|UFDC Home||myUFDC Home | Help ||
|Table of Contents|
|The murder of H. St. G. Galt|
|Excavations at Bigo, 1957|
|Notes on the Kibuga of Buganda|
|Pleistocene environments and early...|
|Excavation of a rock shelter at...|
|Emin Pasha and Bunyoro-Kitara,...|
|Research on the swamps of...|
|Lango folk-tales - An analysis|
|Richard Buchta and early photography...|
|Notes on contributors|
ALL VOLUMES CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
STANDARD VIEW MARC VIEW
|Table of Contents|
Front Cover 1
Front Cover 2
Title Page 1
Title Page 2
Table of Contents
Table of Contents 1
Table of Contents 2
The murder of H. St. G. Galt
Excavations at Bigo, 1957
Notes on the Kibuga of Buganda
Pleistocene environments and early man in Uganda
Excavation of a rock shelter at Hippo Bay, Entebbe
Emin Pasha and Bunyoro-Kitara, 1877-1889
Research on the swamps of Uganda
Lango folk-tales - An analysis
Richard Buchta and early photography in Uganda
Notes on contributors
Back Cover 1
Back Cover 2
The Uganda Journal
THE JOURNAL OF THE UGANDA SOCIETY
VOLUME 24, No. I
THE MURDER OF H. ST. G. GALT H. F. MORRIS
EXCAVATIONS AT BIGO, 1957 P. L. SHINNIE
NOTEs ON THE KIBUGA OF BGANDA P. C. W. GlrrKIND
PLEISTOCENE ENVIRONMENTS AND EARLY h N IN UGANDA
W. W. BISHOP AND NM. POSNANSKY
EXCAVATION OF A ROCK SHELTER AT HIPPO BAY, ENTEBBE
R. NM. BRACHI
EMIN PASHA AND BUNYORO-KITARA, 1877 TO 1889 A. R. DUNBAR
RESEARCH ON THE SWAMPS OF UGANDA L. C. BEADLE AND E. NI. LIN'D
LANGO FOLK-TALES-AN ANALYSIS NM. J. WRIGuT
RICHARD BUCHTA AND EARLY PHOTOGRAPHY IN UGANDA H. B. THOMAS
Lucoro and Min-kwet. An Acholi Folk-tale
PETER ORYEM AND NI. J. WRIGHT
Mohammed Bin -
Bark-cloth Hammers -
The Eastern Boundary of Uganda in 1902
- H. B. THOMAS
E. B. HADDON
-E. C. PLANNING
A. T. MATSON
The Prehistory of Southern Africa (by J. Desmond Clark)
Linant de Bellefonds: Journal dun Voyage a Mirod dans les Anndes
1821-1822 (ed. by Margaret Shinnie) J. A. BURGESS
The Marriage Customs of the Lango Tribe (Uganda) in Relation to
Canon Law (by Rev. Cyprian B. Kibangire) D. E. B. CARR
NOTES ON CONTRIBUTORS -
THE UGANDA SOCIETY
Price Shs. 15 (15s.)
THE UGANDA SOCIETY
OFFICERS FOR 1959-60
His Excellency Sir Frederick Crawford, K.C.M.O., O.B.E.
Dr. A. W. Southall
The Hon. Secretary
The Hon. Treasurer
The Hon. Editors
The Hon. Librarian
Miss L. Bown
Dr. C. Ehrlich
Mr. H. S. S. Few
Messrs. Cooper Bros. & Co.
Corresponding Secretary at Jinja:
Corresponding Secretary at Masaka
Corresponding Secretary at Mbale
Corresponding Secretary at Tororo
Mr. J. C. D. Lawrance
Mr. S. Kajubi
Mr. B. E. R. Kirwan
Mr. D. K. Marphatia
Mr. R. J. NMehta, o.B.E.
Mr. C. N. Mukuye
Mr. D. Stephens
Mr. M. Barrington Ward
Mr. J. M. Weatherby
Dr. W. Elkan
Mrs. M. M. Wallis
Mr. J. C. D. Lawrance
Dr. H. F. Morris
Miss P. Fiddes
Hon. Legal Adviser
Mr. C. L. Holcom
Mr. E. B. Cunningham
Mr. M. J. Wright
S Mr. E. Kironde
Dr. W. H. R. Lumsden
H.H. Mutesa I1, Kabaka of Buganda Sir John Milner Gray
R. A. Sir Tito Winyi Gafabusa IV, Mr. E. B. Haddon
C.B.E.. Omukama of Bunyoro Mr. H. B. Thomas. o.a.E.
Lord Twining of Tanganyika and Professor A. W. Williams
Godalming, G.C.M.G., M.B.E.
Sir A. R. Cook, C.M.O., O.B.E.
Mr. E. J. Wayland, C.B.E.
Dr. H. H. Hunter, c.B.E., LL.D.
Mr. H. Jowitt, C.M.G.
Sir H. R. Hone,
K.B.E., M.C., Q.C.
Mr. J. Sykes, O.B.E.
Mr. N. V. Brasnett
Captain C. R. S. Pitman,
C.B.E., D.S.O., M.C.
Mr. S. W. Kulubya, C.B.E.
Mr. E. A. Temple Perkins
Mr. R. A. Snoxall
Dr. K. A. Davies, C.M.O., o.B.E.
Mr. G. H. E. Hopkins, o.e.E.
Mrs. K. M. Trowell, M.B.E.
Dr. W. J. Eggeling
Dr. G. ap Griffith
Rev. Fr. F. B. Gaffney, w.F.
Professor A. W. Williams
Sir J. B. Hutchinson,
Mr. J. D. Jameson, O.B.E.
Dr. Audrey I. Richards, c.e.B.
Dr. H. C. Trowell, O.B.E.
Mr. D. K. Marphatia
Mr. M. Barrington Ward
Dr. H. F. Morris
Mr. S. W. Kulubya, c.B.E.
Mr. B. K. Mulyanti
Mr. G. P. Saben
Secretary: Mrs. M. M. Wallis
THE JOURNAL OF THE UGANDA SOCIETY
J. C. D. LAWRANCE
DR. H. F. MORRIS
(assisted in England by H. B. THOMAS, O.B.E.)
THE UGANDA SOCIETY
PRINTED IN GREAT BRITAIN BY HEADLEY BROTHERS LTD
IO1 KINGSWAY LONDON WC2 AND ASHFORD KENT
THE MURDER OF H. ST. G. GALT H. F. MORRIS
EXCAVATIONS AT BIGO, 1957 P. L. SHINNIE
NOTES ON THE KIBUGA OF BUGANDA P. C. W. GUTKIND
PLEISTOCENE ENVIRONMENTS AND EARLY MAN IN UGANDA
W. W. BISHOP AND M. POSNANSKY
EXCAVATION OF A ROCK SHELTER AT HIPPO BAY, ENTEBBE
R. M. BRACHI
EMIN PASHA AND BUNYORO-KITARA, 1877 TO 1889 -A. R. DUNBAR
RESEARCH ON THE SWAMPS OF UGANDA L. C. BEADLE AND E. M. LIND
LANGO FOLK-TALES--AN ANALYSIS M. J. WRIGHT
RICHARD BUCHTA AND EARLY PHOTOGRAPHY IN UGANDA H. B. THOMAS
Lucoro and Min-kwet. An Acholi Folk-tale
PETER ORYEM AND M. J. WRIGHT
Mohammed Biri -H. B. THOMAS
Kintu's Pipe -
Bark-cloth Hammers -
The Eastern Boundary of Uganda in 1902
- E. B. HADDON
- E. C. PLANNING
A. T. MATSON
The Prehistory of Southern Africa (by J. Desmond Clark)
Linant de Bellefonds: Journal dun Voyage d Mdroe dans les Anndes
1821-1822 (ed. by Margaret Shinnie) J. A. BURGESS
The Marriage Customs of the Lango Tribe (Uganda) in Relation to
Canon Law (by Rev. Cyprian B. Kihangire) D. E. B. CARR
NOTES ON CONTRIBUTORS
THE MURDER OF H. ST. G. GALT'
By H. F. MORRIS
ON 19 May 1905 Harry St. George Gait, the acting Sub-Commissioner of
the Western Province, was speared while sitting on the verandah of a
rest-house near Ibanda in Ankole District. Despite an exhaustive enquiry at
the time, the circumstances of the murder remained a mystery and, although
the identity of the actual murderer was established, if not completely beyond
doubt, it has generally been believed that he was the mere agent of others more
influential, and no convincing motive could be discovered. Although persons
of the highest rank in Ankole came under suspicion, although a close relation
of the Mugabe was exiled from Uganda for life, and although the people of
Ankole had to pay heavily for the crime committed within their boundaries,
no further evidence ever came to light, and the incident is as perplexing today
as it was half a century ago. To understand the circumstances surrounding
this crime, some account must be given of the years preceding the murder and
of events which may have had, or at one time were thought to have had, a
bearing upon the crime.
The period immediately following the death of the Mugabe Ntare in 1895
was for Ankole one of unrest and distress, with three claimants intriguing for
the throne and with rinderpest destroying the cattle. Out of this confusion rose
the dominant personality of Mbaguta. Determined to see his candidate, the
young Kahaya, firmly established as Mugabe with himself as his Nganzi, he
was prepared to have an administrative post set up in Ankole if this would
secure his aim, and a request for this was made to the protecting power. A
civil station was accordingly established at Mbarara in 1898.
The establishment of the Administration in Ankole assured Kahaya's posi-
tion as Mugabe and the triumph of Mbaguta over his rivals, Kahitsi and
Igumira, the Mugabe's relatives. Imperious and ambitious though he no doubt
was, Mbaguta had also a sincere belief in progress and his ability was out-
standing. Hated by the Bahinda (the royal clan) and soon to be on bad terms
with the Mugabe who, as he grew to manhood, resented his overbearing
Nganzi, Mbaguta was dependent upon the support of the Collector, who in
his turn was dependent upon Mbaguta for the co-operation of the native
authorities in the carrying out of government measures. Though a Muhima,
the head of one of the minor Bashambo clans, the Beneishemurari, Mbaguta
was, in the eyes of the Bahinda, an upstart whose family had not even
originated in Ankole proper but in Mpororo. The leader of the Bahinda,
Igumira, was also a dominant personality and one with far more appeal to
the Banyankore of his day than Mbaguta. Men remembered nostalgically his
victorious campaigns during his uncle Ntare's reign and believed that he was
not, as was officially accepted, Kahaya's first cousin but in fact his father. To
the early Collectors on the other hand, Igumira appeared as the embodiment
1 Adapted from the Uganda Government Essay Competition 1958. Based principally
on records in the Secretariat Archives, Entebbe.
2 UGANDA JOURNAL
of reaction and as an altogether baneful influence on Ankole. Had the British
Administration not been established in Ankole when it was, and with it the
authority of Kahaya and Mbaguta, it is probable that Igumira would have
become Mugabe; as it was he had to content himself, albeit with ill grace, with
his chieftainship of Shema and part of Kashari.
During the first two and a half years of the British Administration in Ankole,
the Collectors R. J. D. Macallister and R. R. Racey, with the assistance of
Mbaguta, steadily expanded the scope of the authority of the Administration
and of the Mugabe; independent or semi-independent areas in the west were
incorporated in the Mugabe's kingdom; order was enforced; and a hut tax
was imposed and collected. The discipline which the new regime introduced
was not to the liking of many of the Bahima chiefs and there was the constant
danger of their emigration with their herds to German territory.
During these years preceding the signing of the Ankole Agreement, there
occurred four events which were later to find an echo during the investigations
into Gait's murder. In September 1900, Racey, exasperated by Igumira's
intransigence the final provocation being an attempt to seize a petty chieftain-
ship belonging to his neighbour the Mukama of Igara, arrested him and had
him removed to Entebbe. This deportation was also to the immediate advan-
tage of Mbaguta who was given 150 head of cattle confiscated from Igumira's
herds and also his chieftainship of Shema, for, as Racey reports, "Mgabuta
deserves something for the friendly part he has played in keeping the Wankole
Trouble was expected when Igumira was removed, but in fact the only
incident was the flight of Kijoma. This Muhinda chief of Kikyenkye in north
Kashari was a strong supporter of Igumira and when he heard of the latter's
exile he determined to leave Ankole. An attempt was made to intercept and
arrest him but Kijoma speared the leader of the Collector's party and made
good his escape together with his cattle across the German frontier.
The last two of these four events were the deaths in tragic circumstances of
the Bakama of Igara and Buhweju. The rulers of these two little kingdoms,
although at one time independent, had during Ntare's reign acknowledged
the Mugabe as their suzerain. They were not, however, prepared to accept
the policy pursued by Racey and Mbaguta of the complete absorption of their
kingdoms in that of Mugabe. In January 1901 Musinga of Igara was persuaded
to agree to come to Mbarara but he had no intention of submitting to Kahaya
and, when he reached the boundary of Igara, he disembowelled himself.
Ndagara of Buhweju was more spirited in his defiance and an armed expedi-
tion had to be sent against him in July. This ended in an attack upon his
village during which Ndagara and two of his sons were shot.2
The Buhweju expedition was followed by the signing of the Ankole Agree-
ment by which Kahaya was recognized as Kabaka over practically the whole
of the present District. On this occasion too Mbaguta did very well, for, of
the twelve counties into which the District was divided, he secured for him-
self, as County Chief, Shema and Kashari, where Igumira had once ruled,
and Ngarama in addition to his post as Nganzi.
2 Morris, H. F. (1957). The Making of Ankole. Uganda J., 21, 1-15.
THE MURDER OF H. ST. G. GALT 3
A month after the signing of the Agreement, Gait, then twenty-nine years
old, took over from Racey as Collector. Shortly after his appointment, Gait
had to deal with a matter which may or may not later have had some bearing
upon his murder. This was the disposal of the chieftainship of Buzimba.
Buzimba had until the Agreement been a small kingdom under the Mugabe's
suzerainty, and by the Agreement, Nduru, the former Mukama, was recognized
as county chief. Nduru was by then a middle-aged man and unable to adapt
himself to the new role of an administrative chief, so Galt appointed as his
Katikiro the able young Muhinda, Henry Ryamugwizi. Nduru resented the
intrusion of this stranger and appointed his own Katikiro, Mucukura his
brother and heir, Ryamugwizi being left to enforce the unpopular government
measures such as collection of hut tax and labour on the roads. After a few
months Nduru found his position intolerable and resigned, departing with his
cattle across the Toro border into Kitagwenda. Galt thereupon had
Ryamugwizi appointed as county chief. Nduru shortly afterwards died and
his heir Mucukura returned with his cattle to Buzimba.
In September 1903, six months after Galt had handed the District over
to C. E. Dashwood, Igumira was allowed to return to Ankole. Some months
later, Dashwood had to deal with another chief, Isaka Nyakayaga, who was
to be connected with the murder and who was indeed to be one of the two
tried for the offence. Nyakayaga was a Muhima brought up in Buganda
by the C.M.S., a prot6g6 of the missionary, the Rev. H. Clayton, whom he
accompanied to Ankole. Educated men such as Nyakayaga were in great
demand and since Ndibarema, son of the last Mukama of Buhweju, who
signed the Agreement and was appointed county chief of Buhweju, was only
a youth Nyakayaga was made his Katikiro. Nyakayaga, however, soon got
into trouble and was fined twelve head of cattle which were given to Mbaguta,
and Dashwood ordered his dismissal. Nyakayaga appealed to the Com-
missioner who directed F. A. Knowles, who had meanwhile taken over from
Dashwood, to make a full inquiry. The result was that Nyakayaga's property
was restored to him but he was fined Rs.100 and Knowles directed that he
should never be employed again as a chief.
In 1904 there occurred the death of Bucunku the county chief of Mitoma.3
The county chiefs had been called into Mbarara and while there Bucunku
contracted smallpox and died. Under the terms of the Agreement, whereby
the county chiefs could nominate their successors, he left his county to his
cousin, the Muhinda Gabrieli Rwakakaiga,4 who was to be the other accused
in the trial for Galt's murder.
In May 1905 Galt, who had been acting as Sub-Commissioner of the
Western Province, was posted for duty in the Secretariat and told to hand
3 Apolo Bucunku was a grandson of the Mugabe Gasyonga I and a first cousin twice
removed of Kahaya II. It was he who, on Ntare's behalf, had made blood brotherhood
with Stanley in 1889 (see U.J., 12 (1948), 29).
4 Throughout the record of the inquiry and trial dealing with Galt's murder,
Rwakakaiga is referred to as Bucunku's brother. This is obviously a loose transla-
tion of the vernacular word of relationship and the same is probably true of most
other references to brother in the record.
4 UGANDA JOURNAL
over the province to F. A. Knowles. He therefore told Knowles that he would
come from Fort Portal to Mbarara, spending the night on the way at Ibanda
rest-camp in Mitoma county close to the boundaries of Buzimba county and
Toro district. He arrived on the morning of the 19th and that evening was
sitting on the rest-house verandah when, at about 6.30, someone entered the
compound and threw a spear which pierced his lung. Gait called for his cook
to whom he said- "Look cook, a savage has speared me"; and thereupon he
fell down dead.
News reached Mbarara the next day. Knowles, accompanied by Captain
C. E. H. Laughlin, the officer commanding troops, Ankole District, went to
Ibanda and there conducted an inquiry, first examining those who had been
present in the camp at the time of the murder from whom he could discover
no clue to or motive for the crime, and then the rulers and chiefs of Ankole
and Toro who had been summoned to Ibanda. Knowles found the Ankole and
Toro chiefs evasive and anxious lest any information they gave might impli-
cate any of their own people, but one piece of information, which seemed to
Knowles to be of the greatest significance, emerged. Two days before Gait
had left Fort Portal, a Mutoro, Kadebu, had on Galt's orders been hanged
for murder by spearing; that Kadebu's son had taken revenge, also by spear-
ing, appeared a reasonable supposition. Orders were given for Kadebu's
relations to be arrested and brought to Mbarara for examination and, on the
instructions of the Commissioner, the rulers and chiefs were given a stern
warning that their attitude was unsatisfactory; that they were personally
responsible for tracing and arresting the murderer; and that a heavy penalty
would fall on them if they were to be remiss in their duty. Inquiries continued
in an attempt to identify the spear by which Galt had met his death and
which was the only available clue.
George Wilson, the Deputy Commissioner, had in the meantime been
instructed by the Commissioner, Hayes Sadler, to carry out a Special Inquiry
into the murder. He reached Fort Portal on the 3 June and the next day held
a baraza attended by Kasagama and his chiefs. Little could be discovered, and
Wilson was obviously exasperated by the repeated assertions that a Mutoro
could not possibly have committed the crime the origin of which must be found
in Ankole. Realizing that little would be learnt in Fort Portal, Wilson directed
Kasagama and his principal chiefs to accompany him to Ibanda which he
reached on the 8th. It is clear from the long and lucid report which Wilson
submitted as soon as it was over, that the inquiry which he conducted in the
area of the crime was a model of thoroughness.5
No evidence had been found to support the earlier assumption that the
5 The official party engaged on the inquiry consisted of Wilson, Knowles, Laughlin,
two members of the C.M.S., Maddox and Clayton, and Police-Inspector Crean. together
with a military force of 59 and a police force of 69. Throughout the inquiry Kahaya,
Mbaguta, six county chiefs from Ankole, Kasagama, his Katikiro, and seven county
chiefs from Toro were required to be present, whilst Wilson had to assist him two
Baganda chiefs from Bunyoro and his highly-prized Luganda interpreter, Abdul Aziz.
