this
species. The fin-margin is straight or very slightly concave in females, whereas it
is markedly convex in males; in consequence the fin is much larger in this sex.
Further, in males, the individual rays are stouter and coarser than those in the
anal fin of females.

8cm.
FIG. 20
Alestes nurse. (Drawn by Barbara Williams.)

Coloration: Silver, bluish-grey to blue-black dorsally. A large black blotch
on the caudal peduncle, with a narrow extension onto the caudal fin; a round
black spot behind the head, often faint or invisible in life but intensified after
death. Dorsal fin grey, anal and pelvic fins with a faint orange-red or yellow
tinge; adipose fin orange; caudal yellow or orange.
Size: Adults between 15-23 cm. long, occasionally as large as 27 cm.
Habitat: The shallow coastal areas of Lake Victoria, in water less than sixty
feet deep (Graham 1929, and E.A.F.R.O. investigations). No data are available
from the Victoria Nile or Lake Albert.
Food: Varied, with insects, especially Chironomid pupae, predominating;
small fishes (chiefly Haplochromis spp. and Engraulicypris argenteus) occur
less frequently. Fragments of plants found in the stomachs of many individuals
suggest that plants may contribute to the diet of A. nurse. (E.A.F.R.O. unpub-
lished records, and Graham 1929.) Occasionally specimens are found which
have fed almost exclusively on snails. (Graham op. cit. and E.A.F.R.O. unpub-
lished.) The stomachs of three fishes from Lake Rudolf contained only zoo-
plankton (Worthington and Ricardo 1936a).
Breeding: The precise spawning seasons and habitats of A. nurse have not
been determined in Lake Victoria, although the species is known to run up
rivers and streams during the rainy seasons (E.A.F.R.O. unpublished).
Abundance: Common; often in large shoals.
Distribution: Uganda: Lakes Victoria and Nabugabo, the Victoria Nile (but
apparently absent from Lake Kyoga), Lake Albert (E.A.F.R.O. collection).
Elsewhere: Lakes Rudolf and Chad, the Nile, Niger, Volta, Gambia, Senegal;
also in Liberia and the Cameroons.

THE FISHES OF UGANDA 151
Alestes sadleri Blgr. 1906
Native names: Confused with A. nurse, hence Nsoga.
In general appearance very similar to A. nurse, with which species it is
generally confused. Alestes sadleri differs in the following characters: smaller
scales (31-34 in the lateral line series and 61 or 7-usually 7-between the
lateral line and the origin of the dorsal fin) and the greater number of rays in the
anal fin (20-21).
Sexual dimorphism in the anal fin is as described for A. nurse.
It should be noted however, that in a small sample of nine fishes from Lake
Nabugabo only three had 6 or 61 scales between the lateral line and dorsal fin;
the remainder had 5 scales along this line. In all other respects these fishes
agreed closely with the description of A. sadleri.
In life, the presence in A. sadleri of a narrow, intensely silver longitudinal
band running the whole length of the body serves immediately to distinguish the
two species. After death this band rapidly darkens; in preserved specimens it is
black. The post-mortem black blotch behind the head of A. nurse is absent in
A. sadleri.
The adult size range of A. sadleri is less than that of A. nurse; sexually mature
individuals measure between 8 and 10 cm. long.
Coloration: Silver, bluish to black dorsally; an intensely silver lateral streak
runs from behind the operculum to the caudal peduncle, where it merges with an
irregular black blotch which extends onto the caudal fin base. All fins grey.
Habitat: Shallow coastal waters of Lake Victoria; also amongst the marginal
vegetation of Lakes Kyoga and Nabugabo. Because of their small size, A.
sadleri is rarely caught by fishermen, except when large shoals are netted in
seines.
Food: Insects, chiefly Chironomid pupae (E.A.F.R.O. unpublished).
Breeding: Unknown, but probably similar to that supposed for A. nurse (see
above).
Distribution: Known only from Uganda: in Lakes Victoria, Nabugabo and
Kyoga; also from the Victoria Nile. (Graham 1929, Worthington 1929a and
E.A.F.R.O. collections.)
Reference must also be made to two other Alestes species: one has been tenta-
tively recorded from Lake Albert, and the other may occur in the Lake Kyoga
basin.
Cunnington (1920) doubtfully lists Alestes macropthalmus as occurring in
Lake Albert. Its presence there has never been confirmed.
The second species, Alestes jacksoni Blgr., is known only from a single speci-
men whose locality is given as "Malawa River, Kavirondo, Lake Victoria". The
headwaters of this river are on Mt. Elgon, and it drains into the Kyoga system
and not into Lake Victoria. Thus, if its river locality is correctly stated, A. jack-
soni should be included in the list of Uganda species.
Alestes jacksoni occupies a position intermediate between A. nurse and A.
sadleri. Its large scales (lateral line 26; 4j between the dorsal fin and the lateral
line) betoken affinity with A. nurse, whilst its coloration suggests relationship
with A. sadleri.
Further specimens of this interesting species would be most welcome.

152 UGANDA JOURNAL
Family: Citharinidae
Body short, deep and laterally compressed (13-31 times longer than deep).
Scales ctenoid or cycloid. Lateral line running along the middle of the flank and
caudal peduncle. Mouth not protractile. Teeth slender and movably implanted.
Cheek covered; or partly covered, by the sub-orbital bones. Fins with soft rays
only; an adipose dorsal fin is present.
The Citharinidae are restricted to African fresh-waters; two genera occur in
Uganda.

KEY TO THE GENERA
Scales ctenoid; teeth strong and bicuspid Distichodus
Scales cycloid; teeth minute and conical Citharinus

Distichodus Miiller and Troschel 1854
Body short, deep (2-31 times longer than deep), and compressed. Scales small
and ctenoid. Teeth bicuspid, usually in 2 series. The two halves of the lower jaw
are broadly united (see Fig. 21 a). Dorsal fin situated above the pelvic fin
insertion. Adipose dorsal and caudal fins covered with small scales.
Two species occur in Uganda.
(i) Lateral line with 90-112 scales; 18-21 scales between the lateral
line and the origin of the dorsal fin D. niloticus
(ii) Lateral line with 80-98 scales; 15-17 between the lateral line and
the origin of the dorsal fin D. rostratus

a b
FIG. 21
(a) Lower jaw of Distichodus niloticus, ventral view.
(b) Lower jaw of Citharinus latus, ventral view.

Distichodus niloticus (Linn.) 1762. (Fig. 22)
Native names: Wachone (in general use).
Depth of body contained 21-31 times in standard length, length of head 3
times (in young fishes) to 5 times in adults. Snout rounded, slightly projecting
beyond the mouth. Maxilla extending to below the nostrils. Upper surface of the

THE FISHES OF UGANDA 153
head, and the operculum ridged with low, radiating striations. Teeth in two
series, outer with 24-36 teeth. Dorsal fin with 22-26 rays, the first 3-6 unbranched.
Adipose dorsal covered with small scales. Anal fin with 13-15 rays, the first 3 or
4 of which are unbranched. Caudal fin forked, its rounded lobes almost entirely
scaled. Lateral line with 90-110 scales; 18-20 between the lateral line and the
origin of the dorsal fin; 15-18 between the insertion of the pelvic fin and the
lateral line.

8cm.
FIG. 22
Distichodus niloticus. (Drawn by Barbara Williams.)

Coloration: Purplish-silver above, silver below; young fishes with several
vertical dark bars on the flank.
Size: The largest recorded individual was 83 cm. long (Worthington 1929a).
Habitat: Shallow inshore waters of Lake Albert, especially in the region of
deltas; also in rivers.
Food: Submerged plants, molluscs and small Crustacea (Worthington 1929a).
Breeding: Unknown.
Abundance: Fairly common. The species forms an important element in the
catches of 8" mesh gill nets.
Distribution: Uganda: Lake Albert and Murchison Niles. Elsewhere: Nile,
White Nile, Lakes Rudolf and No.

Distichodus rostratus Giinther 1864
Native names: Unknown, but probably included in Wachone (= D. niloticus).
Very similar to D. niloticus, differing mainly in having larger scales (80-98 in the
lateral line and 15-17 between the lateral line and the dorsal fin), and an
apparently coarser and deeper head.
Until recently, the inclusion of this species in the faunal list of Lake Albert
depended solely on a head of an adult fish collected by J. S. Budget from the
Murchison Nile in 1902. A second specimen was obtained in April 1956 by
Miss R. Osborn while collecting for Yale University.
Reference should be made to Daget (1954) for ecological data on this species
in the Upper Niger.

154 UGANDA JOURNAL
Distribution: Uganda: Murchison Nile. Elsewhere: Lower Nile, Bahr el Jebel,
Lake Chad, Senegal, Niger and Gambia.

Citharinus Cuvier 1817
Body short, deep (13-21 times longer than deep) and strongly compressed.
Scales cycloid and small. Mouth large; a single row of minute teeth. The two
halves of the lower jaw are narrowly united (see Fig. 21 b). Dorsal fin situated
behind the insertion of the pelvic fins. Adipose dorsal large and scaled. Caudal
fin scaled on its basal part only.
Two species occur in Uganda.
(i) Base of adipose dorsal fin shorter than its distance from the pos-
terior margin of the dorsal fin; lateral line with 77-92 scales C. citharus
(ii) Base of adipose fin longer than its distance from the posterior
margin of the dorsal fin; lateral line with 60-77 scales C. latus

Citharinus citharus (Geoffr.) 1809
Native names: Mpoi (Lunyoro, Lugungu and in general use); Aporo (Alur
and Jonam).
Depth of body contained 13-21 times in standard length, length of head 3-4
times. Upper head profile deeply concave and steeply sloping. Snout short, li to
twice as long as broad, projecting slightly beyond the mouth. Eye with narrow,
vertical adipose lids. Teeth minute and numerous, arranged in a single series.
Gill rakers short, fine and closely set. Dorsal fin with 18-21 rays, the first 4-6 of
which are unbranched; adipose dorsal large, its base 1-1 its distance from the
dorsal fin. Anal fin with 25-31 rays, the first 3 or 4 unbranched. Lateral line with
77-92 scales; 20-25 between the lateral line and the origin of the dorsal fin.
Coloration: Silver, base of adipose dorsal fin dark grey. Pelvic and anal fins
orange-red except at the base. Inferior lobe of the caudal fin with a distinct
orange flush; all other fins grey.
Size: Sexually mature fishes are between 50 and 78 cm. long. Worthington's
records suggest that males reach maturity at a somewhat smaller size than
females.
Habitat: Shallow inshore waters of Lake Albert; absent or rare in deeper
open water. Its off-shore range is probably somewhat greater than that of
Distichodus niloticus.
Food: Macroplankton, particularly those elements such as Crustacea and
diatoms which settle on the lake bottom. The species has often been seen swim-
ming at the surface and apparently feeding on the scum collected there
(Worthington 1929a). However, since Citharinus possesses an accessory
respiratory organ, this behaviour may well be associated with breathing and not
feeding, especially if the dissolved oxygen content of the water is low.
Breeding: No information is available on the breeding habits of C. citharus in
Lake Albert. In the rivers Niger and Gambia, spawning takes place in swampy
areas during the season of heavy rainfall (Daget 1954; Svensson 1933). Young
fishes less than 20 mm. long are quite unlike adults. The body is more slender
(depth about standard length) and the characteristic, steep adult profile is
wanting. Typical adult body-form is acquired when the young have reached

THE FISHES OF UGANDA 155
30 mm. standard length videe Svensson, op. cit.). According to this author early
growth is rapid, about 10-12 cm. in two months.
Abundance: Common. C. citharus is an important commercial fish in Lake
Albert, although there has been a marked decline in the numbers caught since
1944. (Uganda Government, 1952.)
Distribution: Uganda: Lake Albert, the Albert and Murchison Niles. Else-
where: Lower Nile, Blue and White Niles, Lake No, Chad basin, Senegal, Niger
and Gambia.
Worthington (1932a) has described a sub-species Citharinus c. intermedius
from Lake Rudolf, which, as its name implies, is intermediate in some characters
between C. c. citharus (the sub-species in Uganda) and C. latus M. and T. (see
below).