THE MURDER OF H. ST. G. GALT 5
murder had probably been committed by Kadebu's son and this line of
inquiry was therefore dropped, Wilson having from the start inclined towards
a political motive. At the Fort Portal baraza it had been suggested by the
Toro chiefs that the leaders of Ankole had five grounds for wishing to avenge
themselves on the British Administration. These events, which have already
been mentioned, were:
1. The exile of Igumira and the loss of his property and chieftainships.
2. The death of Ndagara of Buhweju.
3. The flight of Kijoma.
4. The suicide of Musinga of Igara.
5. The replacement of Nduru by Ryamugwizi as county chief of Buzimba.
Wilson accordingly set himself to follow up these possible motives. From
the start, he appears to have encountered passive resistance from all
Banyankore concerned. "Beyond the instance of Mbaguta alone," Wilson
records, "there was nothing whatever to show that the country was really
perturbed about the matter. The outlook was terribly black and hopeless.
Time disclosed the fact that the Ankole people adopted alternatively policies
of secretiveness and of mystifying mendacity, which after a time deprived
their most thorough-going partisans of all remnants of belief in them."
Wilson, however, persevered, following up each false trail as it was presented
to him only to abandon it when it led nowhere. First the Toro chiefs claimed
there was evidence to implicate Mucukura, Nduru's brother, but their case
instantly fell through. Then suspicion fell on one Kakyere, a relation of
Ndagara. who was said to have been seen talking to a stranger who carried
a spear similar to that which had killed Gait, this spear having the distinguish-
ing mark that its spike was missing. Many days were spent searching without
avail for this mysterious stranger. The Ankole faction replied to these efforts
of the Batoro to implicate Ankole by producing a seemingly convincing case
to the effect that, when in Toro, Gait had come across a party distributing
an illegally killed buffalo, and had arrested three brothers. While on the road,
two of the brothers had died and the third had sworn vengeance upon Gait.
Witnesses were brought from far and questioned and much time was spent
before Wilson was satisfied that the entire case was based on "conjectural
talk, the whole of the details being absurdly fictitious".
Quite unexpectedly, on the 12 June one Mugwisaki announced that the
murder had been committed by a man called Rutaraka. Mugwisaki was a
peasant of Mucukura, who was still under arrest, and it appeared that one of
Mucukura's wives had induced him to give this information. Parties were
sent out to search for Rutaraka and, though the search was unsuccessful,
another important piece of information was discovered, the spear being identi-
fied as belonging to Kangwagye a relative of Rutaraka. According to
Kangwagye's mother, Rutaraka, who had been staying in her house, had
taken the spear when he had gone out on the evening of the murder. When
she had taxed him later with having taken the spear, he had said that he had
pawned it and, although she had given him cowries to redeem it, he had
not returned it. Now that the identity of the man presumed to have com-
6 UGANDA JOURNAL
mitted the murder had been disclosed, Wilson was inundated with evidence
of Rutaraka's extreme and habitual truculence bordering on insanity. It was
as though they were saying to Wilson, "Here indeed is your murderer, now
be content and leave us in peace". Wilson, however, remained convinced that
persons more powerful than peasants were responsible, if not for the actual
murder, then for its instigation. Rumour spread that Rutaraka had committed
suicide and on the 14th the authorities were led to his grave. The body was
disinterred and identified. It was clear that Rutaraka had met his death by
hanging; but that his death had been a suicide Wilson was extremely doubtful.
The peaceful attitude of the body with its closed mouth indicated that the
body had been arranged immediately after the hanging before becoming rigid,
whilst reports were that the suicide had not been discovered till eighteen
hours after it had occurred. Then followed the interrogation of Rutaraka's
family and in particular of his brothers Binagwa and Ndolere, a hunch-
backed dwarf, "one of the most terrible cases of distorted mind and body I
have ever witnessed" Wilson records. This produced such a bewildering series
of accusations, retractions and contradictions that Wilson and his party were
reduced almost to despair. The identity of Galt's murderer had apparently
been established and also the fact that he in his turn had been murdered, but
as yet there was no clue as to motive or as to who had instigated the crimes.
It was at this time that Wilson's suspicions turned definitely towards
Gabrieli Rwakakaiga and Isaka Nyakayaga. As had already been mentioned,
Nyakayaga had been dismissed from Buhweju. After his dismissal, he had,
without the knowledge of the Collector, or apparently of Mbaguta, been
installed as one of Rwakakaiga's sub-chiefs in the area north of Ibanda. When
Gait had arrived in Ibanda, neither Nyakayaga nor Rwakakaiga were there
to welcome him although they had been warned of his visit. Even on hearing
of the murder, Rwakakaiga, instead of rushing to the spot, had leisurely made
his way to Ibanda taking three days to complete the journey. Then, when
Mbaguta had begun inquiries, he had refused to send men in to be examined
and neither he nor Nyakayaga had made any attempt to identify the spear.
Further incriminating evidence now came to light. Rwaniko, another brother
of Rutaraka who had been arrested on Wilson's orders, had been taken to
Rwakakaiga's camp and from there allowed to escape. When subsequently
re-arrested, Rwaniko stated that he had been given a spear and told to go
off. Knowles, therefore, held an inquiry into the conduct of the two chiefs who
were found guilty of "the most culpable negligence and indifference and in
suppressing knowledge which would have been of the greatest use towards
securing the murderer Rutaraka". They were both accordingly placed under
Wilson then arranged for Igumira to be brought to Ibanda in the hope
partly that he might be persuaded to use his vast influence in furthering the
inquiry and partly so that a closer eye could be kept on him. Wilson was,
somewhat surprisingly, strongly impressed by Igumira, comparing him most
favourably with Mbaguta, and he certainly seems to have been of the opinion
that he had established some degree of intimacy with him. So to Igumira
Wilson confided his suspicion of Rwakakaiga and Nyakayaga, and Igumira
THE MURDER OF H. ST. G. GALT 7
apparently told him that he was of the same opinion. Even Kahaya eventually
admitted that Wilson's suspicions might be justified. This was indeed surprising
for "all their (Kahaya's and Igumira's) interests, from their point of view,
were dependent upon the exoneration of Gabriyeli (Rwakakaiga); and
Kahaya's extreme misery in having to deviate from a sort of neutrality which
he adopted when he found he could not distract us from attention to the
Gabriyeli clue, bore out the inference that the conviction of Gabriyeli was
dangerous, either to his safety in the case of his possible collusion, or to his
popularity with the Bahima of his country".
Once again attention was turned towards Mucukura but the evidence pur-
porting to implicate him was forthwith proved worthless; indeed "the sudden
unanimity among the witnesses against him" convinced Wilson that there was
a plot to frame him and at the same time to put Wilson off the track of
Rwakakaiga. Despairing of ever discovering the truth in an atmosphere so
hostile, Wilson on the 30 June gave orders that the camp was to be struck
and that the official party, together with Igumira and all witnesses and
prisoners, should move to "a less repressive atmosphere". Next day, Wilson
gave a farewell address to the Ankole chiefs telling them that "I had no wish
ever to see them again as they were the most benighted people of the four
kingdoms". Kahaya begged to see Wilson but was told "that I could not
shake hands with him". Then, at Knowles's request, he appointed his inter-
preter, in whom he had unbounded, and as afterwards turned out misplaced,
confidence, Abdul Aziz, as county chief of Mitoma in Rwakakaiga's place.6 All
was ready for Wilson's departure when that evening a piece of information,
as vital and unexpectedly produced as Mugwisaki's statement that Rutaraka
was the murderer, was disclosed.
Kahaya had produced yet further evidence purporting to incriminate
Mucukura, this time to the effect that he had given a bullock to Kangwagye,
and, in pursuing this clue, the cattle of a herdsman called Bitabi were taken for
identification. Finding his cattle gone and seeing the camp about to leave,
Bitabi came to Wilson and poured out the following story.
Some time previously, Bitabi said, he had taken milk to Rwakakaiga, and
at a drinking party had heard Rwakakaiga offer a reward for the death of a
European. This death, Rwakakaiga had said, was to be in revenge for the
death of his uncle Manyatsi who had been killed while fighting for the Kabaka
Mwanga and also for the death of his cousin Bucunku who had been given
smallpox by the magic of the Europeans. Several of the people whom Bitabi
stated to have been present at the party were among Wilson's prisoners and,
when confronted with Bitabi's disclosures, they all confirmed his statement.
THE BARAZA AT HOIMA
Next day the party left for Hoima and on the march further evidence was
given by an old man Biryagi who had been a servant in Nyakayaga's house.
He stated that, as the whole country was tired of the discomfort and suspense
6 This appointment was made without first obtaining the approval of the Com-
missioner for which Wilson was later mildly rebuked.
8 UGANDA JOURNAL
caused by the inquiry and as the culprits had been arrested, there was noth-
ing to be gained by longer suppression of the truth; he then in general,
confirmed Bitabi's story. At Hoima the examination began again. This pro-
duced another important piece of information, one Kaheriko confessing he had
stood by while Binagwa and Ndolere had killed Rutaraka. In the face of this
confession, Binagwa and Ndolere admitted that they had strangled Rutaraka
with Kaheriko's help.
Wilson now felt that sufficient evidence had been procured and that the
inquiry should be concluded by a baraza at which "the evidence would receive
formal magisterial attestation". On the 16th the baraza was convened under the
magistrate J. O. Haldane, with A. B. Fisher of the C.M.S. as informal assessor,
and the Mukama and county chiefs of Bunyoro as native counsellors.
Reference need only be made here to the evidence of seven of the witnesses.
Binagwa's evidence was that, a few days after the murder, Nyakayaga had
told him he had heard that a man from his village had killed Gait. To this
Binagwa replied, "Yes, indeed, but did you not send him?" Nyakayaga's
answer was "Do not tell anyone, go and look for Rutaraka". After some time,
Binagwa found Rutaraka and brought him to Ndolere's hut where on Nyaka-
yaga's instructions he hid him. Binagwa and Ndolere then said to Rutaraka,
"You have told us you are going to kill yourself but who told you to kill the
Bwana?" He replied that it had been Nyakayaga acting on instructions from
Rwakakaiga because the Europeans had exiled Igumira to Entebbe and had
given him food which no man could eat.7 That day Rutaraka killed himself.
Kangwagye stated that one day he, Kakyende, Binagwa and Rutaraka took
bananas to Nyakayaga's house where they met Rwakakaiga and where they
spent the night. In the morning Rutaraka was sent for and Rwakakaiga said
to him, "If you can kill a European I will give you a hundred head of cattle,
some land and a chieftainship." On the way home, the others asked Rutaraka
if he would be able to accomplish his task and he replied "Yes, even if I have
to search morning and night I will get a chance and kill one."
Kakyende, a brother of Kangwagye, stated that he remembered a time when
Bitabi had brought milk to Rwakakaiga. He, Rwakakaiga, Nyakayaga,
Kangwagye, Kaigiri and Rutaraka were then drinking beer. Rwakakaiga
said it would be a good thing to kill a European and that the reward would
be many cattle and that he wanted this done because his cousin had been given
smallpox. Rutaraka then agreed to do the deed.
Bitabi's evidence was that, about five months previously, when Rwakakaiga
had come to Nyakayaga's village, he brought milk for them. He heard
Rwakakaiga say "The European and Mbaguta gave my cousin poison which
gave him smallpox." He then promised a hundred head of cattle and land to
anyone who would kill a European. The men with Rwakakaiga were also
told that they would not have to do any work again. Later, Bitabi's cattle
were seized and he was put to work for five days with a hoe (a great indignity
7 The fact that Igumira when in exile had to eat such food as chickens and eggs,
food which no Muhima could touch without shame, seems to have made almost as great
an impression on the Banyankore of the time as the exile itself, and one of the reasons
for Musinga of Igara's suicide is said to have been that he feared that the Europeans
would m:.ke him, like Igumira, eat eggs and do other shameful things.
THE MURDER OF H. ST. G. GALT 9
for a Muhima) and, resenting the injustice of all this, he decided to tell all that
Biryagi's story was that about seven months previously he had heard
Rwakakaiga say to Nyakayaga and Bagyenda in Nyakayaga's house that he
wanted vengeance because the Europeans had killed Manyatsi and Bucunku
and were aiming at killing all the leading Banyankore so as to get possession
of the country. Shortly after this, Nyakayaga appointed Bagyenda headman
over Bigyera with instructions to watch the road from Toro and report when
any Europeans were seen. When Bagyenda heard that Gait had arrived, he
brought word of this to Nyakayaga. That night, Rwakirenzi, the Ibanda rest-
house keeper, came to Nyakayaga and said that Gait was dead. Nyakayaga
however, would not believe it saying "How has he been killed so soon?";
having dressed he went at once to the camp but was turned away by the guards.
Kaheriko stated that after Gait's death he had been sent by Ndolere to tell
Nyakayaga that that which he had ordered Rutaraka to do he had done and
to ask for Rutaraka's reward. Then Kaheriko, Ndolere and Binagwa strangled
Rutaraka, Binagwa tying the rope round his neck while Ndolere held his
feet, Kaheriko standing by.
Ndolere stated that, after Rutaraka had killed Gait, he sent Kaheriko to
tell Nyakayaga that Rutaraka had done this and would probably kill some-
one else. Kaheriko returned with orders from Nyakayaga for Rutaraka to
be killed. He, Binagwa and Kaheriko accordingly killed him.
Rwakakaiga's defence was that the witnesses had conspired against him in
inventing the story that he had spoken of a European's death in Nyakayaga's
house. Bitabi for example, could not have known of such a conversation since
he had brought milk in the morning and had gone away immediately after-
wards, in confirmation of which he could produce fourteen witnesses. As for
Rutaraka, he was unknown to him. Furthermore, and here Rwakakaiga was
on strong ground, what real motive was there? Why should Bucunku's death
be attributed to witchcraft? His great-grandfather had died of small-pox but
no one had said it was from witchcraft. Why too should he wish to avenge
Manyatsi's death when Manyatsi had left two brothers perfectly capable of
doing so? Why should he, a Christian, wish to kill Europeans with whom he
had had dealings since he was a boy and one of whom he could in any case
have killed long before this had he wanted to? The Europeans had accepted
him as the heir of Bucunku from whom he had inherited 1,700 head of cattle
and a large county.
Nyakayaga in his defence had little to say beyond asserting that the evidence
brought against him was false. The baraza then proceeded to give its verdict
which was that Rwakakaiga and Nyakayaga were guilty of having instigated
the murder of Gait.
Wilson concludes his report of the inquiry with a review of the facts. That
Rwakakaiga and Nyakayaga were guilty of instigating the murder, he had not
the slightest doubt, and indeed he went so far as to suggest a suitable punish-
ment of ten years imprisonment and perpetual exile as appropriate for them,
since for Bahima, death would be an inadequate punishment or deterrent for
others. Furthermore, the Mugabe and the county chiefs should be fined at least
10 UGANDA JOURNAL
1,000 head of cattle and three months labour should be provided on public
works by the people of Mitoma, Buzimba and Kitagwenda. The findings of
the inquiry confirmed the suspicion which Wilson had held all along that the
murder was political in its nature, rather than an isolated act of revenge by
some peasant with a grievance. With the motive attributed to Rwakakaiga,
Wilson was, however, far from satisfied. Though many of his informants
assured him that in Ankole, where superstition was rife, such an attitude of
mind was understandable, Wilson remained unconvinced that the real motive
had, in fact, come to light. He was convinced that more sinister and powerful
forces were at work and he believed the Mugabe and his chiefs were impli-
cated in the crime which in Wilson's opinion was an expression of an
antipathy towards European rule not confined to Ankole where "effervescent
excitability has led it to commit itself at an earlier stage than any of its
On the 10 October 1905 the Entebbe magistrate (Edward Treffry) committed
Rwakakaiga and Nyakayaga for trial by the High Court at Entebbe on the
charge of having abetted the murder of Gait committed by Rutaraka. On the
20th, the case opened before Judge G. F. M. Ennis.8 The case for the prose-
cution rested upon the evidence of the five witnesses, Binagwa, Kangwagye,
Kakyende, Bitabi and Biryagi, concerning conversations which they professed
to have taken part in or to have overheard and during the course of which the
accused had incited Rutaraka to commit the murder. There was, of course,
other supporting evidence such as the accused's behaviour and the lack of
assistance they had given to the authorities, particularly concerning the identi-
fication of the spear, but none of this additional evidence would in itself have
proved the guilt of the accused. The vital evidence which the five gave of
the conversations was vivid, detailed and bewildering. Inevitably, in
describing events which had taken place a considerable number of months
before, all references to time were extremely vague; but, furthermore, in many
of the details the witnesses contradicted one another and themselves.
According to Binagwa's evidence, three months before Galt's death,
Rwakakaiga said to Nyakayaga in his presence, "Go and kill a European".
This was said at Bigyera in Kamugasha's house in which Nyakayaga was
living. Kibanda was the only other person present. In his evidence during the
preliminary inquiry, Binagwa had also stated that he remembered the occasion
when Bitabi had brought milk to Rwakakaiga, when he, the two accused,
Kangwagye, Kakyende and Kataigira had drunk and talked. Rutaraka had
not been present on that occasion. In his statement before the Hoima baraza,
Binagwa had said nothing of any of this evidence.
According to Kangwagye, five months before the murder Rwakakaiga gave
orders direct to Rutaraka to kill a European. These instructions were given
after sunset in the presence of Kangwagye, Rwacende, Bitabi and Kakyende
in Kwirigira's house in which Nyakayaga was staying; Kwirigira, the witness
8 The accused were defended at their own expense by Byramji Rustamji of Mombasa.
His fees were paid in cattle.
THE MURDER OF H. ST. G. GALT 11
told the court, was not the same man as Kamugasha. Later, when he was
with Rwakakaiga in Mbarara, Kangwagye heard Rwakakaiga ask Rutaraka
whether he had forgotten the instructions given him.
According to Kakyende's evidence, six months before the murder
Rwakakaiga told Rutaraka that he would give him land and one hundred
head of cattle if he would kill a European because the Europeans had given
Bucunku smallpox. This conversation took place at sunset in his presence
and in that of Rwacende, Mai, Kangwagye, Bitabi and Nyakayaga in the
enclosure to Nyakayaga's own house at Bigyera on the day Bitabi had brought
milk. Later, he went with Rwakakaiga to Mbarara where the latter asked
Rwacende whether he would keep silent if he whom he had instructed carried
out his orders; Rwacende said he would. After Kakyende's arrest, his house
was burnt and his property taken on Rwakakaiga's orders.
Bitabi's evidence was that when he had brought milk for Rwakakaiga, then
staying with Nyakayaga in Kamugasha's house, he overheard Rwakakaiga, on
the other side of the reed fence, say he would give one hundred head of cattle
to the man who killed a European because Mbaguta had asked the Europeans
to bewitch Bucunku and they had given him smallpox. Rwakakaiga then told
Rutaraka to hang the European saying, "When you have killed the European,
hide yourself until things are quiet and then come and claim your reward."
Kakyende, Kangwagye and Rwacende were also present at this conversation
which took place in the morning.
According to Biryagi, when Rwakakaiga gave land to Nyakayaga, he said
in the presence of Biryagi, Kataigira, Bagyenda and Rwakirenzi that
Nyakayaga was getting the land so as to be able to kill a European who would
come from Toro. Nyakayaga asked what he would get for killing the
European. Rwakakaiga replied that he would get more land, and Nyakayaga
agreed. Rwakakaiga said that the murder was to avenge the deaths of
Bucunku and Manyatsi. This conversation took place in Kamugasha's house.
Hangamaisho, Nyamununu, Kangwagye, Kakyende, Rwacende and Bitabi
were sent out of the house before Rwakakaiga spoke to Nyakayaga; they
did not know that Biryagi was in the kitchen listening. This was the only
such conversation he had heard before the murder.
Such was the main evidence for the prosecution, Ndolere and Kaheriko not
being called as witnesses. The accused gave no evidence in defence and called
no witnesses. Their Counsel's argument was that the principal witnesses were
unworthy of belief and were, in fact, no better than accomplices. Ennis in his
judgment referred to the discrepancies in the evidence but dismissed these
with the words "Such discrepancies are however usual and to be expected
in native evidence", and he adds that to his mind the evidence that there was
a conspiracy between the two accused to commit murder was convincing. On
the question of motive, the Judge evidently felt he was on less sure ground
for he merely says, "There is some evidence to show that the motive was
revenge for the death of Bucunku who died of smallpox, believed by the
superstitious natives to have been given him by the Europeans." The Judge
was in agreement with the two assessors (W. E. Reymes-Cole and Major
L. C. E. Wyndham, both administrative officers) that the two accused had
12 UGANDA JOURNAL
instigated Gait's murder and convicted them and sentenced them to death.
Appeal from Judge Ennis's verdict was made to the Court of Appeal for
East Africa and was heard at Mombasa in January 1906. Judges Lindsey
Smith and R. W. Hamilton in a majority verdict tore to shreds Ennis's judg-
ment. It appeared to them quite preposterous that a court could have convicted
on the evidence of the five principal witnesses. The case, said the judges,
must stand or fall on the evidence of these witnesses as to the conversations
said to have taken place before the murder; no other facts were of sufficient
weight to affect the issue; and as for the motive put forward by the prosecu-
tion they could rely upon it even less than had the trial judge. As for the
evidence of the incriminating conversations, it was that of witnesses who
had only come into court after much pressure had been brought to bear on
them, whilst some of them had strong motives for laying the blame on the
accused. They were, if their story were true, accomplices and their evidence
was therefore entitled to no weight unless corroborated independently. "We
have," said the judges, "gone through the evidence most zealously and find
that not a single witness corroborates the story of the other even broadly. We
find that it is a mass of contradictions and inconsistencies impossible of
reconciliation. These witnesses do not agree as to the period, the time of the
conversation, the village or the house where it took place, the reward offered,
the people who were present, the mode of killing or the reason why
Rwakakaiga desired a murder." It was well known, the Judges agreed, that
Africans give evidence very loosely but also that as a rule they have very good
memories for things that interest them and indeed some of the witnesses had
shown exceptional power of memory. Having dealt in detail with the various
inconsistencies, the judges concluded by reversing the decision of the lower
court and ordering the acquittal of the two accused.
The third appeal judge, J. W. Murison, however, submitted a minority
judgment, because he disagreed not with his colleagues' decision but with many
of their arguments. Unlike the other two judges, he was not at all sure that
the witnesses had been lying, for the many and great inconsistencies which
occurred in their statements might well have arisen from their remembering
parts only of the conversations they had heard; different rewards might
have been offered at different times; a date meant nothing to an
African; and finally the statements had often come through two interpreters.
Nevertheless, he felt that the inconsistencies were such as to raise a reasonable
doubt as to whether they were telling the truth and his judgment concludes
with these words, "I have felt all along a very grave suspicion indeed that
Rwakakaiga and Nyakayaga did in point of fact instigate the murder of Mr.
Gait. But the only question before us is whether or not it has been legally
and conclusively proved that they did so. In my opinion it has not, and I
therefore think that the judgment of the Lower Court must be reversed and
both accused acquitted."
AFTER THE TRIAL
George Wilson, who at the time was Acting Commissioner, pending the
arrival of the new Commissioner, Hesketh Bell, was amazed when he heard
THE MURDER OF H. ST. G. GALT 13
of the acquittal, for, after the evidence produced at Hoima, conviction had
been taken as a matter of course. In a despatch to the Secretary of State, Wilson
first says as much as (perhaps more than) he properly could of his opinion
of the Appeal Court's decision and he then states that, as his Report of the
Inquiry had made amply clear, it was absolutely necessary for the peace and
good order of the country that Rwakakaiga and Nyakayaga should be exiled
for life and that he was accordingly deporting them to the East Africa Pro-
tectorate. At the end of January 1906, Rwakakaiga and Nyakayaga left
Entebbe for Kismayu where ten years later Rwakakaiga died. Nyakayaga re-
mained in exile till 1925 when he was allowed to return to Uganda provided he
lived at Jinja. In 1928 he was allowed to go back to Ankole. Igumira who had
been taken by Wilson to Hoima after the Ibanda Inquiry was never allowed
to return to Ankole and died an impoverished exile in Buganda in 1925.
The fines upon the Mugabe and chiefs of Ankole and the exaction of free
labour from the people of Mitoma, Buzimba and Kitagwenda which Wilson
had recommended in his report were imposed. Furthermore, the Ankole Agree-
ment was suspended. This, it appears, was done not merely as a punishment,
but because it was thought that under the terms of the Agreement it was not
possible for the murderers of Rutaraka to be tried by the High Court. The
Secretary of State, in giving his approval to the suspension, informed the
Commissioner that the High Court would have been perfectly competent to
try the case despite the existence of the Agreement, since under the Uganda
Order in Council it had "full jurisdiction civil and criminal over all persons
and over all matters in Uganda".9 He stated further that the suspension must
"be regarded as an Act of force justified by the necessity of marking the
gravity of the crime which had been committed" since it was not clear that
there had been any infringement of the terms of the Agreement such as would
justify its annulment under Section 3. The Agreement was not restored till
1912. The greatest hardship which the people of Mitoma suffered as a con-
sequence of the murder, however, was undoubtedly the appointment of
Abdul Aziz, Wilson's prot6g6 upon whom he had lavished such extravagant
praise, as county chief. His tyrannical rule is still remembered by the Banyan-
kore with loathing though his period as county chief was short, for he was
dismissed a few years later after a complaint to the Administration had
revealed that an arbitrary punishment inflicted on one of his people had resulted
in the loss of the man's arm.
That the evidence brought against Rwakakaiga and Nyakayaga should have
been found insufficient to sustain their conviction is not surprising to anyone
9 Section 6 of the Ankole Agreement states that: "Justice as between native and
native shall be administered direct by the recognized chiefs." In the opinion of the
Protectorate Government this precluded trial by a Protectorate Court where both the
accused and the murdered man were Africans. Such an argument did not, of course,
apply to the main case concerning Gait's murder, where the victim was non-African.
Those held responsible for Rutaraka's death were sentenced to ten years imprisonment.
14 UGANDA JOURNAL
who has studied the proceedings,10 and it would at this stage, whilst one of
the principal personalities is still alive, be improper to attempt to analyse fully
the conflicting and confusing evidence available. Some points may, however,
be mentioned. One interesting fact is that although it was assumed throughout
the trial that it was Rutaraka who actually did the deed, no proof of this was
ever produced and there is in Ankole a considerable body of opinion which
believes that it was not Rutaraka but another peasant who killed Gait.
Nyakayaga, now nearly eighty, firmly maintains that this is so. Rutaraka, he
asserts, was suffering from sleeping sickness and was far too weak to have
committed the crime. If this is so, then presumably Rutaraka was chosen as
a scapegoat as being a useless member of the community near to death in any
case, and was murdered in the hope, subsequently fulfilled, that it would be
assumed that it was he who had speared Gait.
The most bewildering aspect of the whole case is the apparent absence of
motive, either personal or political. No evidence could be found during the
enquiry that anyone had a grudge against Galt, although this does not prove
that such a person did not exist or that there was not some madman or fanatic
who decided to take his life. Had this been so, it is, however, surprising that
the native authorities in Ankole, certainly when faced with the consequences
of the crime and of the fact that Rwakakaiga was in danger, did not manage
to unearth the necessary evidence. Moreover, such a supposition would leave
unexplained the reason why Rutaraka was murdered. The murderer might,
of course, have indeed been Rutaraka acting, not on the orders of others, but
from some personal motive of revenge or because his mind was unbalanced as
the result of his disease. But here again, if this was so why was he murdered?
From the point of view of motive, the obvious suspect was Mucukura, and
indeed it was upon him that most suspicion rested during the early days of
the inquiry. He had been deprived of his inheritance and had seen the former
kingdom, over which his ancestors had ruled as independent Bakama,
handed over to the control of a rival. He had indeed ample reason for desiring
revenge, not merely against the British Administration and the Native
Government, but against Galt personally, since he had been responsible for
his brother's retirement. Nevertheless, although the members of the inquiry
paid particular attention to the possibility of his implication, and although
Rwakakaiga's supporters were anxious to pin the blame upon him, no really
incriminating evidence could be found against him.
As far as political motive is concerned, Nyakayaga stands firm by the
contention put forward by the Ankole chiefs at the time of the inquiry, that
the instigators of the murder were those in control in Toro. The peasant who
did the deed, he maintains, was persuaded to do it by the Toro chiefs in
order to get Ankole into trouble in revenge for the campaigns which Ntare
10 Sir John Gray in a recent letter to the Uganda Journal, (22 (1958), 86), makes the
following reference to the judgment: "The acquittal must in the circumstances stand for
all time. Not only so, but having examined the record of the proceedings carefully and
critically, I may perhaps after 36 years' experience of trials of Africans in Africa be
allowed to say that in my opinion the conclusion reached by the Court of Appeal was
the only one which could be reached in accordance with justice, reason and common
THE MURDER OF H. ST. G. GALT 15
had carried out against Toro in the previous century. He also maintains that
it was Mbaguta who was responsible for the case which was made against
him. Whilst not accusing Mbaguta of having been in any way responsible for
the murder itself, he maintains that he took advantage of the difficulties into
which Rwakakaiga had got himself and worked up a case which would rid
him for good of a dangerous enemy.
If, however the motive was a political one of trying to ensure that retribution
for the murder would be visited upon the heads of others, then it is more likely
that the murder was instigated by Ankole chiefs. For nearly ten years, the
Bahinda had suffered under the autocratic rule of the upstart Mbaguta. He
had enforced unpopular measures introduced by the Administration which
had curtailed their privileges and had reduced their power. He had been
responsible for the humbling of their hero Igumira and had benefited, at his
expense, from his downfall. Exasperated beyond measure, they may well have
felt that there was no way in which they could attack him whilst he remained
the trusted adviser and mouthpiece of the Administration. The only way in
which they could cause his downfall was by undermining the Administration's
faith in him, and what more effective way of achieving this was there than
that of the committing of some outrageous crime against the peace of the
country for which he was responsible.
Such a theory may seem fanciful but it is in accordance with the opinion
held by many Banyankore that the murder was done to get Mbaguta into
trouble. Furthermore, it would explain the conspiracy of silence which Wilson
encountered among all the Ankole chiefs (except Mbaguta) and the reluctance
of leading Banyankore to discuss the matter in after years, an attitude which
persists among the older men even to this day. It is unlikely that the mystery
of the murder will ever be solved. Those who know the secret are taking it
with them to their graves, whilst among the younger Banyankore there is
supposition but little knowledge; indeed legend now seems to be taking the
place of fact and I should like to conclude by quoting one of these legends.
A European of the White Fathers' Mission, it is said, was staying in the
Ibanda rest-house the day before Gait arrived. That evening, an African came
to the rest-house concealing a spear and professing to have a chicken to sell to
the White Father. The latter at the time was reading his breviary and, seeing
an African come into the compound and not wishing to be disturbed, he curtly
told him to be off. The would-be murderer, thinking that the White Father
had by the magic of his book divined his intention made off as fast as he
11 This may well, however, be no legend but a statement of fact in which case it
confirms the supposition that it was the indiscriminate murder of any European that
EXCAVATIONS AT BIGO, 1957
By P. L. SHINNIE
THE large earthwork site at Bigo is certainly the most spectacular as well as
the best known of the relics of Uganda's past. There is already a consider-
able literature concerning it from the first description by D. L. Baines in the
Official Gazette for 15 May 1909 down to recent articles by Mathew(1) and
Lanning.(2) A full description has been given by Wayland(3) together with a
plan and need not be repeated here.(4) An air photograph has also been pub-
lished(5) and comparison with the plan in Wayland's article (made by the late
A. D. Combe of the Geological Survey in 1921) shows the plan to be remarkably
accurate. It also need not be repeated.
Traditionally, this site and the many other earthworks of the area are asso-
ciated with the Bachwezi although other more or less fantastic stories have
been told of them-such as those crediting them with being of Portuguese or
Abyssinian or even, by an extreme flight of imagination, Roman origin.(6)
There has been considerable theorizing as to the origin of the Bachwezi and
recently Wrigley(7) has even suggested that they are merely a mythical people,
personifying the natural features and elements of the country.
This is not the place to discuss the various views of Bachwezi origins or
existence, but it was in an endeavour to throw some light on the historical
problems connected with these legends and to reveal the material culture of
the builders of the earthworks that archaeological excavation was carried out
at Bigo in August and September 1957.
The excavation was carried out by me as Director of Antiquities to the
Uganda Government with the assistance of my wife and Mr. P. L. Carter. Mr.
S. Kyeyune was appointed as interpreter and also took part in the archaeo-
logical work. The help of Mr. E. C. Lanning was invaluable in the planning
stages and I must also express gratitude to Mr. Todd of the Public Works
Department in Masaka for the loan of the tools without which the exacavation
could not have been carried out. An average of 20 men and 10 boys was em-
ployed throughout the work.
The site of Bigo covers an area of about 5 square miles and with the limited
resources of time and money available only a sample excavation could be
carried out. It appeared that the best results would be obtained by concentration
on the centre of the complex where such surface signs as were visible in the
thick bush suggested that the main occupation had been, and where earlier
visitors had identified two mounds (the 'heaps' of Combe's plan).
Clearance of the bush, itself a considerable task, showed that in this area
there were in fact three mounds partially surrounded by a ditch so silted up as
to be virtually invisible until the vegetation had been cleared. In this respect
the ditch in this area differs markedly from the other Bigo ditches which, though
Bigo. Mound A, Trench I seen from the south.
Bigo. Trench I. The old turf line shows up as the dark layer in the section and
the small trench can be seen cut into it.
EXCAVATIONS AT BIGO, 1957 17
heavily overgrown, appear to have very little silt in them and to retain approxi-
mately their original shape.
Two trenches were laid out: one to cut across the ditch near its end and to
cut into the mound, and another to give a second cut across the ditch (see
Fig. 7). The cutting of Trench I soon showed that, as had always been assumed,
Mound A was man-made. Its true shape, not visible until cleared, showed it
to be virtually flat-topped and of a markedly artificial character and, since
the material (gneiss) of which it was formed is identical with the rock into
0 0 44 P 4 o l8 I c t
rUNCO a I
Bigo. Layout of trenches.
which the ditch is dug, it is reasonable to assume that it was built of the rubble
dug out in the making of the ditch. Absence of stratification, although tip lines
can be seen, show that it was built in one single operation. A suggestion that it
was, in fact, a natural hillock was considered for a time-but the finding of a
few scattered potsherds in the make up disproved this.
Excavation of the ditch showed it to be full of occupation debris. From the
quantity of pottery and bone found it had certainly been used as a rubbish
tip and it is to this cause that must be attributed the silting which occurs only
in this area.
The surprise of the excavation was the discovery that the ditch, originally
assumed from its surface indications to be only a shallow one, had been dug
to a depth of 12 feet at the point where it was cut by Trench I. It must cer-
tainly have been cut out of the rock with iron picks. A part of one of these tools
was found and the marks of them on the rock through which the ditch was cut
could be clearly seen.
18 UGANDA JOURNAL
The ditch (see Fig. 8) is not a wide one-it is just over 10 feet wide from
lip to lip and only 2 feet wide at the bottom. As can be seen from the section,
the lower part of the fill is of gravel, which contained little occupation material,
and is presumably mainly natural rain-wash. Above it is a layer of ashy gravel
and above that again pink earth (decomposed gneiss) which must be material
washed down from the mound. Above that again is darker soil containing a
quantity of pottery and bone.
The mound overlies a layer of dark brown soil which represents the turf
line before the mound was built and is cut through by the ditch. Below this
turf line, were found a series of small, round holes, presumably post-holes, as
well as three irregular excavations of unknown purpose and two small trenches
14 inches deep and 8 inches wide. These can be seen in Fig. 9. The southern
of these two trenches had eight post-holes dug into its fill. The trenches had
been dug through the turf line as can be clearly seen against the side of the
cutting in Fig. 5. The chronological relationship of the small trenches and the
post-holes to the main ditch is not certain-clearly they existed before the
mound was raised on top of them and they are filled with the material of which
the mound was constructed. It seems probable that they also pre-date the ditch
which we assume to have been dug at the same time as the mound was raised.
I suggest that the small trenches were "marker" trenches for a palisade, the
posts of which were set up in them. In the comparatively narrow cutting that
we excavated it is not possible to see any coherent pattern in the arrangement
of the other post-holes, though no doubt if the immense labour of clearing the
mound could be undertaken this would become clear. We can assume, failing
other evidence, that they are the holes for hut posts.
There is no clear differentiation between the pottery found below the turf
line in the post-holes, and that found in the ditch; this suggests that no long
period of time elapsed between the pre-mound occupation and the raising of
Trench II gave another cut across the ditch and told much the same story of
a deep ditch cut into the rock and subsequently filled with debris.
Some 50 yards to the south of the main ditch a number of very low mounds
and a considerable scatter of pottery suggested that here was the main area
of occupation from which the debris in the fill of the ditch had come. There
was unfortunately no time to excavate this area.
The pottery from Bigo was extremely plentiful and would repay detailed
study. There is space for only a brief description here.
The main classification is into coarse and fine ware, the fine ware itself
being subdivided in (i) a highly burnished ware with a red slip and (ii) an un-
burnished black ware. This pottery is identical to that from the nearby site of
Ntusi of which Dr. Mathew has given some description.(8) The Bigo coarse
pottery appears to be the same as his Class C but I am unable to identify his
Classes A and B.
SBrn- E-0 Ash
5^Rnrl. E~l Ashy C&-oU
. I I Tf.ET
; m i
Bigo. Section of Trench I.
* POST .HOLES
Bigo. Plan of Trench I,
Bigo. Section of Trench 1.
20 UGANDA JOURNAL
This ware is either black or brown. The pots themselves appear to have
been mainly globular though few complete ones were found; Fig. 10 (1) is an
exceptional piece. The majority of pots have a pattern made by rolling a piece
of knotted grass on the clay before firing-this pattern is commonly found
on the rims and the upper part of the body of the pot and is not found on the
lower parts. Examples are illustrated in Fig. 6.
A very common decorative feature is the painting of red stripes on the sides
of the pots. This was done by applying red ochre-covered fingers to the pot.
The pot illustrated as Fig. 10 (1) is of particular interest. It is open at both
ends and was pierced for suspension at the narrow end. It is of black ware
and is covered all over with knotted grass patterns. Two similar but smaller
pots were found by E. C. Lanning at Mubende Hill (Kampala Museum Nos.
5328 and 5329). The purpose of the pot is obscure-it has been suggested that
it is a fumigating pot but how it was used has not been explained.
(1) Black fumigating(?) pot from fill of ditch in Trench I. Kampala
Museum No. A.59.10.
(2) Black pot with decorated rim and red paint finger marks. From lower
fill of ditch in Trench I. Kampala Museum No. A.59.12.
(1) Black pot with very irregular pattern on rim. From fill of ditch in Trench
I. Kampala Museum No. A.59.13.
(2) Black pot. Red paint finger marks. From ditch fill in Trench II. Kampala
Museum No. A.59.14.
(3) Black pot. From ditch fill in Trench II. Kampala Museum No. A.59.15.
(1) Black ware. Red slip inside rim, knotted grass pattern below rim.
Mound A, Trench I, level 1.
(2) Gray ware. Traces of red paint applied with finger tip below rim. Knotted
grass pattern on rim. Mound A, Trench I, surface.
(3) Gray ware. Blackened on outside from firing and with traces of red
paint. Knotted grass pattern on rim. Mound A, Trench I, level 3.