Citharinus latus Miiller and Troschell 1845. (Fig. 23)
Native names: Unknown.
Similar to C. citharus except that the adipose dorsal fin is longer (its basal
length greater than the distance separating it from the dorsal fin), and the scales
somewhat smaller (lateral line with 60-77 scales). Coloration of the two species
is identical.

8cm.

Citharinus latus.

FIG. 23
(Drawn by Barbara Williams.)

In Lake Albert the ecological requirements of the two Citharinus species are
probably similar. Certainly this is so in the Niger (Daget 1954). It will probably
be found, however, that in Lake Albert, C. latus is less common than C. citharus.
Distribution: Uganda: Lake Albert. Elsewhere: Lower Nile to Bahr-el-Jebel,

156 UGANDA JOURNAL

Senegal, Niger and Old Calabar. Boulenger (1909) also lists the Kingani River
(East Africa).
TABLE IV
Distribution in Uganda of fishes of the families Mormyridae, Characidae and Citharinidae.
An interrogation mark indicates that the species probably occurs in that area also.

Edward
LAKEs Victoria Kyoga Nabugabo and Albert Elsewhere
George
Mormyridae
Hyperopisus bebe .. P Albert and
Murchison Niles
Mormyrus macrocephalus P P Lake Kwania;
Victoria Nile,
dams in Teso
district
Mormyrus kannume .. P P P P Victoria and
Murchison Niles
Mormyrus caschive .. P P
Mormyrus niloticus .. P
Mormyrops anguiloides P Albert and
Murchison Niles
Petrocephalus degeni .. P P Victoria Nile
Marcusenius nigricans .. P P P Kivuni river
(Semliki valley)
Marcuseniuspetherici .. ? Murchison Nile
Marcusenius grahami .. P P Affluent rivers of
Lake Victoria
Gnathonemus longibarbis P P P
Gnathonemus cyprinoides P Albert Nile
Gnathonemus victoria P P Victoria Nile
Characidae
Hydrocyonforskali .. P Albert and
Murchison Niles
Hydrocyon lineatus .. P Albert and
Murchison Niles
Alestes dentex.. .. P Murchison Nile
Alestes baremose .. P Albert and
Murchison Niles
Alestes macrolepidotus P Albert and
Murchison Niles
Alestes nurse .. .. P ? P P Victoria Nile
Alestes sadleri .. .. P P P
Citharinidae
Distichodus niloticus .. P Albert and
Murchison Niles
Distichodus rostratus .. P Murchison Nile
Citharinus citharus .. P Murchison and
Albert Niles
Citharinus latus .. .. P Murchison and
Albert Niles

REFERENCES
(Only those publications which are not included in the bibliography appended to
Chapter I are listed here.)
Boulenger, G. A. (1904). Teleostei, in 'Fishes, Ascidians, etc.', Cambridge Natural
History, London. (2nd edition, 1932.)
Cunnington, W. A. (1920). The fauna of the African lakes, a study in comparative
limnology with special reference to Tanganyika. Proc. zool. Soc., London,
507-622.

THE FISHES OF UGANDA 157
Daget, J. (1954). Les poissons du Niger supdrieur. Mdm. Inst. Franc. Afr. noire,
No. 36,1-391.
Giinther, A. C. L. G. (1880). An introduction to the study of fishes. Edinburgh.
Marcusen, J. (1864). Die Familie der Mormyren. Eine anatomisch-zoologische
Abhandlung. Mem. Acad. Sc. St. Pdtersb., (7), 7, 1-162.
Ricardo, C. K. (1939). The fishes of Lake Rukwa. J. linn. Soc. (Zool.), 40, 625-657.
Sandon, H. (1950). An illustrated guide to the freshwater fishes of the Sudan.
London.
Svensson, G. S. 0. (1933). Fresh water fishes from the Gambia river. K. Svenska
Vet. Akad. Handl. (3), 12, 1-102.
Vesey-Fitzgerald, B. and Lamonte, F. (Editors) (1949). Game Fish of the World.
London.
Worthington, E. B., and Scientific results of the Cambridge expedition to the East
Ricardo, C. K. (1936a). African lakes, 1930-31.-No. 15. The fish of Lake Rudolf
and Lake Baringo. J. linn. Soc. (Zool.), 39, 353-389.
(1936b). The fishes of Lake Tanganyika (other than Cichlidae).
Proc. zool. Soc., London, 1061-1112.

CHAPTER III

FAMILY CYPRINIDAE: THE GENERA Labeo, Gara, Barilius AND Engraulicypris

Family: Cyprinidae
Body covered by cycloid scales, the head naked. Jaws completely devoid of
teeth. One or two pairs of circum-oral barbels, which are, however, absent in
some species. No adipose dorsal fin; the last unbranched dorsal ray sometimes
ossified and spine-like. Well developed sickle-shaped pharyngeal bones carrying
1-3 series of teeth. In many species, wart-like protuberances develop on the
snout during the breeding season, or at sexual maturity.
This family, comprising numerous genera and species, is wide-spread in the
freshwaters of Europe, Africa, Asia, Northern and Central America. Indigenous
species are notably absent from South America, Australia and Madagascar.
Five genera are found in the lakes and rivers of Uganda.

KEY TO THE GENERA
Cheek below eye naked, not covered by the sub-orbital bones. Lateral
line situated along the middle of the flank, or slightly nearer the ventral
outline, always running along the middle of the caudal peduncle 1
Cheek below eye covered by the thin sub-orbital bones. Lateral line
situated below the middle of the flank and caudal peduncle 2
1 Greater part of the dorsal fin an advance of the pelvic fin base (a)
Origin of the dorsal fin above the pelvic fin base, or very slightly in
advance of, or behind this point Barbus

(a) (i) A flap of skin immediately in front of the upper lip; opening
to the gill-chamber extending to below the pectoral fin base Labeo
(ii) No flap of skin in front of the upper lip; opening to the gill-
chamber not extending below the pectoral fin base; a well
developed, almost circular disc on the chin and confluent with
the lower lip Gara

2 Greater part of the dorsal fin situated in advance of the anal fin Barilius
Almost the entire dorsal fin situated above the anal fin Engraulicypris

Labeo Cuvier 1817
Scales usually large. Lateral line running along the middle of the flank and
caudal peduncle. Dorsal fin with three unbranched rays, none spinous, its origin
well in advance of the pelvic fin insertions. Mouth large, inferior and protrac-
tile, the lips well developed; immediately in front of the upper lip there is a flap
of skin dependant from the snout. Both jaws provided with horny cutting edges.
Barbels hidden.

THE FISHES OF UGANDA 159
The genus occurs in Africa and southern Asia; in Uganda it is represented by
-three species.

KEY TO THE SPECIES
Inner surface of the lips with numerous small transverse folds 1
Inner surface of the lips smooth Labeo horie

1 Eyes equally visible from above and below (a)
Eyes easily visible from above but barely if at all visible from below
L. forskali
(a) (i) Snout swollen and fleshy; caudal peduncle as long as, or
slightly longer than deep L. coubie
(ii) Snout not noticeably swollen or fleshy, caudal peduncle longer
than deep L. victorianus

Labeo horie Heckel 1846
Native name: Karuka (Lunyoro); Posso (Lugungu and Alur) Nduka (Jonam).
Description: Depth of body contained 3-4 times in standard length, length of
head 4-6 times (relatively longer in small fishes). Snout broadly rounded, about
I head length and shorter than the post-ocular part of the head. Eye visible from
above and below, its diameter contained 3-6 times in head length (relatively
larger in small fishes). Lips with numerous rows of papillae. A minute barbel is
present, but hidden in folds of skin at the angle of the mouth. Dorsal fin with III,
12-14 rays, its upper margin straight or slightly convex. Lateral line with 40-44
scales. Caudal peduncle as long as deep.
Coloration: Greenish brown above, golden below; young fishes have a distinct
but ill-defined dark spot on the lateral-line scales above the pectoral fin.
Size: Worthington (1929a) gives 72 cm. as the length of the largest fish
recorded from Lake Albert.
Habitat: Restricted to the inshore waters and sheltered bays of Lake Albert.
(id., ibid.)
Food: Virtually unknown; a colourless slime, mud and sand have been
recorded from the guts of this species (Worthington, op. cit.).
Distribution: Uganda: Lake Albert, Murchison Nile. Elsewhere: Nile.

Labeo forskali Riippell 1835 (Fig. 24)
Native name: Omoruma (Lunyankole).
Description: Depth of body contained 3-4 times in standard length, length
of head 4-5 times. Snout rounded, more or less fleshy and swollen, with a
distinct curved transverse groove on the upper surface; snout length about half
that of the head and somewhat longer than the post-ocular part. Eye not visible
from below, its diameter J-* head length (relatively larger in small fishes). Inner
surface of the lips with numerous transverse folds. Barbels minute and hidden.
Dorsal fin with III, 9-11 rays, its upper margin concave. Lateral line with 37-42
scales. Caudal peduncle 1U-1 times longer than deep.

UGANDA JOURNAL

Fio. 24
Labeo forskali. (Drawn by Barbara Williams.)

Coloration: Olive-green above, lighter, almost yellow below; scales of the
flanks with pinkish centres.
Size: The largest specimen from Lake Edward was 48 cm. long (Worthington
op. cit.).
Habitat: In Lake Edward, the deep water close to rocky shores. (Worthington,
op. cit., Poll and Damas, 1939.)
Food: Plant debris and mud.
Distribution: Uganda: Lake Edward, Agoye river, West Nile (a tributary of
the Ala river which flows into the Nile).

Labeo coubie Riippell 1832
Native name: Kwangurameli (Lunyoro).
Description: Similar to Labeo forskali except that the body is usually some-
what deeper, the eye is visible both from above and below, the caudal peduncle
is as deep as or slightly deeper than long, and the dorsal fin has III, 12-14 rays.
Coloration: Grey-blue above, silvery-white below; scales on the flanks have
deep mauve centres. Young fishes are greyish-silver, with numerous wavy black
lines on the flanks, and a large black spot at the base of the caudal fin.
Size: The largest recorded specimen from Lake Albert was 74 cm. long and
weighed 9 lb.
Habitat: The inshore waters of Lake Albert, particularly in sheltered bays
(Worthington, 1929 a).
Food: Plant debris and mud (id., ibid.).
Distribution: Uganda: Lake Albert, Murchison Nile. Elsewhere: Nile system,
Niger, Volta and Lake Chad.