(4) Gray ware. Traces of red slip on rim. Knotted grass pattern below rim.
Mound A, Trench I, level 1.
(5) Black ware. Covered inside and out with red slip. Decorated with 13
parallel wavy lines on outside of rim. Mound A, Trench I, surface.
(6) Gray ware. Lightly burnished red slip on rim and outside. Mound A,
Trench I, surface.
(7) Very thick black ware. Lightly burnished red slip on rim and inside.
Mound A, Trench I, upper fill of ditch.
(8) Brown ware. Red slip on rim and shoulder. Mound A, Trench I, in
hole in natural.
EXCAVATIONS AT BIGO, 1957 21
' ... :;:" :'
S. : : ,. :* -, .." /
-"':- .. ,^ *..
,, : . .
7 ", '
," ?: ;-
.-.. . ., . ,
EXCAVATIONS AT BIGO, 1957 23
(9) Black ware. Red slip inside and out. Knotted grass pattern on rim.
Mound A, Trench I, level 1.
(10) Gray ware. Red slip on rim and for short distance down on inside.
Knotted grass pattern on body. Mound A, Trench II, level 1.
56 7 8
9 0 1 12 13 14
Bigo. Pot Rims.
(11) Black ware. Lightly burnished black slip on outside. Narrow band of
knotted grass pattern on outside immediately below rim. Mound A,
Trench I, level 1.
(12) Gray ware. Lightly burnished red slip inside and out. Knotted grass
pattern immediately below rim. Mound A, Trench I, hole in natural.
(13) Gray ware. Lightly burnished inside. Knotted grass pattern below rim
on outside. Mound A, Trench I, surface.
(14) Brown ware. Red slip on rim. Knotted grass pattern on body. Mound A,
Trench II, level 1.
24 UGANDA JOURNAL
No complete pots of the finer wares were found although a considerable
number of sherds were collected. A characteristic of this ware is that much of
it had a highly burnished red slip. A selection of rims is illustrated in Fig. 13.
2 3 4
Bigo. Pot Rims.
(1) Black ware. Traces of burnished red slip on outside. Mound A, Trench
I, level 3.
(2) Gray ware. Outside black from firing. Knotted grass pattern on rim.
Mound A, Trench I, surface.
(3) Black ware. Lightly burnished and with traces of red paint. Mound A,
Trench I, surface.
(4) Black ware. Lightly burnished red slip inside and out. Mound A,
Trench I, surface.
(5) Gray ware. Red slip inside and out. Mound A. Trench I, surface.
EXCAVATIONS AT BIGO, 1957 25
(6) Black ware. Highly burnished red slip inside and out, with black patches
from firing. Mound A, Trench I, level 3.
(7) Red ware. Red slip inside and out. Mound A, Trench II, level 1.
(8) Black ware. Burnished red slip inside and out. Mound A, Trench I, hole
(9) Gray ware. Lightly burnished red slip inside and out. Knotted grass
pattern on outside. Mound A, Trench I, hole in natural.
(10) Black ware. Lightly burnished. Mound A, Trench I, surface.
(11) Gray ware. Lightly burnished red slip on outside. Knotted grass pattern
below rim on outside. Mound A, Trench I, level 1.
Bigo. Iron Instruments.
(b) Small Finds
A number of iron objects were discovered, all save one coming from Trench
I. These objects were mainly knife and razor blades and arrow heads but also
included one hoe. They were mostly in a state of advanced decay. The best
preserved are illustrated in Fig. 14.
(1) Iron razor. Uganda Museum No. A.59.28.
(2) Iron razor of a type still used in Ankole. No. A.59.24.
(3) Iron arrow head of a type used for bleeding cattle. No. A.59.21.
(4) Iron arrow head. No. A.59.16.
(5) Iron spear head. No. A.59.23.
26 UGANDA JOURNAL
In addition to the iron objects a copper bangle (59.44) was discovered un-
stratified, and one head was found on the surface. In view of the positions in
which they were found no importance can be attached to these two items.
No stone artifacts were found but a number of loose pieces of rock, some of
which had been used as rubbers were submitted to the Geological Survey for
determination as the presence of non-local rock might have been of signifi-
cance. Dr. Du Bois kindly reported on the specimens as follows:
"19,611 Buff-coloured fine-grained quartzite.
19,612 This rock is greissen and consists of coarse quartz, muscovite,
19,613 Dark green, almost black, medium grained amphilolite.
19,614 Pink, medium-grained, partially weathered granite with coarse
felspathic pegmatite vein.
19,615 Fine-grained, tourmalinized, slightly felspathic quartzite.
19,616 Coarse-grained, glassy, recrystallized quartzite.
19,617 Coarse granitic pegmatite.
19,618 Greissen similar to 19,612.
19,619 Highly tourmalinized quartzite.
19,620 Coarse, grey, recrystallized quartzite.
These rocks are typical Karagwe-Ankole quartzites, post Karagwe-Ankole,
granites, or other rocks of Karagwe-Ankole age modified by tourmaliniza-
tion associated with the post Karagwe-Ankole granites.
No rock is sufficiently distinctive to suggest that it had any origin other
than a local one, whilst the general character of the suite as a whole suggests
it was collected within, say, a ten-mile radius of Bigo."
The number are those given to the specimens by the Department of
Geological Survey, Uganda.
(d) Animal Bones
A considerable number of animal bones were found-mainly in the fill of
the ditch. These have very kindly been identified by Mr. Poole, Mr. Wasawo,
and Mr. Bosa of the Department of Zoology at Makerere College.
The animals represented were:
Trench I. Level 1. Forest Hog or Wild Pig; Buffalo; Ox; Goat; Buck; Chicken.
Above Ditch Fill. Ox; Goat;
Ditch Fill. Ox; Goat; Chicken; Hare.
Trench II. Level 1. Goat; Ox; Buffalo(?).
3. Ox; Goat; Chicken(?).
4. Ox; Goat.
Note. The animal referred to as Ox was one which was very similar (as
regards teeth and jaws) to local present day African cattle.
There is nothing unusual in this fauna and similar remains would no doubt
be found in the rubbish heaps of modern villages.
EXCAVATIONS AT BIGO, 1957 27
The modest excavations described here do not go far to solve the archaeo-
logical and historical problems presented by the large group of earthworks in
western Uganda, ably described and listed by Lanning. We are no nearer to
solving the Bachwezi problem but do now know something of the material
culture of the people, whoever they may have been, who lived at Bigo.
Even the most superficial study of the material now made available shows
that Bigo was certainly built and lived in by an African population not
markedly different in culture from the present inhabitants of the area. They
knew the use of iron and made tools and weapons some of which resemble
types now in use in western Uganda. The pottery shows considerable differences
in detail from the modern pottery of the country, and has certain characteristics
(particularly in the red finger-tip painting) which should make it easily recog-
nizable on other sites, but in its general character it is clearly of purely
African inspiration and, wherever its makers may have come from, there has
been no foreign, that is non-African, influence.
The only indications as to who these Bigo people were are to be found in
the iron arrow head (Fig. 14 (3)) which is very similar, if not identical, to the
specialized arrow head used nowadays by the Bahima for drawing blood from
the neck of cattle. This blood forms an important element in the diet of the
Bahima as it does in that of many other of the cattle peoples of East Africa.
The peculiar pot (Fig. 10 (1)) described tentatively as a fumigating pot may, if
this identification is correct, also suggest Bahima influence in view of the well
known practice of fumigating the milk vessels as a means of cleaning them.
Something can now be said as to the purpose of Bigo. Prior to the excava-
tion it was not clear even that there had been any considerable occupation and
the view had been expressed that the ditches were only for use as a vast cattle
pen. The scarcity of occupation material within the large area enclosed by the
outermost ditches suggests that that portion may well have been used mainly
for cattle, but there can now be no doubt that there was considerable human
occupation at the centre of the complex.
The purpose and function of the mounds remains most obscure and until
they have all been adequately excavated little more can be said. Mathew
(op. cit.) has suggested that our Mound A was raised as a base on which to
build a hut and in default of other evidence this suggestion may be allowed to
stand. Clearance of the top soil from the mound would reveal the post-holes
if a hut had ever stood there. If there was a hut, it must surely have had ritual
or royal significance to have merited the large mound erected to serve as its
As already pointed out, this excavation has not solved the historical problem
of Bigo or of the Bachwezi. But it has, for the first time in Uganda, produced
properly attested material of a past culture. The date of the site still remains
uncertain but, in default of precise information from archaeological sources, the
approximate traditional date of about 500 years may be allowed to stand.
It is unlikely that archaeology can do more to establish a chronology in view
of the lack of foreign imports of known date. But, though it is not impossible
that such objects may be found if further excavation is undertaken, the distance
28 UGANDA JOURNAL
from the coast and the known isolation of Uganda from outside contacts make
this a matter of some improbability.
What can now be done is to define a Bigo culture in material and geographical
terms. The material of this culture is known and an obvious next step is to
make careful surface surveys, with perhaps some excavation, on the other
earthwork sites so that their identity or otherwise with the Bigo culture can be
(1) Antiquity, 27 (1953), 215-17.
(2) Uganda J., 17 (1953), 51-62.
(3) Uganda J., 2 (1934-5), 21ff.
(4) A preliminary account of the excavation and reproduction of this plan has
been published. Antiquity, 33 (1959), 54-7.
(5) Mathew op. cit. Plate III and Cole, S. M. (1954). Prehistory of East Africa,
(6) Uganda J., 2 (1934-5), 226-33.
(7) Uganda J., 22 (1958), 11-17.
(8) Antiquity, 27 (1953), 216-17.
NOTES ON THE KIBUGA OF BUGANDA
By P. C. W. GUTKIND
The Waganda consider their country to be where the King is, and if no
Kabaka of the Royal blood is installed in Mengo, the result would be a
break-up of the people.
Captain F. D. Lugard in Africa. No. 2 (1893). C. 6848, p. 36.
THE dispute between John Hanning Speke and Richard Burton in the
middle of the last century, as to whether the former was right in claiming
Lake Victoria as the source of the Nile, drew attention to the kingdom of
Buganda in the inter-lacustrine region of east central Africa and to the auto-
cratic state which impressed outsiders with its efficient and powerful political
Before the eighteenth century the Ganda tribe had a flexible political struc-
ture in which local authority rested in the hands of the clan, though suzerainty
of the kabaka,' or king, was recognized. In the eighteenth century a more
centralized government under the kabaka emerged which resulted in con-
siderable consolidation of power and, at times, found expression in absolute
One result of the centralization of political power was the increasing import-
ance attached to the kibuga, or capital, of the kabaka. The kibuga was usually
located on a hill-top which could easily be defended against attacks and which
offered routes for escape. The kabaka moved the kibuga frequently3 and its site
was invariably changed at the death of a kabaka, when the lubiri, or palace,
was immediately deserted. For the last 150 to 200 years, the kibuga had re-
mained somewhere within a small area of central Buganda, not far from the
shores of Lake Victoria on Murchison Bay. The mobility of the kabaka was
indicated by the fact that from the time of the death of kabaka Suna II in 1856
to the arrival of Captain (later Lord) Lugard in 1896 the location of the kibuga
was changed at least ten times.
Ahmed bin Ibrahim el Ameri in 1844 was the first Arab to reach the kibuga
of kabaka Suna II (whose reign started in, approximately, 1831).4 He was
followed by Saim, an Arab half-breed, whose party of Arabs and Swahilis
helped Suna's army to defeat the neighboring Soga. Saim was followed by
Isa bin Hussein who arrived in 1850 and subsequently became Suna's personal
bodyguard and was given land by Suna in the county of Buddu. When Snay
bin Amir met Burton at Tabora in 1858, he gave him a brief account of kabaka
Suna's kibuga but did not mention the place where it was located.5
Speke, the first European to visit the kibuga, found kabaka Mutesa I at
Banda-Balogo in 1862, and H. M. Stanley found him at Rubaga in 1875. When
the Rev. C. T. Wilson arrived in 1877, he too found him at Rubaga but he also
occupied another palace at Nabulagala (Kasubi). But since approximately
30 UGANDA JOURNAL
1885, when kabaka Mwanga came to the throne, the kabaka has lived on
Our knowledge of the early history of the kibuga has been limited to brief
descriptions and snatches of information recorded mostly in travelogues,
government documents, mission records and diaries. It was not until 1911,
when the Rev. John Roscoe, of the Church Missionary Society, published a
detailed account of the customs, history and beliefs of the Ganda tribe of the
Protectorate of Uganda,7 that a more detailed description of the kibuga
appeared. Drawing on the memory of the katikiro,s Sir Apolo Kagwa, Roscoe
reproduced two plans, drawn by Kagwa, of the kibuga9 and of the lubirio1 "as
they were known during the reign of King Suna and in the early years of King
The plans indicated the location of the kibuga; the enclosures of the prin-
cipal chiefs were laid out on the side nearest to the chiefs' own districts.12 The
kibuga is in the saza, or county, of Kyadondo, one of twenty in the king-
dom of Buganda. Each saza is divided into a number of sub-counties or
gombolola; of these there are eight in Kyadondo, one being the kibuga. Each
gombolola is in turn sub-divided into a number of parishes or miluka'3 of
which there are twelve in the kibuga.
On Apolo Kagwa's map the kibuga on Mengo hill was contiguous, to the
east with Makerere hill; to the south it included the flat-topped hills of Bukesa
and Namirembe; and to the west the smaller hills of Masanafu and Lugala.
The kibuga is well-watered; to the north the river Lubigi flows from east to
west; while the rivers Jugula and Nsoba start from a deep spring close to the
lubiri and flow north to join the Lubigi. In the west the smaller rivers of
Nabisasiro and Muga also originate from deep springs near the lubiri.
Of the kibuga Roscoe wrote:
The capital was divided into a number of sites corresponding to the country districts;
every leading chief was surrounded by the minor chiefs from his district, and a portion
of uncultivated land was left on which peasants could build temporary huts when they
were required to reside in the capital for state work. By this plan all the people from
a particular district were kept together, and the sites remained the official residences of
the chiefs of the district to which the sites belonged. Chiefs built high fences of reeds
round their estates in the capital; the fence bordering on the main road leading to the
capital was always neatly finished, and the space in front of the gate was kept swept
and free from weeds. Within the enclosure there was a considerable amount of land
cultivated, with plantain trees which were well cared for, and the fruit of these trees
supplemented the food which was brought up from the country estate, and also supplied
the table in any emergency. Every chief built a number of houses within his enclosure,
not only for his own use and that of his wives, but also for slaves and retainers, and a
supply too for casual visitors who might wish to stay with him for a day or two. When
a chief was promoted, or deposed, he had to leave his site on which he had been living
within a few hours; this was so ordered to prevent chiefs from using all the food on
the site, and from causing destruction to the property.14
The Use of the Word Kibuga
Although the word kibuga is translated as capital, today it may also be used
for any town or area with a large number of people living under non-rural
conditions. The word is derived from the Ganda verb okwebuga-to walk to
and fro, or a busy place. Early writers were not very explicit in the use of
NOTES ON THE KIBUGA OF BUGANDA 31
Its first use in the literature available appeared in Burton's The Lake Regions
of Central Africa, who referred to "Kibuga, the capital of Uganda",16 and also
"Kibuga, the capital district of Uganda"7 and quoted Arabs whom he met
at Tabora in 1858 as saying, "Kibuga is the residence of the Great Mkama or
chief of Uganda".'8
Speke translated the word as palace19 and Grant spoke of the "capital of
Uganda-named Kibuga".20 Stanley, who visited the court in 1875, also used
the term to mean palace:
In the afternoon I was invited to the palace . Court after court was passed until
we finally stood upon the level top in front of the great house of cane and straw which
the Waganda fondly term Kibuga or the palace.21
Cardinal Lavigerie too said that "La residence royale porte le nom de
Kibouga".22 R. P. Ashe, the C.M.S. missionary, reported that kabaka Mwanga
ordered him to come up at once to the Kibuga (palace)";23 yet in the same
account he employed the word to describe an area greater than that of the
palace by translating kibuga as the capital.24 In 1905 J. F. Cunningham referred
to kibuga in quite a different sense, i.e., in connection with burial customs of
There still remained the under jaw. The executioner appointed a chief to make a
"kibuga" (enclosure with a few houses in it) and the jaw .was placed in one of the
houses in the "kibuga" . .25
Roscoe, in 1922, defined kibuga as the residence of the kabaka:
The native capital commonly goes by the name of Mengo, and the residence of the
king, or Kabaka, is known as Mbuga, a name which is used to signify not only the
regular royal enclosure on the top of the hill Mengo, but any place where the king
happens to be in residence . .26
Mengo, a name often used interchangeably with kibuga, appeared to be a
generic term given to any hill occupied by the kabaka. Leclercq, in Aux
Sources du Nil, stated this quite precisely:
Au temps de Mt6sa, ce nom de Mengo s'appliquait non a la colline royale actuelle,
mais A ce que 6tait alors la colline royale, laquelle se d6place suivant le caprice de
chaque roi, mais porte toujours, quelle qu'elle soit, ce nom du Mengo.27
Finally, the term, kibuga, as a political and social entity, meant different
things to different levels of Ganda society. To royalty the kibuga was the centre
of power; to the chiefs a place in which they were compelled to live at certain
times of the year; to minor chiefs a place to be visited when so ordered by the
more senior chiefs; to labourers a place where they served the kabaka and
chiefs in the execution of public works; to the peasants a place to be avoided;
and to all it indicated the heart of the kingdom, the seat of authority and
Kibuga: Village or Town
Ashe described this area as "a vast village of gigantic straw built houses"28
and as "one of the great capitals of Africa".29 Dr. W. Junker, the African ex-
plorer, spoke in 1886 of the kibuga as "neither a village nor a town"30 yet "many
32 UGANDA JOURNAL
miles in circumference".3 Colonel Colvile, in 1895, referred to the "capital of
Uganda" as being "built on four hills"32 and Martin Hall, writing in 1898 (pro-
pounding the Roman analogy), said of Mengo, "If Rome was the city of seven
hills, Mengo quite outshines it, for it has as least thirteen hills each one of
which has its special importance"." Sir Harry Johnston also described the
"native capital" as being "like ancient Rome-only much more so-a city
of seven hills34 . Each suburb or portion of the straggling town . a
hillock in itself".35 Leblond, in 1912, spoke of "cette important m6tropole ...
6tait plut8t un ensemble de villages qu'une ville proprement dite",'6 and
Leclercq said, ". . mais tout cet ensemble d'agglom6rations demeure
invisible, parce que Kampala est moins une ville qu'un immense jardin".37
Before the arrival of Captain Lugard in December 1890, visitors to the
kibuga had to obtain permission from the kabaka to enter.38 Visitors were kept
under close supervision and obliged to, live where directed. Consequently they
had little opportunity for free movement about the country. Lugard paid no
attention to this and he crossed the Nile without warning making a forced
march to the kibuga, then at Mengo, and insisted on camping on a hill of his
own choice irrespective of the kabaka's wishes.39
The Approaches to the Kibuga
On approaching the kibuga, Speke declared, "It was a magnificent sight. A
whole hill covered with gigantic huts such as I had never seen in Africa
before."40 A. M. Mackay told us how the approaches to the kibuga were
The neighbourhood of the capital is a series of hills, conical or table-shaped, inter-
sected by swamps, which they call rivers. All the paths approaching the capital must
cross these swamps, a low bank having been generally made on which to cross, water-
ways being left here and there, across which one or two logs of wild palm are thrown.
As all coming to or going from the capital must cross at these points, the executioners
lie in wait there, and capture unwary serfs, whom they tie up for execution.41
Grant stated that "within a radius of thirty miles from the palace nothing
is allowed to be plundered",42 and Jonveaux, a French litterateur who never
actually visited Buganda,43 but annotated the work of others, described how,
in contrast to the rest of the kingdom, there was no evidence of disorder on
approaching the kibuga.
S. we approached the Kibouga or royal residence, and I could not fail to notice
the absence of any signs of disorder: acts of violence being severely punished if com-
mitted within a range of six leagues round the palace. Herds of cattle and flocks of goats
were feeding quite peacefully by the side of the road. Not one of the escort dared lay
his hand on them, although provisions were very scarce indeed.44
Chiefs in the Kibuga
Senior chiefs who "spent a large part of their time in the capital"45 were not
always safe there, as Speke's account indicated. Unfortunately he did not tell
us the reason for the "commotion":
The regions about the palace were all in a state of commotion today, men and women
running for their lives in all directions, followed by Wakungu and their relatives. The
cause of all this commotion was a royal order to seize sundry refractory Wakungu, with
their property, wives, concubines . and families all together.46
NOTES ON THE KIBUGA OF BUGANDA 33
But there were also advantages. In the kibuga the chiefs themselves, Roscoe
S. .looked to the katikiro for orders about work, which would consist either in keep-
ing up their own buildings, or in clearing roads. The katikiro would tell the King when
it became necessary to have roads cleared and a man would be appointed as overseer of
this work. This office was always eagerly desired by the chiefs because there was
profit to be made from it. The man appointed could sublet it if he was a chief, and yet
keep emoluments from the office himself.47
Roads in the Kibuga
All early visitors to the kibuga were greatly impressed with the wide and
well maintained roads which entered the kibuga from all sides. Stanley, for
example, said there were:
S. .broad avenues, imperial enough in width . we came up to one of these avenues,
the ground of which was a reddish clay, strongly mixed with the detritus of hematite.
It gave a clear breadth of 100 feet of prepared ground, and led by a gradual
ascent to the circular road which made the circuit of the hill outside the palace
C. T. Wilson and R. W. Felkin, both C.M.S. missionaries, told of a "magni-
ficent road some eighty yards wide, leading straight up to the palace"49 and
Ashe spoke of a "magnificent sweep of road . between lofty fences".50 In
the kibuga the roads:
S. were kept in order by those whose enclosures adjoined them . .The principal
roads were about twenty yards wide, others were narrower, while the small branch roads
were not more than three yards wide. The katikiro would send orders into every part of
the country telling the people to come and repair the roads.51
The Royal Residence
Once well inside the kibuga the visitors were much impressed by what they
saw. Speke, for example, when he entered the lubiri was "quite surprised" by
the palace's extraordinary dimensions, and the neatness with which it was
kept".52 Stanley summed up his observations by recording:
But this man builds upon a hill that he may look abroad and take a large imperial
view of his land. He loves ample room; his house is an African palace, spacious and
lofty; large clean courtyards surround it . .a cane enclosure surrounds all, and
beyond the enclosure again is a wide avenue running round the palace fences.52
Hutchinson records that the "flagstaff and an immense flag proved a decided
feature in the landscape" rather than the huts "however large" which added
"but little attraction",54 a view shared by Gedge, of the Imperial British East
Africa Company, who, on arriving in the kibuga, found the royal enclosure to
be a "miserable collection of huts, and dirt and filth reign supreme there".55
Very shortly after the Rev. R. H. Walker recorded a similar observation in his
S. none are better houses, and only one of them as large as mine was. None of
Mwanga's are as large as the Pokino's at Masaka.56
Ashe, when he saw the kabaka for the first time, was much impressed as he
S. along a beautiful lane, between lofty fences, "bisakati", formed of the tall tiger-
grass, which grows so abundantly in Buganda, and is used to surround the houses and
gardens of the upper classes.57
34 UGANDA JOURNAL
Fahs thought the "royal hill, superb in itself" and made "especially impos-
ing by the tall fence of tiger grass enclosing it on either side".58 The lubiri itself
. a building forty feet in height and supported on each side by straight wooden
pillars . .with its seventy feet of length the structure was easily the largest in the
However, it was left to Roscoe, to give the most detailed account of the
The king lived upon a hill situated in the neighbourhood of the lake. The summit of
the hill was levelled, and the most commanding site overlooking the country was
chosen for the king's dwelling houses, court houses, and shrine for fetishes, and for
the special reception room. Round these buildings on the lower slopes of the hill other
houses were built; in front were huts for the guards and retainers, and the many houses
for the wives, their maids and slaves, were built on the sides and at the back of the
royal houses. The whole of the royal enclosure was divided up into small courtyards with
groups of huts in them; each group was enclosed by a high fence and was under the
supervision of some responsible wife. Wide paths between high fences connected each
group of houses with the king's private enclosure. In the reign of the famous king
Mutesa there were several thousand residents in the royal enclosure; he had five hundred
wives each of whom had her maids and female slaves; and in addition to the wives
there were fully two hundred pages and hundreds of retainers and slaves. A high fence
built of elephant grass surrounded the royal residence, so that it was impossible for an
enemy with the ordinary primitive weapons to enter. At intervals round the outside
of the enclosure guard houses were built; there were four or five entrances which were
strongly guarded; both inside and outside were huts with soldiers always on duty, to
prevent any person except the slaves and wives from entering. Again inside the enclosure
near each of the gates were other guard houses with soldiers on duty, who had instruc-
tions as to what persons were to be admitted by the gatekeepers. The main entrance in
front of the royal residence was the only way by which the public were allowed to enter
or leave the court. All the land between the royal residence and the lake was retained
for the king's wives and here they grew their plaintains.60
Size of the Kibuga
How large was the kibuga? Was it smaller or larger than the area covered
today? What was its population? Accounts vary. The answer is almost cer-
tain to be that it was a small and compact area with clearly defined boundaries.
The kibuga was one administrative unit under the control of the katikiro;
within it there were smaller divisions roughly equivalent to the area occupied
by members of the royal family, senior officials of the court, stewards and
The earliest reference to the size of the kibuga was given by Burton, quot-
ing Snay bin Amir, who had visited Buganda on several occasions between
1850 and 1858.61
Kibuga, the settlement, is not less than a day's journey in length; the buildings are of
cane and ratten. The sultan's palace is at least a mile long, and the circular huts, neatly
ranged in line, are surrounded by a strong fence which has only four gates.62
Dr. (later Sir) Albert Cook paid a visit to Mwanga's enclosure in February
1897 which he described as a "vast oval enclosure ... about two miles round".63
Thomas and Spencer reported that the kibuga in 1907 and 1908 covered an
area of "about 20 square miles",64 as it does today, and Roscoe, writing in 1911,
but of an earlier period, said:
NOTES ON THE KIBUGA OF BUGANDA 35
There was one plan followed, which has been used by the kings for years without
variation. The enclosure was oval shaped, a mile in length and half-a-mile wide, and
the capital extended five or six miles in front and two miles on either side.65
Apolo Kagwa, writing in 1918, said that the "oval enclosure", i.e., the palace
alone, was about "1105 by 1122 yards by European measure".66 Roscoe, writ-
ing again in 1922, revised his earlier statement and referred to the "native town"
as "extending from the Kabaka's enclosure on the top of the hill Mengo fully
a mile to the north, east and west . .".67 Howell, writing about the kibuga
at the time when Roman Catholic missionaries arrived, in 1897, referred to the
court of the kabaka as an "enclosure surrounded (by) land a mile long and
Some visitors to the kibuga gave an indication of its size by reference to the
large number of houses which covered the slopes and hill-tops. Speke, for
example, reported that the whole area was "covered with gigantic huts"69 (but
gave no number), but Grant spoke of "several hundred houses".70 Hutchinson,
who wrote in 1876, merely referred to a "vast collection of buildings"71 and
Lavigerie said categorically that the royal enclosure alone "ne compete pas
moins de quatre ou cinq cents huttes".72 Apolo Kagwa plotted 496 houses in
the royal enclosure73 and "over a hundred houses"74 belonging to the katikiro
plus approximately 370 plots assigned to officers of state, chiefs and members
of the royal family.75
There have also been estimates of the size of the population. Burton, for
example, in 1858, when given a brief description of the kibuga, was told that
"the Harem (of the kabaka) contains about 3,000 souls . .".76 Linant de
Bellefonds, an emissary of Gordon, on his arrival in the kibuga in 1875, re-
ported that "more than 10,000 people surround us".77 Sir Harry Johnston esti-
mated the population at 77,000 in 1900,78 and Deputy Commissioner George
Wilson, in 1906, as 60,000.79 Sir Frederick Treves estimated 77,000 in 191080
and Roscoe, writing in 1911, yet of an earlier period, reported that "not less
than three thousand people (were) living in the royal enclosure".81 An official
population census in 1911 returned a kibuga population of 32,44182 yet
Uganda Notes, published by C.M.S. missionaries, reported in the same year
a population of 14,852 which had risen to 15,274 by 1912.83 In 1913 Leblond
wrote, "on comprendra qu'h certain moments il pouvait se trouver, aux
alentours de Mengo, de quinze a vingt mille personnel".84 And in the same year
Leclercq referred to "cette m6tropole" with a population of 60,000.85 In 1948
the kibuga population was 34,337,86 a remarkably small increase over the
32,441 of 1911.
These early estimates must be treated as no more than a guide to the size
of the population and the area covered by the kibuga. To obtain an accurate
population count must have been difficult (although in May 190487 a law for
the compulsory registration of African births and deaths in Buganda was
passed and a more effective Birth Registration Law was enacted in 1923),88
because of the mobility of the kibuga population. Ashe, for example, referred
to the "ever moving throng"89 of people in the kibuga, and in 1895 Lugard
referred to the roaming population to be found in this area.90 By 1907 and
1908 the boundaries of the kibuga were clearly demarcated, so that Thomas
36 UGANDA JOURNAL
and Spencer's statement that the kibuga covered some twenty square miles can
be accepted as accurate.91 This would indicate that Natete in the west, Kibuli
in the east, Mulago in the north and Kabowa in the south were all part of the
The Labour Force in the Kibuga
We know from several accounts that a large number of labourers were work-
ing in the kibuga on the kabaka's works, keeping "an army of men employed
the whole year round", and serving the chiefs whose establishments often com-
prised a "total of a thousand people".92 Living houses, cook- and wash-
houses and ancillary buildings had to be constructed and maintained. Fences
needed constant repair as did roads, bridges and culverts. Those called upon
to perform these services "did not care to live long in the capital"93 not only
because of the difficulty of obtaining food,94 or because they had to work hard,
but because they were in "danger of being seized and put to death".95 Walker
referred to the "lesser fry round Mengo",96 and was the first to give an idea of
actual numbers employed on a building project in the kibuga.
The King has lately been enlarging his pond here at the Capital . .it is a poor thing
though 2,000 men have been employed upon it for many months.97
Captain J. R. L. Macdonald, writing in 1897, about observations made
between 1891 and 1894, said:
.. large numbers of workmen were brought annually to the capital to refence the
royal enclosure and repair or renew the palace buildings.98
The construction of the Protestant and Catholic churches called for the
employment of "hundreds of people".99 It was only after the turn of the century
that the Government became the major employer of unskilled labour for
capital development projects. By 1907 "it was estimated that over 10,000 men
were used daily as porters in Kampala alone",'00 and between 1905 and 1909
the Buganda Government too maintained a substantial labour force of approxi-
mately 300 to 400 labourers for road works in the kibuga.10' Most of these
labourers no doubt lived in or close to the kibuga.
In 1914 the editor of Uganda Notes said that the kibuga population:
S.. though a large one is a very shifting one, numbers of people coming to work for a
time for their chief and then returning to the country. The fixed population of the capital
is by no means a large one. Some chiefs always live here owing to their duties, but no
one lives in the capital who can help it . .102
Powesland recorded that at "the end of 1921, hundreds of men were looking
for work in Kampala"'03 and in 1923 the Public Works Department reported,
"A floating labour population seems to be coming into being."104 Again in
1927 the records indicated that "hundreds of labourers were to be found
squatting in and around Kampala . .".105
The kibuga population was evidently a heterogeneous one. In Mutesa's
time slaves from neighboring tribes were to be found as servants to chiefs,
and in the 1880s and 1890s soldiers from a distance were quartered in the
kibuga. Ashe reported that in the kibuga there "were people of all shades and
___ ___ **-^ ---- ---- --- *' ~-- -----------
From: John H. Speke, Journal of the Discovery of the Source of the Nile, 1863.
View of King Mtesa's Palace from my Hut-Uganda.
From: Col. Charles Chailld-Long. Central Africa: Naked Truths of Naked People. 1876. p. 102.
Arrival at the Palace of M'ts6.
K .. r ,,1 .. C -
Fom: The Diary of the Reverend C. T. Wilson of the Church Missionary Society.
(Reproduced by kind permission of the Church Missionary Society.)
From: The Story of the Life of Mackay of Uganda Told for Boys, 1891. D. 172.
The Capital of Uganda at the time of Alexander Mackay.
DE LA CAPITAL a
DCI LUSAFJDA s.^
ota. La riride.r r. d. cd, in w -A
lnn r u qu 'iat ocrupapi tn nuILt la
_JL i Liiiv 'n --de .-Jan r .-,,l)2 -- -' _
:atd .vi"llape ILA. .- .
r.- -- :4 wi
''' f ._
-, ,;e Residene bt
M A a A "--
From: G. Leblond. Le Pere Auguste Achte. Maison-Carrde. Algiers. 1912. p. 131.
from: nlr Aloert LOOK. uganaa Memories, lt-, p. 0,.
Kampala from Namirembe Hill, 1897: Lugard's Fort in middle distance.
NOTES ON THE KIBUGA OF BUGANDA 37
complexions from a light copper colour to a jetty black".106 Dr. Ansorge,
writing in 1899, spoke of a "Swahili settlement .. ." and in 1902 Uganda
Notes reported on a "mixed population within a mile of the centre of our
mission (which) is a gradually increasing population of Swahilis, Arabs,
Indians, Soudanese, as well as Mohammedan Baganda".108 In 1908 Sir Hesketh
Bell, the Governor, wrote that "a considerable number of people from Toro,
Bunyoro and Ankole find there way down to Entebbe and Kampala in search
of work .. .".109
It could well be that it was in response to the needs and demands of a large
labour force that Mackay, in 1882, suggested1 to kabaka Mutesa the establish-
ment of a market in the kibuga. Markets were unknown in Buganda,'" so it
is no wonder that Mackay's suggestion was considered an innovation and
aroused great curiosity.
Immediately it was decreed that, there and then, an enclosure was to be built in the
palace ground where people could buy and sell: but so ridiculous were their ideas of
barter that the court not only decreed that any one selling anything anywhere else would
be chopped in pieces, but they agreed at this sitting what was to be the price, in cowries,
of every article.112
For a number of years the only markets in Buganda were located in the
kibuga and increasingly served the needs of the large population. By June
1903, for example, it was reported that "the Native market is growing apace,
and it is now quite four times its former size"."3
The Traditional Administration of the Kibuga
The traditional political administration of the kibuga, be this defined as the
area covered by the royal enclosure or the larger area covered by the kibuga,
differed from the rest of the kingdom as it was directly under the control
of the katikiro whose power and dignity virtually equalled that of the kabaka
himself. In the kibuga the reins of political control were held more tightly by
the kabaka and katikiro than in the rest of the kingdom where, because of the
distance from the kibuga, it was possible for chiefs to exercise considerable
unchecked power. Outside of the kibuga the great bulk of the population were
peasants who were subject to economic exploitation and the whims and fancies
of the chief to whom they attached themselves and whose land they cultivated.
The chiefs in turn were expected to render aid and protection to the peasants.
When resident in the kibuga, the chiefs were subordinate to the katikiro, and,
in some matters, to the kago,114 the county chief of Kyadondo. The katikiro
was assisted by the kibale and kisekwa. The office of the kibale might best be
described as that of a deputy to the katikiro, although tradition suggests that
his duties were predominantly those of a Lord Chamberlain to the royal house-
hold. The kibale was responsible for much of the day to day administration
of the kibuga, and in particular dealt with all judicial matters in which the
kabaka, members of the royal family and the more important chiefs were in-
volved. When the kabaka and the katikiro were away the kibale, acted as
chief regent of the kingdom. His other duties included the supervision of the
chiefs living in the kibuga, the distribution of plots within the kibuga and the
38 UGANDA JOURNAL
supervision of road construction. This latter function he seemed to have main-
tained for some years after the 1900 Agreement came into force. The kibale
played a considerable part in the allotment of kibuga land to the senior chiefs
and other important persons. Today the kibale is the only person who is
allowed to inform the kabaka of the death of a member of his family; except
for this function the duties of this office are extinct.
Another important office in the kibuga was that of the kisekwa, whose func-
tions can best be described as those of a judicial officer and magistrate. It is
not clear whether his judicial functions covered the whole kingdom or whether
he was only responsible for trying cases arising among the residents of the
Because, from more recent Buganda history, it appeared that the katikiro
was a very busy man and often had to join the kabaka on numerous war
expeditions, it is quite possible that the office of the kisekwa was of relatively
recent origin. Some of the older informants said that the office was established
at the time of kabaka Suna, to take some of the judicial functions off the
shoulders of the katikiro. Otherwise, the kisekwa had no administrative respon-
sibility in the kibuga.
When, in 1900, the kisekwa resigned from his office, his deputy took over his
duties for the times being, and served as magistrate for the kibuga until 1913,
when a special court for the kibuga was gazetted and presided over by Mr. P.
Balintuma, the first asoloza ekibuga, or tax collector, of the kibuga."5
Some Early Social Problems in the Kibuga
The kibuga suffered a great deal of destruction during the religious wars,
between 1888 and 1892. In 1893 the Catholic Union of Great Britain reported
that the kibuga was "a smoking heap of ruins; at least all the Catholic houses
have been fired."16 Two years later Captain Lugard wrote in a similar vein:
There is no one left to till the fields. Besides numerous deaths there are a number
of women who have not been seen since the war of 1892. People, moreover, who have
lost everything are not likely to have cowries enough to buy food for themselves in the
A more hopeful note was struck by George Wilson in 1907, when he was
able to report that the . native capital, Mengo . shows gratifying
evidence of progress .".18 Yet six years later the editor of Uganda Notes
in three months no less than eighty native houses have been destroyed (by
incendiarism) in the immediate neighbourhood of the capital.19
Such conditions naturally contributed to a host of social problems in the
kibuga. A brief comment will demonstrate this.
Roscoe reported that kabaka Suna "made the neglect of certain sanitary
conditions in the capital an offence punishable by death (and) he had put a
number of persons to death for the breach of the rules in question".120
Although Mackay, Wilson, Felkin, Ashe and Walker described, in their
NOTES ON THE KIBUGA OF BUGANDA 39
reports to the Church Missionary Society in London, sanitary conditions in
the kibuga as "squalid", "slimy", "foul", "unclean" and "grimy", the ready
use of these terms may be attributed to missionary zeal. However, plague
broke out frequently, and Howell described how, as a result:
People fled in droves (from the capital), and those who could not go away drove
from their homes those who were stricken with the dread disease. The sick were uncared
for and corpses left unburied. The disease reached the Court and page after page fell
victim to it.121
Dr. Ansorge, writing to the Commissioner, Colonel Colvile, in October 1894,
referred to a camp of Swahili traders as a place of "filth and stench" and an
"incongruous mass",122 and Colonel Trevor Ternan, in 1897, complained to
Apolo Kagwa about the "deplorable conditions I see in the native capital".