Labeo victorianus Boulenger 1901 (Fig. 25)
Native name: Ningu (Lake Victoria); Nsuku (Lunyoro and Lunyara); Okok
(Ludope); Apok (Lango).
Description: Depth of body equal to or slightly greater than length of head,
which is contained 4-5 times in the standard length. Snout rounded; as compared
with L. forskali and L. coubie it is neither fleshy nor swollen; snout length from

THE FISHES OF UGANDA 161
J-1 of the head and less than the post-ocular part. Eye equally visible from above
and below, its diameter contained 5-7 times in the head length (relatively larger
in small fishes). Lips with transverse folds on the inner surface. Barbels minute
and hidden. Dorsal fin with III, 9-10 rays, its upper margin concave. Lateral line
with 37-39 scales. Caudal peduncle 1 times longer than deep.

FIG. 25
Labeo victorianus. (Drawn by Barbara Williams.)

Coloration: Olivaceous above, lighter or creamy-white below. Dorsal, anal
and pelvic fins often tipped with orange.
Size: The largest specimen, recorded by Graham (1929) was 41 cm. long, but
the usual size for fishes caught by African fishermen is between 20 and 30 cm.
Breeding: Although the exact spawning sites of L. victorianus are still
unknown, the large number of sexually active fishes which run up rivers during
the "rains", suggests that the species spawns in rivers at that period.
Habitat: Lake Victoria: shallow inshore waters. Lake Kyoga: open waters
away from the water-lily zone. L. victorianus also occurs in the major affluent
rivers of Lake Victoria and in the Victoria Nile.
Food: Mud, plant debris and unidentifiable slime. Specimens kept in aquaria
have been seen to graze on epizootic Rotifers and fungi growing on the scales
and bodies of other fishes (Fryer, et al., 1955). Mr. Copley, Fish Warden, Kenya,
informs me (in litt.) that in a film taken underwater, he has seen Labeo cylin-
dricus Peters grazing on the backs of submerged Hippopotamus.
Economic Importance: Smoked ningu are relished by most Africans. In many
parts of Lake Victoria the species forms the basis of a flourishing fishing industry
carried out with small mesh nets.
Distribution: Lakes Victoria and Kyoga; the Victoria Nile, and major rivers
flowing into Lake Victoria.

Gara Hamilton Buchanan 18221
Body feebly compressed, somewhat rounded; covered by moderately large
scales. Lateral-line running along the middle of the flank and caudal peduncle.
I The genus Gara was listed by its previous name, Discognathus, in the systematic list
given in Chapter I.

162 UGANDA JOURNAL
Mouth inferior and protractile. Lips thickened; a large, nearly circular suctorial
disc developed on the chin. One or two small barbels on each side. Gill opening
not extending below the level of the pectoral fin base. Dorsal fin with 9-11 rays,
6-8 of which are branched.
The genus occurs in southern Asia and north-east Africa; a single species is
recorded from Uganda.
Gara johnstoni (Boulenger) 1901 (Fig. 26)
Description: Depth of body equals length of head, which is contained 41-5
times in the standard length; head somewhat depressed and flattened above.
Snout rounded and projecting above the mouth. Eye near the upper profile of
the head, its diameter contained 31-5 times in the head length (relatively larger
in small fishes) and l1-21 times in the inter-orbital width; the pupil situated in
the second half of the head. Width of mouth contained 2-21 times in head
length. Lips moderately well developed, the upper with a weak fringe of
papillae, the lower in contact with the suctorial disc, which is longer than broad.
Two minute barbels on each side. Dorsal fin with III, 6-7 rays, its position
equally distant from the nostrils and the caudal fin origin. Pectoral fin from -4,
to as long as the head, widely separated from the pelvic fins which are situated
below the posterior 4 of the dorsal fin. Lateral line with 35-37 scales; 5-5 scales
between this line and the first dorsal ray, 31-4 between the lateral line and the
pelvic fin base. Caudal peduncle as long as, to 1 longer than, deep.

icm
FIG. 26
Gara johnstoni. (The mouth, viewed ventrally, is inset.) (Drawn by Barbara Williams.)

Size: Rarely longer than 110 mm.
Coloration: Dark grey-brown to bronze above, creamy-white below. Fins
greyish; 2-4 black spots at the base of the dorsal fin. In males, the dorsal, anal
and caudal fins are tipped with orange.
Habitat: The species is abundant in the turbulent waters of the Victoria Nile
near Jinja. Otherwise, very little is known of its habitat preferences.
Food: Algae and insect larvae, probably scraped from the surface of rocks.
Distribution: The Victoria Nile and possibly Lake Victoria. A single record
from the Nairobi River, Mount Kilimanjaro (Boulenger 1909) requires con-
firmation.
Barilius Hamilton Buchanan 1822
Body fairly compressed, the ventral surface rounded; scales small. Lateral
line low down, approximately following the ventral outline and running along

THE FISHES OF UGANDA 163
the lower part of the caudal peduncle. Cheek almost covered by the sub-orbital
bones. Mouth large, terminal and rather oblique. Lips undeveloped; circum-
oral barbels absent. Dorsal fin placed behind the pelvic fins but entirely in
advance of the anal fin.
A genus of wide distribution in Asia and Africa; one species occurs in Uganda.

Barilius niloticus (Joannis) 1835 (Fig. 27)
Description: A small species rarely exceeding 100 mm. in length.
Depth of body contained 4-5 times in standard length, length of head con-
tained 3J-41 times. Snout pointed, as long as the eye (at least in adults) which
is contained 3-4 times in the head length. Mouth large, extending to below
the centre of the eye. Barbels absent. Sub-orbital bones covering nearly the
entire cheek. 4-6 gill rakers on the lower part of the first gill-arch. Dorsal fin III,
7-8, situated almost entirely in advance of the anal fin. Lateral line with 36-40
scales; 51-61 scales between this line and the first dorsal ray; 1 or 2 scales
between the pelvic fin base and the lateral line.

FIG. 27
Barilius niloticus. (Drawn by Barbara Williams.)

Young B. niloticus are sometimes confused with Engraulicypris. However,
the species are immediately distinguishable by the position of the dorsal fin
relative to that of the anal.
Coloration: Bright silver; fins colourless. In young fishes, a dark mid-lateral
stripe may develop after death.
Habitat: The extreme inshore waters of Lake Albert, often in water only a
few inches deep (Worthington, 1929 a).
Food: Zooplankton (Worthington op. cit.).
Distribution: Uganda: Lake Albert. Elsewhere: Nile and Niger.

Engraulicypris Giinther 1893
Body strongly compressed, scales moderately large. Lateral line situated low
on the body and running along the lower part of the caudal peduncle. Dorsal fin
inserted almost entirely above the anal fin. Mouth large, terminal and rather

164 UGANDA JOURNAL
oblique; without lips or circum-oral barbels. Cheek covered by the thin sub-
orbital bones.
This genus is confined to Africa; two species occur in Uganda.

KEY TO THE SPECIES
(i) Lateral line with 36-39 scales; anal fin with 11-13 branched rays.
Known only from Lake Albert E. bredoi
(ii) Lateral line with 49-52 scales; anal fin with 14-17 branched rays.
Occurs in Lakes Victoria and Kyoga E. argenteus

Engraulicypris bredoi Poll 1945
Description: Depth of body contained 4-5 times in standard length, length of
head 3J-4 times. Snout pointed, somewhat shorter than the eye which is con-
tained 31-4 times in the head length. Mouth large and terminal, extending to
below the anterior border of the eye. 10 or 11 gill rakers on the lower part of the
first gill-arch. Dorsal fin with II-III, 7 or 8 rays, its origin above that of the anal
which has II-III, 11-13 rays. Lateral line with 35-39 scales; many scales in this
series are without pores.
Size: Rarely exceeding 45 mm.
Coloration: Intensely silver, surface of head darker; a dark mid-lateral stripe
appears after death.
Habitat: Imperfectly known, but probably restricted to the surface zone of
inshore waters.
Distribution: Endemic to Lake Albert.

Engraulicypris argenteus (Pellegrin) 1904 (Fig. 28)
Native name: Mukene (Lake Victoria).
Description: Similar to E. bredoi, except that there are 48-52 scales in the
lateral line (which is always complete), more gill rakers (16) on the first arch
and slightly more branched rays in the anal fin (14-16).

1cm.
FIG. 28
Engraulicypris argenteus. (Drawn by Barbara Williams.)

Size: Rarely exceeding 80 mm.
Coloration: Intensely silver, often with an overall nacreous sheen. After death

THE FISHES OF UGANDA 165
a fairly distinct dark mid-lateral stripe appears. Caudal fin yellow, other fins
colourless.
Habitat: The distribution of E. argenteus within Lake Victoria is poorly
known. Most records are from the surface of inshore and coastal waters, although
there are some from surface waters over great depths (Graham, 1929 and
E.A.F.R.O.). In Lake Kyoga, E. argenteus inhabits the open water away from
water-lily swamps. The species is also caught in the more turbulent regions of
the Victoria Nile.
Breeding: The information available suggests that E. argenteus spawns in the
lake and produces floating eggs (Graham op. cit.). This habit is unusual amongst
fresh-water fishes, whose eggs are generally heavier than water.
Food: Mainly planktonic Crustacea. In turn, E. argenteus is heavily preyed
upon by birds and several species of predatory fishes including some Haplo-
chromis spp., Bargrus docmac, Clarias mossambicus and Schilbe mystus.
(E.A.F.R.O. and Graham, 1929.)
Distribution: Lakes Victoria and Kyoga; the Victoria Nile.

TABLE V
Distribution in Uganda of species of the genera Labeo, Gara, Barilius and Engraulicypris
(Cyprinidae)

Edward and
Lakes Victoria Kyoga George Albert Elsewhere
Labeo horie P Murchison Nile
Labeo forskali P Agoye River, West Nile
Labeo coubie P Murchison Nile
Labeo victorianus P P Victoria Nile; major
rivers flowing into Lake
Victoria
Gara johnstoni ?P Victoria Nile
Barilius niloticus P ? Murchison Nile
Engraulicypris bredoi P
E. argenteus P P Victoria Nile

REFERENCES
Fryer, G., Greenwood, P. H., and Trewavas, E. (1955). Scale eating habits of African
cichlid fishes. Nature, 175, 1089-90.

(To be continued)

STUDIES IN UGANDA SWAMPS

By E. M. LIND, B.Sc., PH.D.(LIV.)

Sis not possible to travel far in Uganda without encountering swamps. Indeed,
o the early explorers, these great areas of water-logged vegetation which
occupy about 6 per cent of the Protectorate must have presented a formidable
obstacle. They are found in the river valleys and also occupying bays and inlets
of the lakes where, in many places, vegetation is advancing into the shallow
marginal waters.
The origin of these swamps is obscure. There is evidence that, in early Tertiary
times, the rivers flowed from a point somewhere east of the present Lake
Victoria westwards towards the Congo. But the Tertiary was a time of violent
earth movements which resulted in the formation of the Rift Valleys and quite
changed the configuration of East Africa. As a result of an upthrust of the land
at this time somewhere to the east of the Western Rift Valley, a new watershed
was formed from which part of the original river system continued to fall to the
west, while in the remainder of the valley the flow was reversed towards the
shallow basins now occupied by Lakes Victoria and Kyoga. The slow flow of
the rivers down these much levelled waterways would allow the accumulation
of silt along the banks upon which the swamp plants could establish themselves,
impeding the drainage still further by their tangle of stems and underground
rhizomes and building up organic mud on the valley floor. The flooding of the
valleys and lake-inlets which occurred during the Pluvial periods of the Pleisto-
cene no doubt contributed to the spread of the swamps which are now such a
conspicuous feature of the Uganda landscape.
In some places near the towns swamps have been drained and used for
cultivation. In other places aquatic vegetation is encroaching on shallow lakes
and dams to form new swamps. Although they occupy such a large part of the
Protectorate these water-logged areas have hitherto attracted little attention,
being considered useless except to provide a few fish and building material in a
country where good agricultural land was plentiful. Apart from a paper by
Eggeling (1935) little attempt has been made to study the ecology of the swamp
vegetation or the relation of the various plant communities to the underlying
muds. Recently, however, the possibility of draining the swamps and using them
for crop production has been engaging the attention of agriculturists, and in
Ruanda-Urundi this has already been done, on a large scale and with some
success.
It was felt that some knowledge of the plant communities of swamps was
essential if such developments were to take place in Uganda. The account of
swamp vegetation which follows is based on studies made in the neighbourhood
of Kampala and Lake Nabugabo and also in an area at a higher altitude in the
Kashambya Valley north of Kabale. Some observations are also included on the
very interesting grass swamps in Teso District.
166

STUDIES IN UGANDA SWAMPS 167
THE PLANT COMMUNITIES
The most widely distributed of the swamp plants is Papyrus (Cyperus
papyrus). If it does not cover the whole valley, it always occurs along the edge
where it may be quite dry. It can stand, and probably requires, considerable
flooding and its existence at the margins no doubt depends on periodic inunda-
tion. Round the lake shores in sheltered bays, it sends out its long floating
rhizomes into the water but is often separated from the open water by a belt of
grass. In the swamps it spreads by large underground rhizomes bearing many
roots and if cut or burned it will grow up again to a height of about 10 feet in
little over 4 months.
Papyrus has its own community of associated plants which, in an uncut stand
is relatively poor in species. The fern Dryopteris striata (Schumach) C. Chr. and
species of Polygonum are constant components, while other plants such as
Ipomoea spp. Mikania cordata (Burm. f) B. L. Robinson and Melanthera
scandens (Schumach and Thonn) Brennan, climb up its tall stems. Papyrus has
an altitudinal limit of about 6,500 ft.
Another very characteristic swamp plant is Miscanthidium violaceum,
K. Schum (Robyns) a tall grass with handsome purple plumes. This is often
mixed with papyrus, but frequently it forms pure stands many acres in extent,
as may be observed in the swamps occupying some of the bays and arms of Lake
Victoria and round Lake Nabugabo and the small lakes in its vicinity. A good
example is seen in the bay behind Entebbe airport where there is a fringe of
papyrus along the water edge, but the remainder of the swamp is dominated by
Miscanthidium with another grass, Loudetia phragmatoides (Peter) C. E. Hub.
with brown flower heads in the drier parts.
Except where the growth is old and very dense, Miscanthidium has a larger
variety of associated species than papyrus, some growing on the mud and others
in humus which accumulates on the top of the grass tussocks. Among the
common species is Dryopteris thelypteris (L) a rather more delicate species than
D. striata of the papyrus community, the clambering grass Leersia hexandra,
Schwartz, the showy purple Dissotis incana, E. Mey, and many species of
Cyperus, Pycreus and other members of the Cyperaceae. Another interesting
plant which seems to be associated with Miscanthidium swamps is the bog-moss
Sphagnum. It appears first in stagnant pools between the tussocks at the land-
ward edge and spreads inwards until in some places it forms a carpet through
which the grass plants project. This condition is well seen on the eastern shore of
Lake Nabugabo and in parts of the Kashambya Valley near Kabale. It is recog-
nized that Sphagnum requires an acid environment, and that once established,
it increases the acidity of the habitat. It would seem that it first gains a footing
in stagnant pools which receive acid drainage from the adjoining grassland and
are not subject to flooding by the lake water. In the swamp between Lake
Nabugabo and Lake Victoria, however, the Sphagnum comes almost to the
water edge. The marginal region where the bog moss first makes its appearance
is also remarkable for the presence of several plants not found in the central
part of the swamp. Among them are Lycopodium carolinianum L, and
Mesanthemum radicans (Berth) Koern. A characteristic group of plants growing

168 UGANDA JOURNAL
in the Sphagnum carpet includes the African Sundew, Drosera madascariensis
D.C.
A frequent constituent of Uganda swamps is the 'bullrush', Typha australis,
Schum. and Thonn. Though it often occurs mixed with papyrus and Miscanthi-
dium it sometimes dominates the vegetation, as in the upper part of the
Kashambya Valley. Its distribution in this valley suggests that its growth is
favoured by flooding and heavy silting. It resembles papyrus in the poverty of its
associated species.
Cladium jamaicense, Crantz, a near relative of Cladium mariscus (L) Pohl,
the characteristic plant of the English Fen country, forms scattered patches in
the reed swamp round Lake Victoria, but in the Kashyamba Valley it dominates
the vegetation for a distance of some miles. Pure Cladium swamps have not
been observed elsewhere.
Phragmites mauritianus, Kunth, a plant very similar to the English Phrag-
mites communis, Trin, occurs in many swampy areas especially round the lake
edges, but it seems to be most prevalent in western Uganda where it dominates
some of the swamps in the Fort Portal area and is found round Lake Bunyonyi
and many of the crater lakes.
Cyperus latifolius, Poir, is one of the larger members of Cyperaceae. With its
rather coarse leaves and large brown inflorescences, it is characteristic of swamp
edges which have been disturbed or are subject to silting.
The plant communities so far outlined are those found in swamps in the
region round Lakes Victoria and Nabugabo or in the Kashambya Valley in
Kigezi. Readers familiar with the valleys of the Lake Kyoga system round
Iganga or with those in the neighbourhood of Soroti will recognize that here
there is quite a different kind of swamp, where papyrus and Miscanthidium are
rare or absent and the vegetation is dominated by grasses, particularly Echi-
nochloa spp. and Hyparrhenia spp. This type of swamp is seen at its best in
Teso where, in the wet season, there is deep flooding which makes collecting
difficult while in the dry season the grass covering affords good pasturage for
cattle.
Lastly, a word must be said about the swamps occupying the valleys in the
higher parts of the Protectorate, above 6,500 ft. Here the dominant vegetation
is composed of grasses with members of the Cyperaceae which now include the
genus Carex, a predominantly temperate genus. Associated with these as the
altitude increases are plants of mountain habitats such as Helichrysum spp.,
Alchemilla spp., Lobelia and Senecio spp. together with Sphagnum and other
mosses.

THE SWAMP MUDS
Enough has now been said to indicate that the area of waterlogged mud
round the lakes and in the valleys may bear a variety of plant community and
it will be well to enquire what are the factors which control the distribution of
these types of vegetation. Attention turns first to the nature of the mud upon
which the plants grow and from which they acquire much of their nutriment.

STUDIES IN UGANDA SWAMPS 169
In many of the lake edge swamps, and in some valleys, the plants do not rest
on mud but form vegetation mats floating on a considerable depth of water. The
mat of rhizomes gives under one's weight but is sufficiently compact to bear it,
and among the rhizomes there collects silt and mud derived from decaying plant
material. In the mud are numerous small roots some of which must penetrate
into the water beneath. Some borings were made with a peat augur through a
floating mat of this description in the Kashambya Valley, and these revealed that
at the bottom of the water there had collected a considerable depth of detritus
peat containing vegetative remains, seeds and pollen grains.
As far as evidence goes to date, in most of the valley swamps round Kampala
the vegetation rests on quite a shallow layer of peat, often less than a foot in
depth, and overlying clay. But in parts of Kigezi there are known to be over
50 feet of peat under the papyrus covering.
The word 'peat' is used here to denote a mass of partly decomposed organic
matter containing recognizable plant remains and often mixed with silt, but not
having the definite structure of a bog peat of temperate climates. Very little is
yet known about the processes of organic decomposition under water-logged
conditions in the Tropics.
In order to investigate the chemical composition of the muds, samples were
collected by hand from among the roots at a depth of about 6 inches in various
plant communities. These were analysed spectrographically and the results are
given in Table I. It will be noted that, except in the Typha community and the
Teso swamp, calcium was the only exchangeable base revealed. Iron was not
measured but the colour of the ash gives some indication of its distribution.
Though insufficient data are yet available to furnish conclusive evidence, it is
clear that there is some correlation between the plant covering and certain
features of the underlying muds. Papyrus can grow under a wide variety of con-
ditions but with a pH1 rarely less than 5-5 or more than 6-4. Miscanthidium is
associated with a lower exchangeable calcium and a pH between 4-6 and 5-6.
The ash is usually of a pale pink colour compared with the deeper pink of
papyrus muds. Phragmites and Cladium both grow on mud of higher pH (6-4 to
6-9) and higher calcium content and the ash is yellow to brown. Manganese
appeared in the Typha mud from the head of the Kashambya Valley but was not
present in mud from the Typha region lower downstream. Grass swamps in Teso
reveal quite a different kind of substratum, where sodium, magnesium, potassium
and manganese are present as well as calcium. The mud is grey and heavy and
sticky in contrast to the soft black muds of the other communities.

THE KASHAMBYA VALLEY
This long narrow valley lies north of Kabale in Kigezi at an altitude of about
5,500 ft. A small stream enters the valley at its southern end and flows through
swamp for about 15 miles emerging as the Rushoma river to fall over the steep
shelf of the Kisizi Falls. Although there is common drainage throughout the
1 pH is a term used to express acidity, pH7 being neutral. The lower the pH the greater
the acidity, and the higher the pH the greater the alkalinity.

UGANDA JOURNAL

SSYMBOLS
. :. Cyperus popyus
,A Typho sp.
-- Misconthidium violoceum
11111 C/odium jomaicense
.xxx
xxx Cyperus latifolius
C Cultivated

Core B.
LAKE
C SCALE. CALE .ie

Core C.

ore D.
reCt

KISZI KASHAMBYA
FALLS C ROAD
FIG. 1
Upper Part of Kashambya Valley.