Ternan was particularly concerned about the "dangerous conditions of many
of the houses both from the point of view of construction and sanitation".123
Sir Harry Johnston, however, expressed a somewhat different view, for he
observed that "sections of the town inhabited by the little king and his court,
the native gentry, and the common people are clean . .". The Ganda were
"one of the few Negro races", he continued, "who attempt anything like sani-
tary measures to keep their surroundings free from filth . the surroundings
of their houses (are) very clean."'24
In 1899 Ternan took firmer steps to deal with the "danger presented by the
insanitary conditions in the kibuga" when he signed a notice under article 99
of the Queen's Regulations, Africa Order in Council, 1899, which set out rules
for the prevention of small pox and "the control and purifications of buildings
and areas, and the general improvement of the sanitary conditions of the
district and its inhabitants". The Principal Medical Officer was given authority
over an area "contained within a radius of 3 miles from Mengo",125 i.e., the
greater part of which is now, and, presumably, was then, the kibuga. Although
Apolo Kagwa expressed the view that this legislation "gives the Europeans . .
great power over our kibuga", he was evidently grateful to the Government
for the help "to destroy this very bad sickness".126
Sanitary conditions in the kibuga had evidently deteriorated, and health
measures were not enforced effectively, despite Ternan's order, for in August
1901 Bishop Tucker requested Sir Harry Johnston to "consider more effective
measures to improve the deplorable conditions in the native town, remove
unsightly hovels everywhere close to Kampala, and to protect the health of the
native residents from disease and constant sickness".27
An initial difficulty in any inquiry into the question of beer drinking in
Uganda arises from the use of the term mutamivu which in Luganda is used
to denote an habitual drunkard, or an occasional drinker, and by many it is
applied indiscriminately to all who are known to drink at all, on the ground
that in Buganda such a thing as moderate drinking is unknown.'28 Whereas
Grant reported, "There was very little drunkenness visible (in the kibuga)",129
C. T. Wilson talked of "a drunken race".130
40 UGANDA JOURNAL
Concern with the consequences of heavy drinking has consumed a good
deal of time, and since 1900 much legislation has been passed and a number
of official inquiries into excessive drinking have been held. Although drunken-
ness can certainly express itself in very undesirable behaviour, it has taken some
time for missionaries and administrators to understand, and accept, the socio-
logical importance of the preparation, the giving and consumption of beer in
ritual and social life. In the kibuga, in particular, the missionaries lost little time
in attempting to stamp out the "evil of drink", as they saw it. For example in
1901, as a result of sermons on this subject, delivered in Namirembe Cathedral:
S. .the chiefs are taking active steps to put it down . the native parliament has
decided to stop at once all selling of native beer in all the markets in Mengo, and also
at the roadside .. .131
When, after this decision was taken by the Great Lukiko, the Acting Sub-
Commissioner, Stanley Tomkins, was told of beer being sold in a market near,
or in, the kibuga "contrary to regulations":
He at once jumped on his horse, and followed by several men made a raid on the
place, the men breaking all the calabashes and gourds, and spilling the beer in all
Thirteen years later mission workers conducted a house to house survey
S. .the large amount of drunkenness there is in the Church and the so-called "Clubs"
for drinking purposes which exist in the villages round the capital.133
Roscoe, who lived in the kibuga for many years, summed up the changes,
which by the early 1920s, had overtaken it.
It (the native capital) was in those days a fine native town . but it has now become
a somewhat neglected and untidy place. The fences are broken down, the roads need
repair, and in many places there are large forlorn-looking houses . 134
But whatever the changes, the editor of Uganda Notes hardly exaggerated
when, in 1914, he wrote of the "immense influence that Mengo must always
exercise on the whole country" and that "nowhere else, with the exception of
Entebbe, are the Baganda more independent; and in no other part of the
country can individuals be so easily lost in the crowd".135 Certainly the
kibuga has changed from a small and compact area to a sprawling urban
1 Throughout this article the old orthography of Luganda, the language spoken by the
(Ba-) Ganda, is used where appropriate, i.e. for Ganda titles and offices. Although this
does not always make for easy reading, translation of a vernacular term into English is
often impossible without lengthy discussion, so that the retention of the Luganda term
is both more accurate and more convenient for the present purpose.
2 C. C. Wrigley. Buganda: An Outline Economic History. The Economic History
Review, 2nd series, vol. 10, no. 1, 1957, pp. 71-2.
3 C. T. Wilson and R. W. Felkin (1882).Uganda and the Egyptian Soudan, vol. i, p. 176.
4 J. H. Speke (1863). Journal of the Discovery of the Source of the Nile, p. 154: and
Sir John Gray. Ahmed bin Ibrahim-The First Arab to Reach Buganda, The Uganda
Journal, 11 (1947), pp. 80-3. See also: J. V. Taylor (1958). The Growth of the Church
in Buganda, p. 25.
ngn9 3g /Wyv
The Capital of Uganda. Plan drawn by Sir Apolo Kagwa, K.C.M.G.
From: J. Roscoe. The Baganda, 1911.
NOTES ON THE KIBUGA OF BUGANDA 41
5 R. F. Burton (1860). The Lake Regions of Central Africa, vol. ii, pp. 186-88.
6 Church Missionary Intelligencer, 1885, p. 727.
7 J. Roscoe (1911). The Baganda.
8 The Uganda Agreement, 1900, provided for the appointment of "Three native officers
of State" to assist the king in the Government of Buganda. The officers were: a prime
minister (sometimes known as chief minister) or katikiro; a chief judge or omulamuzi;
and a chief treasurer or omuwanika. (The Uganda Agreement 1900. Revised Edition of
the Uganda Laws, vol. VI, 1951, p. 16.)
9 Roscoe. The Baganda, Plan 1. (Reproduced as Fig. 7 of this article.)
10 Ibid, Plan 2.
11 Ibid, p. 523.
12 Ibid, p. 523.
13 Singular: muluka.
14 J. Roscoe (1921). Twenty-Five Years in East Africa, p. 192.
15 In the early accounts there was a tendency to use Kampala, Mengo and Kibuga
interchangeably in describing the environs of Kampala hill, now Old Kampala, and
Nakasero hill on both of which European settlements were established before the turn of
the century. Leclercq speaks of "Kampala est le nom sous lequel on la d6signe com-
munement". (J. Leclercq (Paris, 1913). Aux Sources du Nil, p. 155.) Today the name
Kampala is reserved for the non-African municipality; Mengo for the headquarters of
the Buganda Government and immediate environs; and Kibuga for the nineteen square
miles comprising one of the divisions of the county. Kibuga and capital can be, and
often are, used interchangeably.
16 Burton. The Lake Regions of Central Africa, vol. ii, p. 175.
17 Ibid. p. 186.
s1 Ibid, p. 188.
19 Speke. Journal of the Discovery of the Source of the Nile, p. 283.
20 J. A. Grant (1864). A Walk Across Africa, p. 208.
21 H. M. Stanley (1878). Through the Dark Continent, vol. i, p. 200.
22 S. Em. le Cardinal Lavigerie (Paris, 1886). Prbs des Grands Lacs, p. 49.
23 R. P. Ashe (1889). Two Kings of Uganda, p. 198.
24 Ibid, p. 104.
25 J. F. Cunningham (1905). Uganda and its People, p. 226.
26 J. Roscoe (1922). The Soul of Central Africa, p. 45.
27 J. Leclercq. Aux Sources du Nil, p. 160.
28 Ashe. Two Kings of Uganda, p. 50.
29 Ibid, p. 52.
30 Sir Harry Johnston's definition of a Uganda town was "a series of villa residences
surrounded by luxuriant gardens". The Uganda Protectorate (1902), vol ii, p. 656.
31 W. Junker (1890-2). Travels in Africa during the Years 1875-1886, vol. iii, p. 538.
32 H. Colvile (1895). The Land of the Nile Springs, p. 48.
33 M. J. Hall (1898). Through my Spectacles in Uganda, p. 57.
34 These three accounts, covering a period from 1895 to 1902, defined the kibuga in
terms of a number of hills ranging from four to thirteen. The differences did not reflect
a change in the size of the kibuga, but rather the view of the observers, i.e., what they
thought constituted the kibuga.
35 Sir H. H. Johnston. The Uganda Protectorate, vol. i, p.104.
36 G. Leblond (Maison-Carree, 1912). Le Pere Auguste Achte, p. 130.
37 J. Leclercq. Aux Sources du Nil, p. 156.
38 Speke. Journal of the Discovery of the Source of the Nile, pp. 283-4; and Grant
(1864). A Walk Across Africa, p. 216.
39 F. D. Lugard (1893). The Rise of our East African Empire, vol. i, pp. 373-7.
40 Speke. Journal, p. 283. See also Ashe. Two Kings of Uganda, p. 50.
41 A. M. Mackay, Pioneer Missionary of the Church Missionary Society. By his
sister (1890), p. 197.
42 Grant. A Walk Across Africa, p. 216.
43 H. B. Thomas. Emile Jonveaux-An Armchair African Explorer. The Uganda J..
10, (1946), 152-4.
44 E. Jonveaux (London, 1875). Two Years in East Africa: Adventures in Abyssinia
and Nubia with a Journey to the Sources of the Nile, p. 331.
45 L. P. Mair (1934). An African People in the Twentieth Century, pp. 177, 192.
46 Speke. Journal, p. 385; see also: Roscoe. Twenty-Five Years in East Africa, p. 192.
47 Roscoe. The Baganda, p. 241.
48 Stanley. Through the Dark Continent. vol. i, p. 199.
42 UGANDA JOURNAL
49 Wilson and Felkin. Uganda and the Egyptian Soudan, vol. i, p. 95.
50 Ashe. Two Kings of Uganda, p. 49.
51 Roscoe. The Baganda, p. 243.
52 Speke. Journal, p. 287.
53 Stanley. Through the Dark Continent, vol. i, pp. 201-2. See also: S. G. Stock (1892).
The Story of Uganda, p. 55.
54 E. Hutchinson (1876). The Victoria Nyanza, p. 15.
55 E. Gedge. Diary of 15 April 1890.
56 R. H. Walker. Diary, vol.XII, letter 177, 27 February 1892, p. 84, (Archives C.M.S.,
London). A picture, which appears in the Church Missionary Gleaner, vol. 22 (1895),
p. 68, shows the "Two-Storied House of Mwanga, King of Uganda".
57 Ashe. Two Kings of Uganda, p. 49.
58 S. L. Fahs (New York, 1907). Uganda's White Man of Work, p. 74.
59 Ibid, p. 75.
60 Roscoe. Twenty-Five Years in East Africa, pp. 88-9.
61 Burton. The Lake Regions of Central Africa, vol ii, p. 193.
62 Ibid, vol. ii, p. 188.
63 Sir Albert Cook (Kampala, 1945). Uganda Memories, p. 43.
64 H. B. Thomas and A. E. Spencer (Entebbe, 1938). A History of the Uganda Land
and Surveys, p. 40.
65 Roscoe. The Baganda, p. 200.
66 Sir Apolo Kagwa (New York, 1934). The Customs of the Baganda, edited by May
Mandelbaum (Edel), p. 74.
67 Roscoe. The Soul of Central Africa, p. 45.
68 A. E. Howell (1948). The Fires of Namugongo, p. 46.
69 Speke. Journal. p. 283.
70 Grant. A Walk Across Africa, p. 220.
71 Hutchinson. The Victoria Nyanza, p. 15.
72 Lavigerie. Pres des Grands Lacs, p. 50.
73 Roscoe. The Baganda, Plan 2; see also p. 366 where Roscoe gives the figure of
74 Kagwa. The Customs of the Baganda, p. 75.
75 Roscoe. The Baganda, Plan 1.
76 Burton. The Lake Regions of Central Africa, vol. ii, p. 188.
77 L. de Bellefonds. Itindraire et Notes. Bulletin Sociedt Khediviale de Geographie,
Cairo, 1876-77, p. 40.
78 Johnston. The Uganda Protectorate, vol. i, p. 104.
79 Entebbe Archives, Secretariat Minute Paper, 986/06.
80 Sir Frederick Treves (1910). Uganda for a Holiday, p. 211.
81 Roscoe. The Baganda, p. 366.
82 Uganda Census Returns, 1911, p. 2. This is the only census (others having been
held in 1921, 1931 and 1948) in which the kibuga is returned as one of the Chief Towns
of the country. As such it might well include Kampala Township, as the latter is not
83 Uganda Notes, vol. 14, no. 3, 1914. p. 55.
84 Leblond. Le Pkre Auguste Achte, p. 132.
85 Leclercq. Aux Sources du Nil, p. 155.
86 Uganda Census Returns, 1948.
87 Reprinted in 1927 Supplement to the Laws of Uganda, p. 480.
88 General Notice 282 on 1923, The Buganda Births and Deaths Registration Law,
Revised Edition of the Uganda Laws, vol. VII, 1951, p. 1237. An Ordinance making
compulsory the registration of births and deaths of non-Africans came into force in 1905.
89 Ashe. Two, Kings of Uganda, p. 52.
90 Africa. No. 7 (1895). C. 7708, p. 84. See also Uganda Notes, vol. 13, no. 1, 1913,
p. 20; and Roscoe. The Baganda, pp. 246-7.
91 Thomas and Spencer. Uganda Land and Surveys, p. 40.
92 Roscoe. The Baganda, p. 366. See also Mair. An African People in the Twentieth
Century, p. 196.
93 Roscoe. The Baganda, p. 246.
94 Africa. No. 7 (1895). C.7708, p. 84.
95 Roscoe. The Baganda, p. 246.
96 C.M.S. Archives. Rev. R. H. Walker to Captain F. D. Lugard, 3 February 1891.
97 R. H. Walker. Diary, vol. XV, letter 289, 7 November, 1895, pp. 68-9. (Archives
NOTES ON THE KIBUGA OF BUGANDA 43
98 J. R. L. Macdonald (1897). Soldiering and Surveying in British East Africa, p. 139.
99 Mengo Notes, vol. 2, no. 3, 1901, p. 62.
100 P. G. Powesland. History of Migration in Uganda, in Economic Development and
Tribal Change. A. I. Richards, Ed., (1954), p. 20.
101 Entebbe Archives. S.M.P. D.1369, Director of Public Works to Chief Secretary,
27 September 1910.
102 Uganda Notes, vol. 15, no. 13, 1914, p. 18.
103 Powesland. History of Migration in Uganda, p. 26.
104 Annual Report of the Director of Public Works, 1923, para. 41.
105 Powesland. History of Migration in Uganda, p. 33.
106 Ashe. Two Kings of Uganda, p. 52.
107 W. J. Ansorge (1899). Under the African Sun, p. 121.
108 Uganda Notes, vol. 33, no. 1, 1902, p. 4.
109 Powesland. History of Migration in Uganda, p. 20.
110 Proceedings of the Church Missionary Society, 1882-3, p. 55.
111 Mair. An African People in the Twentieth Century, p. 130. Burton. The Lake
Regions of Central Africa, vol. ii, reported, "Suna greatly encouraged, by gifts and
attention, the Arab merchants to trade in his capital; the distance (from the coast) has
hitherto prevented more than half-a-dozen caravans travelling to Kibuga . ." p. 193.
112 The Story of the Life of Mackay of Uganda-Told for Boys. By his sister (1891),
113 Uganda Notes, vol. 4, no. 6, 1913, p. 30.
114 Roscoe. The Baganda, p. 248.
115 No. 222 of 1913, Uganda Gazette, 1913, p. 228.
116 Catholic Union of Great Britain (London, 1893). Notes on Uganda, p. 75.
117 Africa. No. 7 (1895). C. 7708, p. 84.
118 G. Wilson. The Progress of Uganda. Journal of the African Society, 6 (1907), 134.
119 Uganda Notes, vol. 14, no. 12, 1913, p. 268.
120 Roscoe. The Baganda, p. 227.
121 Howell. The Fires of Namugongo, p. 76.
122 Kampala Municipal Council Archives. Dr. Ansorge to Colonel Colvile, 12 October
1894. Manuscript headed, The Swahili Town.
123 Colonel Ternan to Apolo Kagwa, 27 September 1897. From the private papers of
Sir Apolo Kagwa.
124 Johnston. The Uganda Protectorate, vol. i, p. 105 and vol. ii, p. 646.
125 No. 22 of 1 December, 1899, Uganda Laws, 1899, p. 75.
126 Buganda Government Archives. S.12/00, Apolo Kagwa to Johnston, 22 February
127 C.M.S. Archives. Bishop Tucker to Johnston, 17 August 1901.
128 Uganda J., 14 (1913), p. 131.
129 Grant. A Walk Across Africa, p. 230.
130 Wilson and Felkin. Uganda and the Eygptian Soudan, vol. i, p. 184.
131 Mengo Notes, vol. 2, no. 2, 1901, p. 56.
132 Ibid, vol. 2, no. 6, 1901, p. 81.
133 Uganda Notes, vol. 15, no. 13, 1914, p. 17.
134 Roscoe. The Soul of Central Africa, p. 45.
135 Uganda Notes, vol. 14, no. 3. 1914, pp. 55-6.
136 A. W. Southall and P. C. W. Gutkind. Townsmen in the Making-Kampala and
its Suburbs. East African Studies, no. 9, East African Institute of Social Research,
Kampala, Uganda, 1957 (Second Edition).
PLEISTOCENE ENVIRONMENTS AND
EARLY MAN IN UGANDA
By W. W. BISHOP AND M. POSNANSKY
THE term Pleistocene, meaning the most recent of the main geological
periods, was introduced by Lyell in 1839. Although the term is only 120 years
old, the period itself has generally been considered as spanning approximately
the last one million years. Recently absolute methods, including radiocarbon
dating, have enabled a more accurate time-scale to be constructed for the last
40,000 years. This in turn has been used to calibrate deep sea cores and it now
seems possible that the Pleistocene may have commenced little more than
300,000 years ago. (Emiliani, 1955.) However, another new method using
Potassium/Argon dating yields an age of 400,000 to 450,000 years B.P. (Before
Present) for deposits of the Antepenultimate Pleistocene Glaciation in North
America. (Evernden, 1959.) Nevertheless, in Africa it is probable that the
whole Pleistocene period, which is now usually defined in relation to the
degree of evolution of fossils of the elephant, horse and buffalo families, may
prove to be of the order of only 500,000 years in length, or possibly less. This
is the time-span which it is proposed to use for East Africa in this paper.
As absolute dating is still in its infancy it is necessary to depend largely upon
a relative time-scale which is based on changes in fossil fauna as animals
gradually developed in the course of time or in response to variations in their
environment. One advantage in sub-dividing the Pleistocene arises from the
fact that it includes the period during which man, defined as a tool-making
animal, was developing rapidly and stone tools frequently survive to be avail-
able as fossil tools to help in correlating Pleistocene deposits and in giving
to them an approximate age.
The object of this paper is to describe the areas in Uganda from which stone
tools or fossil remains have been recovered in stratified deposits and to recon-
struct, as far as possible, the conditions under which early man lived. Part I
deals with the geological and fossil evidence of former environments at each
locality and tries to show how conditions may have differed from those of the
present day. Part II indicates how man's tools developed with the passage of
time and how his mode of life changed to meet the challenge of his environ-
ment, until eventually he was able to modify the conditions themselves as he
became more skilled.
The observations in Uganda are incomplete and individual localities tend
to be separated by areas where evidence is lacking. Thus links between
different regions remain only tentative and there are many gaps in the time-
sequence, while we have deliberately restricted ourselves to areas where the
geological evidence is clear and the archaeological data not in question. Our
acquaintance with East Africa is of only three years duration and we both
PLEISTOCENE ENVIRONMENTS AND EARLY MAN 45
owe a great deal to previous workers and in particular to E. J. Wayland. With-
out his detailed observations and collecting since 1919 the picture that we are
able to present would have been much more fragmentary.
The problem in Pleistocene Geology is basically the same as that in all
geological work, to observe the fragmentary details of past periods in the
history of the earth and where possible to draw conclusions as to the condi-
tions in an area, in this case Uganda, at the time when a particular series of
deposits was forming. It is possible to present a series of about half-a-dozen
tentative and sketchy outlines of parts of Uganda as they were during the
period in which man was developing through the 'stone age' and to show from
these how the present landscape of Uganda gradually evolved in the last
The Ancestors of True Man?