6*7

6.6

655

4-5

STUDIES IN UGANDA SWAMPS 171
valley the vegetation changes abruptly at intervals, communities of Typha,
papyrus, Miscanthidium and Cladium succeeding each other. Soil samples were
analysed from each community. The results are shown in Table II and the
measurements of pH and conductivity are also indicated on the map (Fig. 1),
alongside the community in which they were taken. The pH figures for samples

A B
Misconthidium Cladium

C
Papyrus

vvv Rootss,hizomes
= Sub-surace water
=SE Surface water

C1
Papyrus

we Sphagnum
::: Clay
IIIII Peatydepios

FIG. 2
Peat cores from Kashambya Valley.

collected on 3 March 1953 were obtained after one week's storage in airtight
jars. The others were measured in the field.
The explanation of these sudden changes of vegetation with related variation
in mud composition must be sought in the history of this valley. Cores taken
through the plant covering in various places revealed varying depths of water
under a floating plant mat (Fig. 2) and in the Cladium zone there is still a small
lake. The presence of lake terraces in parts of the valley lends support to the

D
Typha

VVI-
VVV

tj

5-

8-
9'-i

172 UGANDA JOURNAL
theory that the valley was once occupied by a chain of small lakes separated
from each other by land-bridges formed by land-slides or by silt brought down
the now dry tributary valleys. Each of these lakes has been filled in by encroach-
ing vegetation dominated by a particular plant. This can be explained if one
assumes that the nature of the land drainage into the various lakes differed so
that in one case papyrus was favoured, and in another Cladium or Miscanthi-
dium. If, as is likely, the proportions of the inorganic constituents of the plant
body vary in different swamp plants, the return of these substances to the soil
during decomposition would help to perpetuate the differences between the
muds of the different plant communities.

DEVELOPMENT OF VEGETATION
A study of bogs in temperate countries reveals a gradual succession of plant
communities as the habitat changes. The change may be brought about by the
accumulation of mud bringing the surface above the level of flooding so that it
receives its water only by precipitation; or the cutting back of drainage channels
may cause a drying out of the peat surface. Various bog plants succeed each
other and the climax of the succession is usually woodland. This process is well
seen in the development of alder and oak woodland on fen.
Although swamp forest is not uncommon in Uganda especially along the
swamp edges, it is remarkable what a large proportion of the water-logged areas
have maintained their covering of papyrus or Miscanthidium even when there
is forest at the margin. Most trees of swamp forest seem able to stand consider-
able flooding and their saplings are often found in the wettest parts. So the lack
of forest is probably not due to periodic inundation. The explanation is rather
to be found in the fires which frequently sweep across old papyrus in the dry
season and destroy the young trees. These fires may have a more far reaching
effect, for they temporarily open up the canopy of the papyrus or Miscanthidium
allowing other plants to gain admittance at a time when the soil is enriched by
the ash of the burnt plant material. The drying out of the surface and resultant
oxidization probably results in increasing acidity of the soil and this seems to be
the most likely explanation of the replacement of papyrus by Miscanthidium in
many areas. It will be remembered that the latter plant is favoured by an acid
environment and its prevalence in the district round Lake Nabugabo may be
due to the fact that land drainage into the lakes is from Loudetia kagarense
grassland of high acidity. It is where this drainage reaches the swamp that
Sphagnum first makes its appearance. Where the bog moss forms a complete
carpet, Miscanthidium makes poor growth and appears to be overwhelmed,
while trees, especially Syzigium cordatum, often occur in this habitat. It is inter-
esting to speculate on the ultimate fate of these tropical Sphagnum bogs. It is
possible that they are the forerunners of Syzigium forests of the type seen to
cover the valley in the neighbourhood of Mile 12 on the road from Kabale to the
Kanaba Pass. So far there is no evidence of Sphagnum underlying this forest
but it may not be preserved under tropical conditions.

[Photo: E. M. Lint
FIG. 3
L. Victoria. Papyrus advancing into the water with Miscanthidium behind.

FIG. 4
Papyrus and Miscanthidium.

IPnto: E. M. Lind

STUDIES IN UGANDA SWAMPS 173
This account of the Uganda swamps and of some factors in their ecology is of
necessity somewhat tentative. There is no doubt that swamps serve as important
reservoirs for nutrients draining from the surrounding land as well as for sub-
stances released by the decomposition of the vegetation, and for plants adapted
to this kind of environment they are exceedingly fertile.
Whether the muds would maintain their fertility if drained is open to question,
it is known that in some cases drainage resulted in greatly increased acidity
(Chenery 1952). There are many problems which merit further investigation,
particularly in relation to the structure, physiology and nutritional requirements
of swamp plants and the processes by which the plant material is returned to the
soil on the death of the vegetation. It is hoped that answers may be found to
some of these questions before the natural productivity of the swamp areas is
seriously interfered with by drainage and cultivation.
The author wishes to acknowledge with gratitude the help of Mrs. Edna
Norman in recording vegetation and making peat transects in the Kashambya
Valley and for directing her attention to this interesting area. She is also
indebted to Dr. Chenery and his staff at the Agricultural Research Station,
Kawanda, for the spectrographic analysis of the muds, to Professor L. C. Beadle
of the Department of Biology at Makerere College for his constant interest and
encouragement, and to the staff of the herbarium at Kew and at the East African
Herbarium, Nairobi, for their help in identifying the swamp plants.

TABLE I
Mud Data for Swamps other than Kashambya.

Dominant Locality pH Exch. Other Ash Colour
plant Ca. bases
Papyrus .. Mile 13, Masaka Road 5-5 0-48 None 69-3 Deeppink
Papyrus .. Mile 6, Fort Portal Road 5-7 0-84 None 50-3 Deep pink
Papyrus .. Kawanda 5-7 6-1 None 81-0 Pale pink
Papyrus .. Namulonge 5-9 12-6 None 74-7 Pale pink
Papyrus .. Nawanga, Iganga 5-9 20-16 None 77-0 Deeppink
Papyrus .. Namanve (lake arm) 5-0 21-0 None 39-5 Palepink
Miscanthidium L. Nabugabo 4-6 2-7 None 54-1 Pale pink
(some Sphagnum)
Miscanthidium L. Kyanja, Nabugabo 4-8 2-3 None 40-3 Pink
Miscanthidium Mile 6, Fort Portal Road 5-6 0-3 None 74-6 Pale pink
Phragmites .. Nyakumea, Fort Portal 6-9 54-0 None 12-4 Brown
Echinochloa .. Nawanga, Iganga 6-1 20-2 None 82-3 Deep pink
Echinochloa .. Mile 12, Soroti-Lira Road 6-6 2-9 Mg. 108 21-9 Deeppink
K. 0-29
Mn. -04
Na. -07

Figures are expressed as milli-equivalents per cent of dry mud.

174 UGANDA JOURNAL
TABLE II
Data for Kashambya Valley Swamps.

WATER MUD
Con- Calcium
Date pH duct- pH Ash Colour Other
ivity Exch. Sol. of Ash bases

Typhasp. 9/53 6-7 320 6-3 12-8 25-0 Deepred Mn. 0-69
Typha and
Papyrus ,, 300 6-7 13-4 12-3 Redbrown Mn. 0-73
Papyrus ,, 6-6 295 6-4 5-2 0-3 44-8 Deep pink None
Mixed swamp
Lutobo Road ,, 655 250 -
Miscanthidium
and Sphagnum ,, 45 98 -
Miscanthidium
and Sphagnum 3/53 4-3 20 1-8 5-9 Palepink None
Cladium 9/53 6-3 270 6-4 -
Cladium 3/53 6-9 67-0 2-2 16-7 Lightbrown None
Typha near
Cladium 3/53 7-1 23-0 2-1 65-5 Deep red None
Papyrus
Kashambya Rd. 9/53 6-5 320 -
Papyrus
100yds. S. 3/53 6-6 67-0 1-3 21-9 Pinkbrown None
Above falls 9/53 6-5 190 -

Figures are expressed as milli-equivalents per cent of dry mud.
A dash in the column indicates that no test was made.

TABLE III
List of commoner plants occurring in the main swamp communities.
x x x occurring in at least 50 per cent of quadrats.
x x occurring in at least 25 per cent of the quadrats.
x occurring in less than 25 per cent of the quadrats.
The figures after the names refer to the number of the specimen as identified by Kew or
the East African Herbarium.

Acriulus madagascariensis. Ridl. Lind 134 .. .. .. -
Bulbostylis zeyheri (Boeck) C.B.C1. E.A.H.8 .. .. .. x -
Carex abyssinica. Chiov. Norman 185 .. .. .. .. -- x -
Cladium jamaicense, Crantz. Lind 143 ... .. x x xxx x
Cyperus denudatus L.f. Lind 132 .. ... x xx xxx -
C. distans L. Lind 157 .. .. .. .. .. .. -
C. latifolius Poir. Lind 130 .. .. .. .. x x x x xxx
C. nudicaulis Poir. Norman 143 .. .. .. .. x x -
C. papyrus L. ... .... ..xxx xx x x
Fuirena pubescens (Lam.) Kunth. Lind 138 .. .. x -
F. umbellata. Rotb. E.A.H.6 .. .. x -
Pycreus lanceus. (Thunb.) Turrill. Lind 129 .. .. .. x -

STUDIES IN UGANDA SWAMPS 175

E

P. mundtii. Nees var. distichophyllus (Steud) Lind 137 .. x x -
Typha australis Schumm and Thonn. .. .. .. .. x x xx x
Echinochloa pyramidalis (Lam.) Hitch and Chase. E.A.28 .. x -
Hyparrhenia bracteata (Hunb. et Bonpl. es Willd.) Lind 164 .. x -
Leersia hexandra (Swartz) Stapf. .. .. .. .. .. xx xxx -
Loudetia phragmitoides (Peter) C. E. Hub. Lind 163 .. .. -
Miscanthiduim violaceum (K. Schum) Robyns Norman 78 .. x x x x x -
Panicum parvifolium Lam. E.A.68 .. .. .. x x -
P. trichocladum Hack. ex Engl. E.A.4. .. .. .. x x -
Phragmites mauritianus. Kunth... .. .. .. ..- xxx
Mesanthemum radicans (Benth) Koern Norman 54 .. .. -
Commelina sp. .. .. .. .. .. .. x x -
Floscopa rivularis (A. Rich.) C.B. Cl. Norman 37 .. .. x -
Eulophia porphyroglossus Bollus. .. .. .. .. x x
Polygonum pulchrum Blume E.A. 72 .. .. x -
P. strigosum R. Br. E.A.32 .. .. .. .. x x xx -
P. salicifolium Brouss ex Willd. .. .. .. .. .. x x x xx x
Cissampelos mucronata A. Rich .. .. .. x x x -
Stephania abyssinica Rich. .. .. .. .. .. .. x -
Oenanthe palustris (Chiov.) Norman E.A.22 .. .. xx -
Peucedanum scottianum Engl. Lind 161. Norman 97 .. .. x -
Rorippa microphyllum (Boenn. ex Reichb.) Hylander E.A.63 .. x
Sium thunbergii D.C. E.A.64 .. .. .. .. x
Crassocephalum montuosum (S. Moore) Milne. Redh. Lind 56 x x XX
Ethulia conyzoides L. .. .. .. .. .. x -
Helichrysum fruticosum (Forsk.) Vatke. Lind 19 .. .. -
Melanthera scandens (Schumach. et Thonn) Brennan .. .. xx x x
Mikania cordata (Burmf.) B.L. Robinson. Lind 84 .. .. x x x x x
Desmodium salicifolium (Poir) D.C. E.A.25 .. .. .. x x
Vigna sp. .. .. .. .. .. .. .. .. -
Dissotis incana (E. Mey) Triana E.A.10495 .. .. xxx -
D. macrocarpaGilg.E.A.10534 .. .. .. .. .. x -
Oldenlandia goreensis (D.C.) Summerh. Norman 114 .. .. xxx xx -
Otomeria elatior (A. Rich.) Verdcourt E.A.10506 .. .. x -
Pentodon pentander (Schumach. et Thonn) Vatke. Lind 110 .. x -
Hyptis brevipes Poir.. .. .. .. .. .. x x -
Mentha aquatica L... .. .. .. .. .. x x xx x
Pycnostachys caerulea Hook E.A.54 .. .. .. .. -
Rubus sp. .. .. .. .. .. .. .. .. x x x -
Impatiens eminii Warb. .. .. .. .. .. .. x x
I. procidoides Warb. .. .. .. .. .. x -
Sopubia simplex Hochst. .. .. .. .. x -
Torenia thouarsii (Cham. et Schlecht) Kuntze Norman 2 .. x x -
Ipomoea spp. .. .. .. .. .. .. .. -
Hibiscus diversifolius. Jacq. E.A.10498 .. .. x X -
Melochia mollis L. (K. Schum.) Hutch. and J. M. Dalz x x -
Sauvasegia erecta L. Norman 19 .. .. .. .. .- x -
Limnophyton obtusifolium Miq. .. .. .. ..- -
Dryopteris striata (Schumach) C. Chr. .. .. .xxx x x x
D. thelypteris (L.) A. Gray var. squamigera (Schlecht) C. Chr... x xxxx x xx x
Osmunda regalis L. .. .. .. .. .. x x -
Lygodium microphyllum R.Br. .. .. .. .. .. x x -
Sphagnum spp. .. .. .. .. .. .. .. x -
Number of quadrats examined .. .. .. 27 37 13 5

176 UGANDA JOURNAL
REFERENCES
Chenery, E. M. (1952). A digest of the chemistry of dead swamps in Kigezi district,
S.W. Uganda. Ann. Report Hydrological Survey Dept., Entebbe, Uganda.
Eggeling, W. J. (1935). The Vegetation of Namanve Swamp, Uganda Journal of
Ecology, 23, 422-435.

HISTORIC SITES IN ANKOLE
By H. F. MORRIS

ANKOLE is fortunate in still possessing a considerable wealth of historic sites
with visible remains in the form of trees or earthworks. This is largely due
to the fact that there has been little pressure on agricultural land and that the
most important part of the country, historically, is now largely unpopulated.
There is, however, an evergrowing danger of the knowledge and existence of
these sites being lost for ever for the following reasons:
1. The older generation which still venerated these sites, and possessed
traditional knowledge concerning them, is all but extinct, whilst a generation
which appreciates such things has not yet appeared.
2. The expansion of agriculture is engulfing many historic places.
3. The danger to ancient trees from the annual bush burning.
4. The systematic bush clearance in tsetse reclamation.
The visible remains at places of historic interest are normally trees. At the
sites of royal kraals will be found the remains of bark-cloth trees which formed
the enclosure together with earthworks. Some bark-cloth trees at which cere-
monies were performed also survive. Sacred places, such as enclosures for
Bagyendanwa (the royal drum) and the spirit huts of the Bagabe were often of
finger cactus (which was reserved for royal and sacred use). The royal burial
grounds of the Bahinda surrounded by candelabra euphorbia can still be traced
at Ishanje, Kabaigarire and elsewhere. Burial and other sacred places can
usually be identified by the presence of a creeping plant, emigorora, which
seems invariably to have been planted at such sites. In Igara, the dracaena
(mugorogoro) replaces the candelabra euphorbia at the burial grounds. In the
Barisa kingdoms of Buhweju and Buzimba, bark-cloth trees were planted in the
centre of the royal kraals and at the grave of each Mukama (for here, unlike
Ankole proper, the Bakama were buried underground). In Buhweju a fine series
of these trees, each connected with a former Mukama, still exists at which cere-
monies were performed periodically, though one has recently been badly
damaged by bush fire.
Despite the perishable nature of such relics, preservation should be possible
and worthwhile. Once it is realized that an interest is being taken in such things,
then something will have been done to counteract the feeling that they are
beneath the notice of an educated Christian.
The Protectorate Government, though accepting responsibility for the com-
memoration of sites associated with the early explorers and administrators, does
not, normally, concern itself with those of tribal significance. The Ankole Local
Government, however, has recently included in its estimates a small sum for
"the preservation of historic sites and relics" intended, in the first place, to
ensure the adequate preservation of the most valuable and interesting regalia of
the sub-kingdoms which, certainly in the case of Buzimba, have been in danger
of mouldering away through neglect. A certain extension might be possible to

178 UGANDA JOURNAL
include perhaps the paying of keepers of the more important sites of tribal
importance, but this cannot be a complete answer to the problem.
The problem can really only be tackled, first by a complete recording of all
existing historic sites in the district, and then by instilling into the people the
need to preserve, where practicable, what remains at the sites, preventing
ignorant vandalism and unnecessary damage by fire. As far as the Tsetse Depart-
ment officers are concerned, they are very ready to co-operate provided they are
given the necessary information. Wholesale clearance is seldom carried out and
the sparing of certain specified sites does not at all interfere with their task. It is.
unfortunate that some years ago the kagondo (sacred kraal site) of Bagyendanwa
at Mabare was swept away in such a clearance. Had the Department known
of its existence and exact locality, it could easily have been spared.
The following is a list of the more important historic sites known to me. In
most cases I have visited the site and checked that visible remains exist. The map
shows the approximate location of each. There are, however, very many other
sites of interest in the district which merit recording.

ISINGIRO CoUNTY
1. Ishanje. A sacred forest to which the bodies of the Bagabe were brought..
A circle of nkukuru (candelabra euphorbia) can still be traced about 20 yards
from the lake shore.
2. Kabaigarire. A sacred forest to which the bodies of the Baigarire were
brought. There are three ngyemeko (small washing vessels) said to have belonged
to Kiboga (wife of the Mugabe Ntare).
3. Kabahinda. A sacred forest for the bodies of the Bahinda.
4. Kabainyiginya. A sacred forest for the bodies of the Bainyiginya.
5. Kagondo. Close to the. Rukinga landing is a circle of miyenje (finger
cactus) marking the site where the keeper of Ishanje lived and where the ndaro
(spirit huts) of the Bagabe were.
6. Kagarama Hill. This was a favourite site for the early Bagabe to have
their ndembo (palaces). There are various sites of interest in the area.
(i) The kinabiro is a shallow pool where the Bagabe washed and there is a
mutoma (bark-cloth tree) nearby under which the Mugabe is said to have
sat. There is another mutoma said to have been the Mwigarire's.
(ii) In Mr. Rwaitirimba's mailo there is a huge mutoma in the middle of a
circle of earthworks which is said to have been the birthplace of the
Mugabe Rumongye. About 20 yards away is the remains of a kraal site,
though the mitoma, which is said to have been Gasyonga I's, have been
cut down.
(iii) Rwoga where Bagyendanwa was brought for washing.
7. Munywanyanji. Here the bodies of the Bagabe were washed on the way
to Ishanje. There is a large mushasha tree.
8. Omukihonoka kya njeru. Here Ntare Kitabanyoro defeated and killed the
Munyoro general Burungi-Buhita-Iguru whose blood is said to have prevented
the grass from growing.
9. Nyamitsindo. Here chalk is obtained for Bagyendanwa.

ANKLE DISTRICT

Scale of Miles *49

MILES 10 0 10 20 ILES---

-48 *47

9 -- -*4----- I
Sa--------

,BUNYARU UR '55
BUN GR 2U, 2 54 NYABUSHOZI
*44

.404

*37

S30 *51

I -2 a Mbarara _
--- -- --- -- -- --- -
-19 *13*12
3S .36 -*24*22 .*1 1 ,0
---- *17 "

/ *6 *7 4 4

*16
Sf A S IN G IR O

.25 20

:24
*26
'23

FIG. 1
Historic sites in Ankole.

HISTORIC SITES IN ANKOLE 179
10. Rwemigoshora. Here sticks for placing under Bagyendanwa are
obtained.
11. Enju y'enaku. There are three caves in which Ntare Kitabanyoro took
refuge when pursued by the Banyoro. The ledge on which he placed his pipe is
still pointed out.
12. Mabare. This was the favourite home of Bagyendanwa. Nothing now
remains on the site, the miyenje trees having been cut down in tsetse clearance.
13. Kabukondore. These are large flat rocks where the warriors brought
their plunder to the Mugabe. Nearby is Rubeho where offerings were made to
Bagyendanwa.
14. Akenjeru. There are the remains of a circle of mitoma on the slopes of
Bihunya where the sorcerers consulted the oracles.
15. Ijumuriro. Here the coronation ceremony of turning Bagyendanwa the
right way up, the period of mourning being over, was carried out. Nothing now
remains on the hill.
16. Omutoma rwengabo. Ntare Kitabanyoro planted a mutoma here and
sacrificed the calf of the famous ngabo cow when he left Muzaire, where he had
hidden from the Banyoro. A mukunyu tree now grows where the mutoma once
was. The mutoma died in 1928.
17. Ihunga. Here Ntare sheltered from the rain in a cave on leaving
Rwangabo on the way to Kagarama.
18. Engyereka. This is a large rock about 30 feet high with foot-holds up it.
The rock is associated with the Bacwezi and is held in great awe.
19. Rukoma. The site of the rurembo (palace) of Gasyonga I. Earthworks
and several mitoma exist.
20. Kantsyore Island. Here Ntare Kitabanyoro took refuge from the
Banyoro.

RWAMPARA COUNTY
21. Itabe. Ancient mitoma mark the site of two kraals said to be those of
Wamara and his mother. This is the reputed birthplace of Ruhinda.
22. Kichwamba. The clay used for smearing Bagyendanwa is obtained from
here.

KAJARA COUNTY
23 and 24. Matongo (kraal sites) of Kitami.
25. Ihunga Nyondo. Itongo (kraal site) of the Mwenekihondwa Rubunda.
Here is the stone he used for kishoro (the mweso of the Baganda).
26. Nyabugando. The mutoma here is said to mark the birthplace of Kahaya
Rutindangezi, Mukama of Mpororo.

IGARA COUNTY
27. Bumbaire. Graves of the Bakama of Igara are marked by migorogoro
(dracaenae).
28. Buturo. The graves of the Baigarire.

180 UGANDA JOURNAL
29. Bushenyi. The rurembo of the last Mukama Musinga is in the banana-
garden near the rest-camp. There are the remains of earthworks and the grave
of the Mukama's sister, Kitunga.
30. The rurembo of Rutondo. There is a large mutoma and earthworks.
31. Lake Katunga. Sacrifices used to be made in this lake.
32. Kitembe. Here Kihoza, the royal drum of Igara, used to be kept.
33. Irembezi. Mitoma mark the itongo of Rwihura.