Evidence from Uganda of the existence of man's earliest ancestors, or
'pre-men', is virtually non-existent. However, a fossil site dating from the Lower
Miocene period, possibly about twenty to twenty-five million years ago, was
recently discovered at Napak in Karamoja by J. G. Wilson of the Uganda
Department of Agriculture.2 Excavation by the Geological Survey has now
yielded numerous fossils including teeth of two genera of early ape, Proconsul
nyanzae and Limnopithecus sp. The former was an ape about the size of a
chimpanzee while the latter more closely resembled the modern gibbon. Both
are true apes but they may have been closely related to some common stock
which eventually gave rise to the modern great apes and to true man, 'the
These two unspecialized fossil apes lived and died on the slopes of the then
active volcano of Napak, which between eruptions became clothed temporarily
with vegetation of which fossil wood survives as evidence. Other animals
inhabiting the area at this time included the early cone-toothed elephant, four
genera of carnivores, three rodents, a lemur, a pig and a chalicothere. (Bishop,
The fossil assemblage is important as its similarity with that from the well-
known Lower Miocene localities in Kenya allows a tentative date to be estab-
lished for a fixed point in the development of the landscape of Uganda. The
surface upon which the new Miocene volcano of Napak was built up is pro-
bably equivalent in age to that of much of central Uganda. (The detailed
evidence for this correlation will be presented elsewhere.) Thus it is possible
to infer that many of the surviving features of the landscape of Uganda had
their origin in pre-Miocene times. The flat laterite hill tops such as Nsamizi
and Kololo, so characteristic of Buganda and Masaka districts, represent a
once continuous land surface that was already being dissected by broad valleys
as long as twenty-five million years ago. Since that period the rivers have merely
deepened and widened their valleys along the lines of the old pre-Miocene
drainage pattern. It is in these valleys, and in the Western Rift Valley, into
which all the major east-west river systems of Uganda eventually flow, that the
sediments, fossils and stone tools of the Pleistocene period are to be found.
46 UGANDA JOURNAL
For convenience the period is divided into three sub-divisions, Lower, Middle
and Upper Pleistocene but these terms in East Africa can at present be used
only in a purely relative sense and cannot be correlated with the similarly
named but more accurately known faunal sub-divisions of more temperate
latitudes in Europe and elsewhere.
A Period of Negative Evidence
(i) Kaiso. Lake Albert. The Kaiso beds, named from a fishing village on
Lake Albert, were first described in detail by Wayland (1925, 1926). His work
followed discoveries by prospectors between 1919 and 1921 of pieces of fossil
bone in the Kaiso area. The deposits are a series of horizontally bedded grey
to buff clays, silts and fine micaceous sands which must have been laid down in a
large lake ancestral to Lake Albert. The series includes ironstone bands one to
two feet in thickness which contain the majority of the fossils. These are made
up of abundant water snails and bivalves, fish, reptiles and mammals. The
fish include enormous Nile Perch, the vertebrae of which are very numerous
and often exceed in size those of their modern counterparts. Crocodiles were
abundant, while the remains of hippopotamus, including those of a large and
a pigmy species, are the most frequent among the mammals. Two genera of
elephant, three pigs, a three-toed horse and a zebra, a rhinoceros, a carnivore,
a baboon, various ruminants and a chalicothere have also been identified,
(Hopwood, 1926 and 1939) and some of them are shown in Fig. 2. This
assemblage dates the beds as Lower Pleistocene in age while the presence of
land mammals, the nature of the ironstones which are sometimes sandy but
frequently oolitic, plus the fine grained nature of the whole sequence, show
the area to have been occupied by a shallow lake or series of lagoons into
which the rivers flowed sluggishly carrying only fine detritus.
This could hardly have been the case so near to the escarpment if the rift
valley boundaries formed as pronounced a feature in Lower Pleistocene times
as they do today, and it seems likely that a more eroded escarpment formed the
boundary of the rift valley at this time. This is represented by the upper portion
only of the present scarp above the 'hanging base-line' of Wayland (1925,
Fig. 8). Broad valleys had been incised in this feature and slow flowing rivers
found their way from central Uganda bearing a load of fine grained clay and
silt which was deposited in the shallow lake to form the Kaiso beds. The drain-
age pattern of Uganda suggests that at this time the Kafu, Katonga and Kagera
flowed westwards and were the rivers in question.
Identical deposits with similar fossils occur near the west end of the Kazinga
Channel and southward for some ten miles along the eastern shore of Lake
Edward, and again west of Lake Edward near the Semliki River in the Belgian
Congo. This shows that this 'Kaiso environment', with conditions suitable for
the preservation of abundant fossils, was widespread. To date, however, no
definite stone tools or artifacts have been found in the Kaiso deposits and no
fossil remains of any primate that might have made tools.
Unless such evidence is forthcoming with further investigation it must be
concluded that at this time western Uganda, although an area with abundant
PLEISTOCENE ENVIRONMENTS AND EARLY MAN 47
fauna, did not support a population of tool-making hunters although man was
probably already in existence in Africa.
(ii) The Kafu River. It has been suggested that the Kafu River gravels
contain abundant evidence of a 'Kafuan' stone age culture based upon pebble
tools (Hirst, 1927, 1928; Wayland, 1927, 1934; Lowe, 1952.) Such 'tools',
however, have been under criticism recently in view of the difficulty of
recognizing simple pebbles of quartzite split by human means when they occur
within river gravels where abundant natural fractures also occur. (Oakley,
1957; Burkitt, 1956; Bishop, 1959; Clark, 1957.) It cannot be denied that the
later hand-axes and easily recognizable tools must have had a beginning in
more primitive forms, possibly very similar to the split pebbles illustrated by
Van Reit Lowe (1952, Plate 1). However, the stratigraphical evidence on the
River Kafu does not support the existence of such a Pre-Chellean pebble
culture and similar problems exist in the case of the so called pebble-tools from
the quartzite gravels of the Kagera River. Until further evidence is available
of the existence of true Pre-Chellean tools in an unquestionable stratigraphical
relationship it must be concluded that there is no firm evidence for the presence
of man in Uganda during the Lower Pleistocene.
The Period of Hand-Axe Man
With the commencement of the Middle Pleistocene a considerable change
occurred in the type of deposit that was formed in the Western Rift Valley.
The name Semliki Series (Lepersonne, 1949) has been given to these pre-
dominantly sandy deposits, with horizontal bedding but with some false
bedding within individual bands. They contain some clay portion and frequent
gravel lenses, while ironstone horizons are only rarely present. Towards the
top of the series dark palaeosols occur. Fossils are rare but in the upper Semliki
area of the Congo they include Palaeloxodon recki, a Middle Pleistocene
elephant. (de Heinzelin, 1955.) The deposits are coarser than the preceding
Kaiso series and the presence of much sub-angular quartz in the gravel suggests
rapid derivation by fast-flowing rivers, most of the clay portion having been
The most likely explanation of this sharp change in lithology is that the
deposition of the Kaiso series was followed by a major movement of the rift
valley boundary faults. This rejuvenated the lower reaches of all the rivers
flowing into the rift, causing waterfalls and rapid erosion of the Basement
rocks. Further evidence of this movement is seen in the fact that the Kaiso
beds exhibit low dips suggesting tilting by tectonic disturbance. The Semliki
series remain flat-lying and virtually undisturbed as this post-Kaiso faulting
was the last major movement of the rift valley faults. Since it occurred the
eastern escarpment of the rift valley has been virtually in its present form with
only minor fault movements and gradual incision by river erosion to modify
Thus at this period the western rift presented an attractive environment not
unlike the present one but with a more starkly precise and fresh escarpment
48 UGANDA JOURNAL
and numerous waterfalls plunging into a Semliki series lake which extended
to the foot of the escarpment in most places. There is evidence from the Congo
(de Heinzelin, 1955) that the shores of this lake were living sites of early man
and similar localities existed in Uganda.
(i) Mweya. In 1956, during excavations to widen a park road, A. C. Brooks
of the Game Department found an Acheulean hand-axe in gravels capping a
narrow col some five hundred yards north-east of Mweya Lodge in the Queen
Elizabeth National Park. A search in the area yielded a series of artifacts both
those eroding out naturally and those removed in the course of the road work.
In 1958 the writer was able to excavate a small section of the gravel and
obtain hand-axes, hammer stones, cores and waste flakes in situ, while surface
finds included hand-axes and a cleaver.3
The deposit in which the artifacts occur is a dirty coarse to medium, sub-
angular river terrace gravel. It is only a few feet in thickness and contains
much secondary limonite which often encrusts the pebbles and tools. The arti-
facts occur at the base of the deposit at a height of 90 feet above the Kazinga
Channel. The gravel was deposited by a river which flowed along the line of
the present Kazinga Channel but considerably above it.
The artifacts have all been considerably rolled as a result of their inclusion
in the river gravel, since they were derived from their original source at or near
the surface of the underlying Semliki beds. It seems probable that man lived
at some site or sites on the shore of the Semliki series lake not very far removed
from the present safari lodge at Mweya. Towards the end of the Middle
Pleistocene the lake became much smaller and similar in size to the present-
day Lake Edward, while soils were developed on the exposed flat top of the
former Semliki series lake bed. Across this surface rivers extended establish-
ing the present river pattern some 100 feet above modern lake level, and into
the valleys were swept the tools of Acheulean man to be rolled and mixed with
the river gravels.
(ii) Paraa. It is an odd coincidence that the other locality from which hand-
axes and artifacts are known at present in the Semliki series of the Uganda
section of the western rift, is in the vicinity of Paraa Lodge in the Murchison
Falls National Park.
In 1958, in company with Dr. J. de Heinzelin, the writer found a hand-axe
in vein quartz, together with numerous waste flakes and worked pebbles, in
a gully eroding a series of gravels and sands which represent the Semliki series in
the Paraa area. The beds rise to about 100 feet above Lake Albert near Paraa
and occur very extensively both north and south of the Victoria Nile and in
West Nile where they rise to about 300 feet above the lake. Other discoveries
seem likely in this area and Dr. Posnansky recently found a hand-axe in the
southern area of the park. The lithology in the Paraa area is almost identical
to that of the deposits in the Lake Edward area and in the Congo and con-
ditions must have been very similar, with Acheulean man hunting and living
along the lake margins.
(iii) The Kagera River. The third and richest area in Uganda from which
Acheulean artifacts have been recovered is away from the western rift, in the
Kagera valley and particularly in the vicinity of Nsongezi.
PLEISTOCENE ENVIRONMENTS AND EARLY MAN 49
I have suggested that during the Lower Pleistocene the Kagera river was
one of the major rivers which drained westwards from central Uganda into
the western rift. With the major post-Kaiso fault movement which rejuvenated
the rivers and caused a relative lowering of the rift valley floor, there seems
also to have been associated a rise of the shoulders of the rift which resulted
in an uplifting of the area east of the rift valley. This tilting may have gone
on intermittently but its ultimate result was to pond-up lakes, or increase
greatly the area of lakes that already existed in the headwaters of the major
rivers. These were the forerunners of lakes Kyoga and Victoria.
As the uplift continued well into the Upper Pleistocene, the lakes initially
extended much further to the west than they do at present. They may for some
time have continued to discharge to the west until the uptilting forced the lake
waters to retreat eastwards, stranding their former shore-lines and deposits.
This caused the lakes to overflow from outlets elsewhere as uplift to the west
exceeded the rate of downcutting of the river beds.
An arm of Lake Victoria occupied the Kagera Valley during the Middle
Pleistocene and extended as a long finger-like embayment some 70 miles west
of the present shoreline of Lake Victoria into the Nsongezi area. West of this
point the river passes through a narrow gorge cut in Karagwe-Ankolean
quartzites which with uplift made an excellent natural dam behind which a
lake was impounded.
The tilting, which initiated and eventually ended the lacustrine phase at
Nsongezi, also resulted in the reversal of the Kagera river causing it to
flow in its present direction eastwards to Lake Victoria across the former
lake deposits. These remain underlying extensive 'flats' on either side of the
river and about 100 feet above it.
In 1930, the cutting of a new road from Kagera Port to Kikagati resulted in
borrow pits near Nsongezi exposing a rich artifact horizon of which Wayland
(1934, etc.) and Solomon (1939) have described the stratigraphical relations,
while O'Brien (1939) and Lowe (1952) have described the typology. The lower
deposits of the series are not well exposed but extensive pitting by Wayland
has made available a number of good sections through typical lacustrine or
estuarine sands and clays. The artifact level, which is probably best referred
to as the 'M-N horizon', includes abundant Acheulean material with also a
proportion of Sangoan tool types. The artifacts are usually slightly rolled and
are mixed on the M-N horizon with sub-angular boulders and blocks of grey
quartzite which formed the raw material for the industry. The horizon marks
a temporary fall in the lake-level during which rocks moved as a scree down
the steep quartzite slopes and out over the lake deposits as a creep formation
which descends gradually towards the centre of the valley. Thus again we have
evidence of hand-axe man living close to a lake, at the head of an embayment
bounded by steep quartzite hills.
The reason for the fall in lake-level may have been either climatic or tectonic
but was certainly only of short duration and small amplitude since the deposits
under the M-N horizon exhibit little weathering and only slight evidence of
erosion of the unconsolidated lake beds at this period. When the lake-level
rose again it effectively sealed the artifact horizon and only subjected the tools
50 UGANDA JOURNAL
to slight rolling. Lenses of beach-like gravel and sand sometimes overlie the
stone tools but normal lacustrine or estuarine conditions were soon re-
Man becomes more specialized
(i) The Kagera Valley. In the period which followed the formation of the
M-N horizon, the tilting which continued throughout the Middle and into the
Upper Pleistocene gradually converted the Nsongezi area into a vast swamp.
The top of the lacustrine sequence consists of variable, slightly gritty, grey
swamp-clays. These are well seen in a series of erosion gullies near Nsongezi
which expose up to 20 feet of the deposits. The sections in the grey clay show
irregularly reddened horizons and others with ferruginous mottling or dark
humic layers resembling palaeosols. Abundant artifacts, derived from tempor-
ary land surfaces within the swamp clays, can be collected in the erosion
gullies. The assemblage is middle Sangoan (Lowe, 1952; O'Brien, 1939) and
includes beautifully made tranchets, lances and points in grey quartzite and
white vein quartz. These date from a much later period than the M-N
horizon material and show a great advance in man's technique of working
stone and also in his development of smaller more specialized tool types in
addition to the hand-axe, which with the cleaver was the multi-purpose tool of
the Middle Pleistocene. The assemblage must be assigned to the later part of
Upper Pleistocene and at this time the head of the lake embayment at Nsongezi
had become a shallow swamp which, at least during dry periods, allowed
man to hunt, make stone tools and possibly live on the flats. However with the
return of moist conditions he had to abandon the valley bottoms leaving his
artifacts to become buried and preserved under 6 to 12 feet of swamp clay.
It was only after this period and possibly within the last 25,000 years that
the uptilt to the west finally caused the swamp fringes at the head of the valley
to dry out completely. This allowed the present Kagera river to commence
cutting down below the flat top of the former lake beds until this now forms
platforms, approximately parallel to the thalweg of the river but 100 feet above
it and having the appearance of river terraces. At Nsongezi the negative break
of slope on the strand-line, at the junction of the former lake deposits and
their old cliff line, is now 400 feet above Lake Victoria. The feature appears
to grade eastwards, down valley, into a strand-line of Lake Victoria at about
100 feet A.L. (Above Lake), and this implies a relative westerly uptilt of about
300 feet in the 70 miles between the lake-shore and Nsongezi.
This amount of uptilt has taken place at least during the later half of the
Upper Pleistocene and is a visual expression of the continued tectonic move-
ments associated with the rift valley, which commenced at the beginning of
the Middle Pleistocene. The contrast between the rapid, rejuvenated western
sections of the former westerly flowing rivers Kafu, Katonga and Kagera and
their sluggish, reversed, swamp-filled sections which now flow east to Lakes
Victoria and Kyoga is very striking. The swamp divides in the through-valleys
are indicated in Figure 1.
PLEISTOCENE ENVIRONMENTS AND EARLY MAN 51
(ii) Lake Victoria. As might be expected the upwarping has given rise to
a series of strand-lines round the western shore of Lake Victoria. These rise
gradually away from the lake and are found at heights from 220 to less than
100 feet. Below this last height, additional strand-lines occur but they main-
tain their general height above the lake and would seem to be controlled by
Pleistocene Environments in Uganda. Locality Map.
stages in the down-cutting of the Jinja outlet rather than by tectonic
The evidence of strand-lines is preserved in the form of cliffs, caves and off-
shore deposits and these former shore-lines show signs of man's activities.
Many of the caves were once occupied by man but few of them have been
excavated to date. Numerous waste flakes of a microlithic industry based upon
quartz pebbles as raw material have been recovered from the surface of the
52 UGANDA JOURNAL
higher flats and from beneath barren material in caves associated with them.
Caves up to 50 feet above the lake have yielded pottery and microliths, while
at Hippo Bay, Entebbe, in a cave nine feet above the lake, iron and abundant
pottery occurs without microliths.4 The former lake-levels probably preserve
a detailed, and as yet virtually untouched, record of Man's development
through the Upper Pleistocene down to the historic period.
(iii) The Western Rift. By the close of the Middle Pleistocene, Lakes
George, Edward and Albert had already assumed virtually their modern form
and the present river pattern had also been established in most areas. The
Upper Pleistocene was a period of gradual river down-cutting and the deposi-
tion of scattered terrace deposits.
In the Lake George-Edward vicinity an outburst of volcanic activity
occurred with the formation of the Katwe and Bunyaruguru volcanic fields on
the flanks of the rift valley. These comparatively recent volcanic deposits cul-
minated in the formation of the characteristic explosion craters of this region
and the deposition of up to twenty feet of fine, volcanic wind-blown ash as a
surface mantle over the surrounding area. The period of explosive activity
probably took place approximately 10,000 years ago. (de Heinzelin, 1957.)
At about this time also a renewed movement of about 100 feet occurred along
the rift valley boundary fault in the Kichwamba area. A series of small fault
movements, giving rise to steps of between 30 and 50 feet in height, also took
place on both sides of the Kazinga Channel near Katunguru. It is also possible
that the climate in this part of the rift valley was a little wetter some 10,000
and more years ago than it is now, as red palaeosols occur within and beneath
the volcanic ashes while grey to brown soils are forming at the surface at the
Thus the Lake George-Edward area suffered a series of comparatively
recent changes during the Upper Pleistocene but as yet we know little of man's
activities in Uganda in this rapidly changing environment. However, in the
adjoining area of the Congo, near the point at which the Semliki river leaves
Lake Edward, a most interesting site has recently been excavated by Dr. J.
de Heinzelin (1957). Here there is evidence that the Ishango culture, which
included bone harpoons in addition to stone tools, may have been terminated
by the explosive phase of the Katwe and Bunyaruguru volcanic fields. A
picture of the effect that such outbursts from over 200 craters would have upon
the surrounding country and upon the environment of Ishango Man can be
obtained from Dr. Du Bois's account of the recent activity at the single volcano
of Kitsimbani in the Belgian Congo. (Du Bois, 1959.)
In Uganda man's changing environment through the Pleistocene has been
largely controlled by tectonic movements and volcanic activity associated with
the western Rift Valley. It is seldom possible to infer the type of climate that
contributed to man's environment at any particular time. Despite the detailed
Pluvial-Interpluvial sequence established for Kenya and Tanganyika, it is not
yet possible to recognize similar climatic sub-divisions in the Pleistocene of
Uganda although some climatic changes have undoubtedly taken place.