SHEMA COUNTY
34. Kitagata. Here there are the famous hot springs.
35. Musale Shema. There are the following sites in this gombolola:
(i) Mutanoga. A mutoma marks the itongo of Mwira of Kigiro.
(ii) Kigabagaba. A mutoma marks the itongo of Bwesharire of Kigiro.
(iii) Kashoma. Earthworks mark the itongo of Rwikina, Mukama of Igara.
(iv) Kitembe. Itongo of Katwahwa of Kigiro.
(v) Kinyuma. Itongo of Buruguriya.
36. Kajare. The itongo of the Mwenerukari Bukondore.
37. Sabawali Shema. There are the following sites in this gombolola:
(i) Kanyamutambara. A mutoma marks the itongo of Mweneishemurari
Katairerayo.
(ii) Kitoha. A mutoma marks the itongo of Muhunga.
(iii) Katagata. Migorogoro marks the site where sacrifices were made to
Mureju.
(iv) Katoma. A mutoma and a well mark the site of the itongo of the Muhinda
Tujungye.

BUNYARUGURU COUNTY
38. Kazinga. Nkukuru mark the itongo of Kaihura.
39. Kichwamba. The remains of mitoma mark the site of another itongo of
Kaihura.

BuHwEU COUNTY
40. Ntobora. A large mutoma marks the grave of the Mukama Karamagi II.
Nearby is the site of the kraal of the Mugabe Ntare which he had when he fled
from Mukwenda.
41. Kitoha. A large mutoma and migorora mark the grave of the Mukama
Rusharabaga.
42. Butare Roman Catholic Mission Grounds. A large mutoma marks the
site of Ndagara's rurembo.
43. Kyamutara. A mutoma and migorora mark the grave of Kashoma III.
44. Kikondera. A mutoma marks the grave of Ndagara.
45. Kyamujumi. A mutoma and migorora mark the grave of an unknown
Mukama.

MITOMA COUNTY
46. Kyakahoza, Bwenda. Mitoma mark the graves of the Buzimba Bakama
Nyakairu, Mugarura and Mukindo.

HISTORIC SITES IN ANKOLE 181
47. Kijongo. Here is the pool in which the Bagabe bathed after their corona-
tion. There are also mitoma and migorora.
48. Ibanda Hill. On the top is the site of the kigabiro (place for sacrifices) of
the Mugabe's witches of Ibanda.
49. Mugoye. This was the scene of the final battle in the civil war between
Ntare and Mukwenda. Mukwenda's grave is nearby and also the site of Ntare's
camp.

KASHARI COUNTY
50. Kaigoshora. Some years ago a building was erected over the grave of
Ntare and in it are kept certain of his relics such as his bow and his knife.
51. Rukiri. The site of a rurembo of Ntare marked by mitoma.
52. Omukirunganwa. The site of a rurembo of Ntare marked by mitoma.
53. Katete. The site of a rurembo of Ntare marked by a muyenje.

NYABUSHOZI COUNTY
54. Haragatwa. The site of another of Ntare's ndembo. There is also a
mutoma at Kitoma which was a kigabiro.
55. Byanamira. The site of Kiboga's kraal. There are mitoma and
migorora.

TWO SURVEYS OF KASILANG ERONY, TESO,
1937 AND 1953'
By P. N. WILSON2 AND J. M. WATSON3
INTRODUCTION. Tothill (1938) presented a detailed report on nineteen
surveys conducted in different areas of Uganda by members of the Department of
Agriculture. These surveys were conducted to estimate the extent of soil
deterioration and soil erosion in the Uganda Protectorate. Each survey followed
a similar plan, in which every taxpayer within a known small administrative
area was interviewed and his answers to a number of agricultural questions were
recorded. In addition, all plots cultivated by the taxpayer, together with any
plots resting at the time of the survey, were paced and their acreages ascer-
tained. In this manner much information was obtained, not only about soil
erosion, but also concerning the patterns of peasant farming in different parts of
the Protectorate.
The present work describes a resurvey of Kasilang Erony, a village in Teso
District first surveyed by Watson in 1937. Unfortunately the data of the original
survey were not published as one of the nineteen surveys placed on record by
Tothill, since they were received too late for inclusion. It is therefore necessary
to present Watson's data in the present work.
METHOD. In 1937 the survey was conducted by Watson together with five
Makerere students and subordinate field staff. In 1953 Wilson and six Makerere
students undertook the resurvey, assisted by five interpreters. Each student was
made responsible for surveying one hamlet within the Erony (see below).
In both years the survey was conducted in two parts. The first in May when
the long-rain crops were growing, and the second in September when the cotton
and short-rain crops had been planted.
All plots were measured by pacing mean-length and mean-width. The normal
pace-length of each student was calculated both before and after the survey and
individual correction factors were used to transform 'square paces' into 'square
feet'. This was thought to give a more accurate assessment of acreage than
assuming all paces to be one yard in length. It was found that the mean pace of
the African students was only 30 inches.
NOMENCLATURE. Watson used Luganda nouns for describing divisions.
1 The authors' thanks are due to the Professor of Agriculture, Makerere College, for
initiating the 1953 survey, and to the District Commissioner, Teso District, and the District
Team for their co-operation and interest, and for the loan of Ateso-speaking field-staff
who acted as interpreters. The following Makerere undergraduates formed the team for
the 1953 survey: Messrs. S. Byakika, W. Gowa, R. A. Irigiei, J. Maina, J. Mburu, and
P. Sibyetekerwa. The project was financed by the Makerere College Research Grants
Committee. The authors also wish to express their thanks to Professor Wilson for the
provision of the photographs which illustrate this article. Mrs. M. E. Wilson assisted with
the preparation of the manuscript.
Finally, the authors wish to record their thanks to the chief and peasants of Kasilang
Erony for their willing co-operation and hospitality during the two surveys.
2 Makerere College, University College of East Africa.
3 Formerly of the Department of Agriculture, Uganda.
182

TWO SURVEYS OF KASILANG ERONY, TESO 183
Ateso terminology is used in this paper; the Luganda equivalents as used by
Watson and the approximate English translations are as follows:
English Ateso Luganda
County4 Ebuku Saza
Sub-county Etem Gombolola
Parish Eitela Muluka
Village Erony Kitongole
Hamlet Ireria Ekyalo
Clan Ateker Ekika
The first four are administrative divisions recognized by Government, each
under the control of a paid Chief. The hamlet is a tribal division owing some
allegiance to an unpaid 'Clan Leader'. One clan is generally in the majority in
any one hamlet.
BOUNDARY CHANGES. In 1941 Kasilang Erony was amalgamated with the
neighboring Erony of Okimo. The 1953 survey was confined to the area sur-
veyed in 1937, known as 'Old Kasilang'. The survey, therefore, covers only two-
thirds of the area at present known as 'Kasilang Erony'.
DESCRIPTION OF AREA. Kasilang Erony is situated between long. 33* 33' and
330 36' E. and lat. 1o 32' and 1 34. N. It occupies an area of 5-8 square miles,
and straddles the Soroti-Mbale main trunk road between mile-posts 67 and 69.
INCHES OF RAINFALL

X-X 1937 RAINFALL TOTAL 59-8 INCHES)
1--- 1953 RAINFALL CTOTAL 459 INCHES)
12- --0 MEAN RAINFALL FOR PERIOD 1937-1953
II-

7 Q
-i =

J F M A M JN JY A S O N D MONTH
FIG. 1

A botanical description of Kasilang has been published elsewhere (Wilson
and Watson, 1955). There are no rainfall records taken near Kasilang, the
nearest place being Serere Experiment Station, situated 13 miles to the west.
4 A County in Uganda represents the largest administrative unit in a District. There
are generally some half dozen such Counties in a District.
I ^y /\

are generally some half dozen such Counties in a District.

184 UGANDA JOURNAL
The Serere rainfall data for the years 1937 and 1953 are presented graphically
in Fig. 1, as giving some indication of the onset of the rains at Kasilang in
these two years. The mean monthly rainfall for the period 1921-1953 is also
shown.
A map of the Erony has been prepared, showing the position of every house
and kraal. This is presented as Fig. 2.

MINOR DIVISIONS OF KASILANG. Although the Erony was the smallest
administrative unit, the land surveyed was sub-divided by the inhabitants into
nine hamlets, each associated originally with a pioneer patriarchal family unit.
There was very little evidence of a stabilized corporate life in separate hamlets.
The names of all hamlets were in everyday use, but the boundaries between
hamlets were often obscure. The position in 1953 was that there were six well-
defined hamlets, and three satellite communities which, although at one time
capable of separate identification, were no longer clearly defined.
The names of the six major hamlets were:
1. Kasilang. Derived from esilang (Ziziphus marutiana), a common thorn-
tree of the area.
2. Obiit. Derived from biityono, a 'greedy man'. Watson recorded that the
original inhabitants of this hamlet were characterized by their greediness,
preferring to eat by themselves and refusing to share their food.
3. Akoboi. Derived from ekoboi (Terminalia velutina), a common tree of
the area.
4. Akalen. Derived from akalele, a 'flood of rushing water'. (Reason
obscure.)
5. Ikwii.s Derived from akwii, 'leaves', or a tree characterized by prolific
foliage.
6. Aasi. Derived from aasi, meaning 'outside', or 'over there'.
The geographical arrangement of the hamlets is shown in Fig. 2.

AGRICULTURAL DATA6
The data presented in Table I show that about 64 per cent (1937 estimate)
and 67 per cent (1953 estimate) of the land is capable of being used for cultiva-
tion. Of this land which is deemed suitable for arable farming, 40-2 per cent was
in fact farmed in 1937 and 45-5 per cent was tilled in 1953.
These figures allow the theoretical crop: rest ratio at Kasilang to be deduced.
This ratio was 1: 1-5 in 1937 and 1: 12 in 1953. Judged by department standards
of a 3:3 crop-rest rotation, this ratio is adequate to maintain soil fertility. There
is, however, very little margin to allow for an increase in the population of the
area. Population trends will now be considered, together with other sociological
data relevant to the study.

5 This hamlet appears as Okwi on the map (Fig. 2). The information that Ikwii is the
correct spelling was received too late for the map to be altered.
6 Al the livestock were counted in the kraals or compounds of the erony. The informa-
tion concerning livestock data and animal husbandry will be published in a second article.

TWO SURVEYS OF KASILANG ERONY, TESO

o I on

S 0 A 0 x a
o o oo o *i,
KOBOI OU
o E AKALEN .-

oo AOBo AAS, e ^ S-

.y *.. M.'- FOREST

0 0 / KEY0 t
I* I 0 BI

y CYCE TM
( o / v ; SWAMP
y0y 0 0 0 V
S\ -.--.--... BOUNDARY
\o0 BETWEEN HAMLETS

V \ o 1 FOREST
BS\ v BS S00 GRANITIC ROCKS
BSo l OABI, 0 A

\ v ^ 0 KRAAL

V V y. O0 HOMESTEAD

STREAM DRAINING BS SWAMP BOUNDARY
TO L.KYOGA OF ERONY

00 I MILE

FIG. 2

186 UGANDA JOURNAL
TABLE I
Land Usage at Kasilang.

1937 1953
(a) Total acreage of Erony .. .. .. .. 3,763 3,763
(b) Total area of cultivable land .... .. .. 2,399 2,529
(c) Total area of non-cultivable land . .. .. 1,364 1,234
CULTIVABLE LAND (b)
(d) Maximum total area cultivated at any one time ..i 966 1,151
(e) Maximum total area resting at any one time .. 1,433 622
(f) Minimum total area of cultivable land not cultivated and
not resting .. .. .. .. .. .. 0 756
2,399 2,529
NON-CULTIVABLE LAND (c)
(g) Total area under forest or suitable for forest (including non-
agricultural land such as hill-tops and semi-swamps) .. 1,299 1,179
(h) Total area under paths, roads, bomas .. . .. 65 55
1,364 1,234
Figure (g) has been further subdivided:
(i) Semi-swamps .. .. .. .. .. .. .. 781 781
(j) Kasilang Hill (rocky outcrop) . .. .. 48 48
(k) Area around Kasilang hamlet of low fertility, not suitable
for agriculture .. .. .. .. .. .. 195 150
(I) Acacia forest .. .. .. .. .. .. .. 275 200
1,299 1,179

NoTE.-The big discrepancies in the 1953 and 1937 figures for (e) and (f) are due to the
fact that in 1953 land was only classified as 'resting' if it showed signs of previous cultiva-
tion. In 1937 Watson classified all land suitable for arable cultivation but not in fact
cultivated as 'resting'. Some of the land was virgin bush.

Table II shows that there has been an overall increase of 21-9 per cent in the
population between 1937 and 1953; i.e. a mean increase of 1-4 per cent per
annum. This increase is evenly divided between children and adults. The num-
ber of adults has increased by 82, the number of children by 76.
Part of the increase in the adult population must be ascribed to immigration
into Kasilang between 1937 and 1953. Of the 222 taxpayers in the Erony 166
(74-7 per cent) were already resident there in 1937, 13 (5-9 per cent) had been
there from eleven to fifteen years, 14 (6-3 per cent) had been there from six to
ten years and 29 (13-1 per cent) had been there for five years or less. Thus 25
per cent of the 1953 population immigrated to Kasilang between 1937 and 1953.
The data for Kasilang children presented in Table I (e) indicate that the
number of children was approximately half the number of adults. This suggests
that the population is not replacing itself, a misleading assumption since allow-
ance is not made for children away from home at school, college, or learning a
trade. The child-bearing data for Kasilang are set out in Table III, which shows

[Photo: Fergus Wilson
FIG. 3
Teso District. View taken from a granite outcrop showing typical vegetables
and cultivation areas near Kasilang.

FIG. 4
A ploughing team preparing for cotton.

:. JL, - . L "
^*~ l l1-.. A. -.'^ ^ .

IPhoto: Fergus Wilson
FIG. 5
Sowing cotton. Note the action of the foot covering the
seed dropped into a shallow depression.

FIG. 6
Harvesting cotton.

ill,,r : rle W,9 ln

FIG. 7
Weeding cotton.

FIG. 8
Harvesting eleusine.

[Photo: Fergus Wilson
FIG. 9
A typical grain store containing eleusine. The cover is on the left.

[Photo: Fergus Wilson
FIG. 10
A simsim drying rack. The woman is pouring out the seed from the dried
plants.

[Photo: Fergus Wilson
FIG. 11
Pounding dried cassava with a stone on a flat rock-surface.

FIG. 12
Teso cattle.

Irtoro: r erg, wI son

ISAW
'* L|(ii~~it-, ^ r ta'

[Photo: Fergus Wilson
FIG. 13
A Muhima woman treating milk vessels with smoke.

" "^ 'L e, iB.
,A? "
JA 2:j !L

FIG. 14
Catching small fish with a basket-trap.

TWO SURVEYS OF KASILANG ERONY, TESO 187
TABLE II
Population Data for Kasilang.

1937 1953
(a) Total number of Taxpayers .. .. .. .. 218 222
(b) Total number of Families .. .. .. .. 141 173
(c) Total number of Adult Males.. .. .. .. 225 258
(d) Total number of Adult Females .. 262 311
(e) Total number of Children resident in Kasilang .. .. 235 311

(f) Total Population .. .. .. .. 722 880
(g) Increase in population 1937-1953 .. .. .. .. 158
(h) Density of population per square mile .. .. .. 122-8 149-7
(i) Mean number of children per family .. .. .. .. 1-67 1-80

NOTE.-The term 'taxpayer' is here defined as an adult male, resident in Kasilang who,
having rights of possession over land, cattle or a home in Kasilang, would normally
pay African Poll Tax. It includes some people who, for temporary reasons such as sickness,
did not in fact pay Poll Tax in 1953.

that in 1953 the mean number of live children per family was
family, and not 1-80 as suggested in Table II ().

in fact 2-5 per

TABLE III
Children Born to Residents of Kasilang, 1953.

Living at the time of survey:
Children under school age
Children at school or college:
Primary .
Secondary .. .
College ..
Children of school age not at school
Children above school age but not adult
Children of Kasilang residents above school
age and of adult status
Dead children of Kasilang residents
Total
Total mean number of children per family..
Total number of living children per family

Boys

61
43
3
2
41
42
48
95
335
1-9
1-4

Girls

Total

108
56
3
2
73
69
112
172
595
3-4
2-5

In 1937, all but nine of the taxpayers earned their livelihood by peasant agri-
culture. In 1953, 42 taxpayers engaged in enterprises additional to agricultural
production. It is of great interest to note that no taxpayers were found who did
not possess and cultivate at least I acre of agricultural land. In 1953 there were
13 herdsmen, of whom 12 were Bahima, 6 headmen, 4 traders, 4 fishmongers,
3 carpenters, 3 bricklayers, 1 bicycle repairer, 1 African Local Government
runner, 1 laundry employee and 1 driver. In 1937 none of these occupations was

188 UGANDA JOURNAL
represented, but in that year there were 3 tailors, as compared with one in 1953;
2 roadmen compared with 1 in 1953; one schoolmaster compared with 2 in 1953;
one Erony chief in both years; and one eitela chief and one storekeeper in 1937
but none in 1953. The Bahima herdsmen constituted the largest group of non-
Iteso in Kasilang.
Of the 222 taxpayers, 201 (90-4 per cent) were Iteso, 12 (5-4 per cent) were
Bahima, 6 (2-7 per cent) Kumam, one Musoga, one Mutoro and one Half-caste.
An attempt has been made to estimate the number of people capable of doing
agricultural manual work at Kasilang. The figures are only approximate, and
there are doubtless discrepancies between the 1937 and 1953 methods of estimat-
ing working ability. The data in the Table IV tends to show that there has been
a significant increase in the proportion of non-workers due to (a) the larger
number of live children in 1953 and (b) the larger expectation of life of the adults
in 1953.
TABLE IV
Workers and Non-Workers.

1937 1953
(a) Non-Workers:
Children .. .. .. .. .. .. 132 169
Adults .. .. .. .. .. .. .. 74 126
Total Non-Workers .. .. .. 206 295

(b) Workers:
Children .. .. .. .. .. .. 103 142
Adults .. .. .. .. .. .. .. 413 443
TotalWorkers .. .. .. .. .. 516 585

(c) Labour Units per family and per taxpayer (counting
children as half labour unit):
Labour Units per family .. .... .. 3-3 3-0
Labour Units per taxpayer .... .. 2-1 2-3

It will be seen that there has been little change in the amount of domestic
labour available per family unit or per taxpayer during the period. About three
units are available per family as compared with approximately two units per
taxpayer.
Having considered in detail the resources of land and labour available for
agricultural enterprises, the use made of these available resources by the Kasi-
lang peasants will now be examined. No attempt has been made to consider the
factor of capital since, owing to primitive systems of land tenure and absence of
mechanization or improvement expenses, very little capital is employed by the
peasants, and the little capital that is employed (e.g. for the purchase of ploughs)
is difficult to evaluate with any degree of accuracy.

TWO SURVEYS OF KASILANG ERONY, TESO 189

CROP PRODUCTION. Table V gives details of the total acreage and the number
of individual plots of every crop grown at Kasilang in 1937 and 1953. From this
raw data, the mean plot size for each crop has been calculated, and these figures
are shown compared to the district mean-plot-size for each crop calculated by
the Department of Agriculture, obtained from the Teso District Agricultural
Annual Report, 1953.
TABLE V

Mean Plot sizes (Acres)
Acreage Number of 1953 only
Plots
1953 only (b) Teso
1937 1953 (a) Kasilang District
CROP
Cotton .. 446-0 382-0 511 0-75 0-84
Finger Millet 359-0 521-1 737 0-71 0-74
Simsim .. 107-0 116-0 203 0-57 0-48
Sorghum .. 36-1 83-2 157 0-53 0-81
Cassava .. 40-5 71-9 143 0-50 0-67
Groundnuts 74-8 59-6 90 0-66 0-82
Sweet Potatoes 43-0 57-7 189 0-31 0-38
Cow-peas .. 188-0 52-0 88 0-60 0-61
Bananas .. 19-2 5-9 20 0-30 -*
Maize .... Nil 5-8 23 0-25 0-10
Tobacco .. Nil 0-2 3 0-07 -*
TOTALS
Food Crops 867-6 973-2 1650 0-59 -
Cash Crops 446-0 382-2 514 0-75 -
Grand Total 1313-6 1355-4 2164 0-63 -

Acreage of these minor crops not calculated for Teso District.

TABLE VI
Cropping Data Per Taxpayer and Per Family.

Percentage of Acreage of Acreage of
Crop Land under Crop Crop per Taxpayer Crop per Family
1937 1953 1937 1953 1937 1953
Cotton .. 33-94 28-18 2-05 1-72 3-15 2-21
Finger Millet.. .. 27-36 3842 1-65 2-35 2-54 3-01
Simsim .. .. 8-14 8-56 0-49 0-52 0-76 0-67
Sorghum .. .. 2-75 6-14 0-16 037 026 0-48
Cassava .. .. 3.08 5-31 0-19 0-32 0-29 0-42
Groundnuts .. .. 569 4-40 0-34 0-27 0-53 0-34
Sweet Potatoes .. 3-27 4-26 0-20 0-25 0-31 0-33
Cow Peas .. .. 14-31 384 0-86 0-23 1-33 0-29
Bananas .. .. 1-46 0-44 0.09 0-03 0-14 0-03
Maize .. .. Nil 0-43 Nil 003 Nil 0-03

The changes during the period 1937-1953 are most marked in the decrease in
acreage of cow peas, from 14 per cent to 4 per cent of the total arable land. In
addition, less land was under the perennial banana crop in 1953 than 1937. More
land was sown to finger millet than in 1937.

	Front Cover
	Title Page
	Table of Contents
	Notes on Contributors
	Articles
	Notes
	Correspondence
	Obituary
	Reviews
	Index to Volume 20 (1956)
	Back Cover