PLEISTOCENE ENVIRONMENTS AND EARLY MAN 53
The evidence upon which the above sequence of events is built up owes a
great deal to the observations of amateur workers such as J. G. Wilson, A. C.
Brooks, and R. M. Brachi. In acknowledging my indebtedness to them I would
express the hope that others may become interested through the medium of
The Uganda Journal, for many more living sites of early man must await
discovery in Uganda.
Since 1939, when Van Riet Lowe of South Africa completed his joint
researches with Wayland and when O'Brien's survey of Uganda's prehistory
was published, little work on prehistory has been undertaken in Uganda,
except for Wayland's excavations at Nsongezi in 1953 and 1954. The archaeo-
logical emphasis has shifted to protohistoric research (Posnansky, 1959). A
great deal of prehistoric research has, however, been undertaken in Tanganyika,
the Sudan, Kenya and, in recent years in particular, in the Belgian territories.
In Uganda the greatest number of early Stone Age finds come from gravel
beds, some stratified in sequences of lake deposits and others along river
valleys. Most unfortunately no true living floors occur, that is sites which
contain not only stone tools but also food debris in the form of bones and
other valuable indications of man's way of life. The number of find spots is
not at present large, and noticeable gaps, where research has been slight, exist
in the Eastern and Northern Provinces. At present in Uganda, though
numerous caves and rock shelters have been found, none of the few that have
been excavated, as has Nsongezi, contains stratified sequences of cultures so
valuable for relative dating. Nevertheless, from the stray surface finds, such
as the tools incorporated in the gravels like those from Mweya and Paraa, or
from land surfaces as at Nsongezi or from the rock shelters and caves like
Magosi, a pattern emerges to which can be added information from outside
Uganda as to chronology and prehistoric ways of life. It must be emphasized
from the outset that cultures can only be recognized from tools repeatedly
found in definite associations. Many stone tools are completely undiagnostic so
that it is only the occasional type tool of a culture that can be dated and not
the mass of waste flakes and general scrapers and parers that accompany any
Stone Age hunting and food-gathering culture.
From the Kaiso beds no stone tools have been found. In fact nowhere in
Africa have diagnostic tools been discovered associated with an Early
Pleistocene fauna.6 The first implements from a stratigraphical sequence in
Uganda come from the lacustrine deposits of the Kagera river valley and
the Semliki series. A great problem regarding Man's antiquity concerns his
first tools. We have to suppose that Man began as an occasional tool-user.
Australopithecus of South Africa (Robinson, 1958), a creature with a small
brain, living in Early and Middle Pleistocene times, probably occasionally
chipped stones to get a sharp edge to cut up his meat if none naturally sharp
was readily at hand, and he possibly also made use of food bones as pounders
and piercing tools. But occasional tools are not readily recognized from tools
flaked by the mechanical processes of nature, such as rolling in a river gravel.
It is only possible to say whether or not pebbles with occasional flakes detached
54 UGANDA JOURNAL
are tools if they are found on a definite undisturbed living floor. In this sense,
the Kafuan Culture of Western Uganda, as has been shown by Dr. Bishop,
contains no diagnostic tools-its representatives may have been tools, but they
may just as easily be the products of nature. No living floors anywhere in
Africa have as yet been found with Kafuan tools but if they do exist then the
Kaiso beds, so conducive to the preservation of bone, must prove to be ideal
for their occurrence. Streams flow and wash the tools from their banks, whereas
lakes like that of Kaiso often rise relatively slowly leaving undisturbed the
living floors on their shores before slowly inundating them with fresh deposits.
The Period of Hand-axe Man
Prior to 1956, hand-axes assignable to the Acheulean Culture had been
found in Uganda along the Lower Kagera valley, the Jinja area and in the
Lower Semliki area of Toro (O'Brien, 1939).7
At Paraa the raw material used for tool making is primarily vein quartz in
pebble form in contrast to the Kagera river area where the most readily avail-
able and preferred material is grey quartzite. This to some extent influenced
the general aspect of the total assemblage particularly with regard to size.
The quartzites which occur as massive scree boulders allow for much larger
tools to be made from the big flakes that can be detached. At Mweya both
quartzites and quartz were in common use.
The Nsongezi assemblage has been well described by both Van Reit Lowe
and O'Brien, whilst a detailed re-examination of the content of the M-N
horizon, as a result of recent excavations, is in preparation by the writer. The
assemblage is culturally mixed in that type tools of two different cultures have
been found; nevertheless, there is no firm typological evidence to suggest the
existence of two separate cultures. The recent excavations have, however,
confirmed that the lower slopes of the M-N horizon contain an assemblage
with more Sangoan elements than the upper slopes. Both typologically evolved
and primitive tools are found together in similar states of abrasion. Simply
stated, the Kagera river assemblages, best evidenced at Nsongezi, are those of
a well developed Acheulean facies with certain non hand-axe culture features.
These latter may perhaps be a result of the introduction of Levalloisian tech-
niques which gives rise to the emergence of what in its pure state would be
termed Early or Proto Sangoan.
The industry consists of hand-axes ranging in shape from fusiform to ovoid;
cleavers predominantly of the type possessing a parallelogram cross section
and a transverse cutting edge; Sangoan type picks; large asymmetrical cores
from which flakes, presumably for use as scrapers and paring tools have been
struck off in all directions; round-topped, high-backed cores and large side-
struck flakes which probably served as knives and heavy scrapers. The most
striking feature about the assemblage is the very large number of amorphous
pieces from which only a few flakes have been detached and which were never
fabricated as standardized tools. The amorphous character of the industry
reflects the easy availability of raw material from the screes of the steep-sloped
Ankole hills. Many of the asymmetric cores were probably also used as chop-
ping and pounding tools and many show signs of heavy usage.
PLEISTOCENE ENVIRONMENTS AND EARLY MAN 55
Though there are some finely flaked hand-axes, the flaking technique is
simple rather than elaborate and there is a lack of the elaboration of form or
fine secondary work that is found at Olorgesailie, Kenya, or Olduvai Gorge,
Tanganyika. There is a surprising absence of the stone balls interpreted by
Leakey (Leakey, 1948) as Bolas Stones.8 The large number of very crudely
flaked pieces, or of pieces with few flakes detached, must be considered not as
a primitive feature but as a reflection of the ready availability of a good quality
raw material for tool manufacture which led to tools being quickly fashioned
for the job in hand and just as quickly abandoned.
At Mweya (Fig. 4) the industry, which is far smaller in total number9 and
thus allows for far fewer generalizations, includes hand-axes, cleavers, stone
balls, asymmetric cores and core choppers and large numbers of flakes, includ-
ing several with facetted striking platforms produced by using Levalloisian
techniques. The Paraa assemblage (Fig. 2),10 which is purely a surface collec-
tion from four localities, is similar to that of Mweya but far less extensive and,
as yet, includes neither stone balls nor cleavers, though it does include several
flakes (Fig. 2, no. 6) produced by a prepared core technique. The Paraa assemb-
lage is marked by a relatively large proportion of 'pebble tools' (Fig. 2, nos. 2,
3, 5). These, if found in isolation, would typologically be assigned to the Kafuan
and Oldowan pebble tool cultures which began in the first part of the Middle
Pleistocene period. As with the primitive features of the Kagera river
assemblages, they reflect the availability of raw material, in this case pebbles
of vein quartz. The pebble tool content of the Paraa assemblage includes many
of the least abraded pieces. A certain number of these 'late' pebble tools are
found in the Mweya assemblage. A similar continuance of pebble tools has
been shown by Leakey at Olduvai (Leakey, 1951) where vein quartz was also
in common use. With any Stone Age industry the relative date of the total
assemblage must be judged by its most recent features. The fact that the
Mweya assemblage includes definite Levalloisian flakes seems to indicate that
the site post-dates that of Nsongezi. The hand-axes at both sites show an
absence of the elaboration of form and flaking that one associates with the full
development of the Acheulean, termed Acheul IV by Leakey, and one would
on this evidence call them late Acheulean.
The Paraa assemblage can be compared, however, with that from Ruindi to
the south of Lake Edward in the Congo. The Ruindi material is all from the
Semliki series. The Mweya material is similar to that found by de Heinzelin
(1957) in the Upper Terrace (Ts) at Ishango which overlies the Semliki Series.
Using this comparative evidence from outside Uganda, it would appear that
the Paraa assemblage is older than that of Nsongezi, and the Mweya more
recent or equivalent in age to the Nsongezi assemblages.
The presence of type Sangoan tools at Nsongezi, like the heavy picks which
developed from the hand-axes and various large scrapers and pounding tools,
could indicate, as has been suggested (Cole, 1954), a functional variation of
the Hand-axe Culture to a basically forested environment. It is of interest to
note that the Sangoan reaches its full cultural development in the most forested
areas of Africa like the Congo, Angola and Western Uganda, though it must
be borne in mind that the Sangoan has also been found in the Sudan
p . *' .
'0 i 2
Surface finds from erosion of gullies at Paraa: 1 and 4, hand-axes; 2, 3 and 5, pebble
tools; 6, a Levalloisian flake showing facetted striking platform; 7, a bifacially
flaked tool. All tools are of vein quartz except 4 which is quartzite.
'P~ ~ fi-
[Photo: Uganda Museum
Fossils of Kaiso Age (above) hippopotamus; part of upper jaw with teeth, well-worn
tooth, part of lower jaw with little worn teeth. (Below) two pieces of elephant molar,
two tusks of hippopotamus, part of a pig tooth.
[1nlloo: uganaa Museum
Artifacts from Mweya (left to right, top row): core, flake, core, bolas stone.
(Bottom row): 2 hand-axes and a cleaver. All are of quartzite except the two top right
which are of quartz.
[Photo: Uganda Museum
Implements from the erosion gullies Nsongezi. 1, broken lance; 2 and 3, points: 4, lance;
5 to 7, tranchets. I is of quartzite and chert: 2 and 7 of quartzite; 3 to 6 of vein quartz.
PLEISTOCENE ENVIRONMENTS AND EARLY MAN 57
(Arkell, 1949). It is impossible to say what exactly the picks and other type
Sangoan tools were used for, though it would be reasonable to assume that
many must have been used for wood-working. The presence of bolas stones
at Mweya and their absence at Nsongezi could be a further indicator of environ-
ment. Bolas stones are common at both Olduvai and Olorgesailie, both areas
where the vegetation must have been open plain or bush, and it is tempting
to suggest that the openness of the landscape dictated Acheulean Man's choice
of hunting tool.
The Kagera River assemblages with their apparent developmental forms of
the Sangoan show a greater affinity to the material from Isimila near Iringa
excavated by F. Clark Howells in 1957-58 than to the material from the Kenya
Rift sites such as Kariandusi and Olorgesailie and Olduvai Gorge. Typologi-
cally the assemblages" like those of Mweya are clearly later than these Eastern
Rift sites, though we find at later sites in the same area an absence of the
Sangoan. The writer would hesitate to assign this complete absence to the
environmental factors that have hitherto been adduced, but would suggest that
the cultural influences affecting western Uganda and the Congo following
Hand-axe times were primarily from the Nile valley area, whereas the Horn
of Africa primarily influenced the bush country of the Kenya Rift and northern
Tanganyika. It could thus be expected that new tool-making techniques such
as the Levalloisian would affect the area west of the Great Lakes sooner than
the Eastern Rift Valley area with its secondary influences from the Horn. A
possible origin of the Levalloisian technique has been cited in the Kharga
Oasis (Caton Thompson, 1952) sites of the Western Desert.
The evidence from living sites outside Uganda indicates that hand-axe users
did not know the use of fire and that wooden spears and perhaps late in the
period digging sticks for roots (Clark, 1959), implemented the stone tools and
were perhaps the more important. From the Semliki series in the Congo and
from Kenya (Posnansky, 1959) and Tanganyika a non hand-axe flake culture,
the Hope Fountain Culture, has been found. It is thought possible that, rather
than representing a different group of people, it may represent a specialized
occupation of the hand-axe users resulting in different tools. In Uganda the
Sangoan elements persisted and on many hills in the Sango Bay Region as
well as on the hills in the Gayaza region of Ankole large assemblages of
Sangoan material have been found. Unfortunately, it is impossible to date
these surface assemblages which are practically all fashioned from the grey
quartzites of the Karagwe-Ankole hills.
The Upper Pleistocene
Archaeological material has only been found in a stratigraphical context
for the period under review in the Kagera valley. The assemblages from the
erosion gullies show a marked advance upon those of the M-N horizon. We
have to envisage Man using mounted missiles like lance heads and perhaps
some form of arrow head or bolt. A noticeable feature of the assemblages
from the erosion gullies is the very small number of fine tools like lance-heads
and tranchets (Fig. 5) compared with the large numbers of flakes and cores.
Many of the lance-heads are broken and it must be assumed that the rarity
58 UGANDA JOURNAL
of these forms indicates their loss or breakage away from what may have been
the factory site areas in the vicinity of the gullies. The technical proficiency
of the tool-makers is highly advanced; basically the tools are made by a
Levalloisian technique followed by beautiful pressure flaking in the case of
the lances. It is probable that many pieces were fashioned to a perfection that
was unnecessary for the functional efficiency of the implement itself.
Tranchets (Fig. 5, nos. 5-7) are the most numerous of the fine tools and have
been recovered from all the larger erosion gullies. The larger ones (Fig. 5, no. 7)
were presumably used as wood-working tools and are possibly the final develop-
ment of the cleaver. The larger quartzite tranchets, as the one illustrated (Fig.
4, no. 7), all have an adze edge, rather than an edge that is central to both
faces. The smaller ones, consistently made of vein quartz and often showing
very fine secondary flaking, may well have been some kind of bolt rather than
being smaller variants of the larger tranchets of quartzite which seldom show
as fine secondary flaking.
Hand-axes were still being made and would indicate that this evolved
Sangoan12 is basically a development of the main Hand-axe Culture tradition
from which it presumably sprang. The hand-axes are mainly of a pointed form
and include the 'duck-head' variety which has a heavy butt and a thin blade
giving a duck-head profile in cross section. All are skilfully and economically
flaked. Scrapers, mainly on flakes, are also common as well as small
Levalloisian tortoise cores.
Several of the lances bear parallel pressure flaking of a form and beauty that
one normally associates with certain tools of the Lupemban II Culture of the
Congo. Caton Thompson (1946) had envisaged a possible Congo source for
the Aterian Culture of north Africa and the Stillbay Culture of eastern and
southern Africa. Her reasons for suggesting this had been partly based on
the long survival of the biface (hand-axe) traditions in this area. Recent exten-
sive surface finds made by John Wilson in Karamoja"1 are of interest in that
they largely consist of well-developed Stillbay tools but with the occasional
lance-head forms and a single tranchet, whilst among the recent Nsongezi finds
typical Stillbay (Fig. 5, no. 2-3) forms occur.
These finds indicate some culture contact between the Sangoan area of the
Congo and western Uganda and the Stillbay world of the Horn and east and
southern Africa. The importance of these recent Nsongezi and Karamoja finds
is that they allow for some cultural synchronization. Typologically the material
from the erosion gullies predates the Late Lupemban (Lupemban-Tshitolian)
material of the Congo to which a date of 13,143 B.C. has been assigned by
radio-carbon dating. Towards the surface of the swamp clays O'Brien found
tanged points of quartz, perhaps Uganda's first real arrow-heads, which are
close in form to the Lupemban points and to the Aterian points of north Africa.
In eastern and northern Uganda at this time, the Stillbay Culture was begin-
ning to receive influences from the Horn or from Kenya in the form of tech-
niques for making microliths or pygmy tools designed to be mounted, often in
composite form, for arrow barbs or even cutting and scraping tools (Clark,
1958). This new culture is known as the Magosian and several Magosian points
have been found as surface finds around the shores of Lake Victoria. The
PLEISTOCENE ENVIRONMENTS AND EARLY MAN 59
final Stone Age Culture and the one which continued until virtually the present
times, the Wilton, is largely a development of Magosian. Unfortunately, except
for the late cave and rock shelter sites around Lake Victoria mentioned else-
where in this Journal, neither of these cultures has been found in a firm
In conclusion, the recent archaeological finds from Mweya and Paraa allow
us to fit the story of a Man as Tool-Maker more closely into the picture of
landscape change which Dr. Bishop has outlined for part of western Uganda,
whilst the new evidence from Nsongezi and Karamoja makes it possible to
postulate the origins and cultural relations of Uganda's Later Stone Age
Cultures. The typological evidence indicates the importance of considering
the palaeolithic cultures in relation to those of the Congo and north and east
Africa. It would seem that at an early date cultural diffusion was taking place
around the west of the Great Lakes from the Saharan fringe of the Sudan.
De Heinzelin (1957) has shown the continuance of this zone of cultural diffusion
in his study of the assemblages at Ishango with particular reference to the dis-
tribution of barbed fish-points of both bone and iron.
1 By W. W. Bishop. Published by permission of the Director of the Uganda
2 For the location of places mentioned see 'Locality Map', Fig. 1.
3 I am indebted to the Warden, Mr. F. Poppleton, for allowing me to examine material
that he collected from the site and to the Director, Mr. R. M. Bere, for permission to
work in the park.
4 This site is fully described by R. M. Brachi at pp. 62-70 post.
5 By M. Posnansky.
6 The Olduvai Gorge Bed I fauna is now reputed to be of Early Pleistocene age,
similar to that of Kaiso. Oldowan tools and the hominid Zinjanthropus boiseae would
thus be Early Pleistocene on this correlation. Earlier tools of the Osteodontokeratic
(prelithic) tradition were made by Australopithecus prometheus at Makapansgat. (Infor-
mation based on the proceedings of the Fourth Pan-African Congress of Prehistory
at Leopoldville, 1959.)
7 The Lower Semliki implements were described by O'Brien (1939, p. 96) as Chellean,
though his illustrated specimens prompt the writer to assign those diagnostic pieces to
an Acheulean facies.
8 Of the 225 flakes from the writer's excavation, none showed a Levalloisian tech-
nique. Only one disc core was found and no true tortoise cores. Van Reit Lowe illustrates
several flakes and cores that could be Levalloisian.
9 The total assemblage numbers 328 pieces the bulk of which are in the Geological
10 From four different erosion gullies. Collections of material exist both in the
Geological Survey and the Uganda Museums.
11 The Isimila site consists of several distinct land surfaces.
12 A confusion exists (Cole, 1954, p. 168) over the nomenclature of the Sangoan
Culture. The erosion gully material would be assigned by certain authorities to the Upper
Sangoan. On recently examining material at Leopoldville, the writer would have no
hesitation to ascribing the non hand-axe elements to the Lupemban II Culture.
13 Paper in preparation by J. Wilson and the writer on recent material found in
Arkell, A. J. (1949). The Old Stone Age in the Anglo-Egyptian Sudan. Sudan
Antiquities Service Occasional Papers No. 1.
60 UGANDA JOURNAL
Bishop, W. W. (1958). Miocene Mammalia from the Napak Volcanics, Karamoja,
Uganda. Nature, 182, p. 1480.
(1959). Kafu Stratigraphy and Kafuan Artifacts. South Af. Jour.
Sci., May, 1959.
Brachi, R. M. (1960). Excavation of a Rock Shelter at Hippo Bay, Entebbe.
Uganda J., 24, 62-70.
Burkitt, M. C. (1956). Some Early Prehistoric Cultures South of the Sahara.
Scientia, Series VI.
Caton-Thompson, G. (1946). The Aterian Industry; its place and significance in
the Palaeolithic World. Jour. of the Royal Anthrop.
Inst., p. 44.
(1952). Kharga Oasis in Prehistory.
Clark, J. D. (Ed.) (1957). Proceedings of Third Pan-African Congress of Pre-
history, Livingstone, 1955.
(1958). Some Stone Age Wood-working Tools in Southern
Africa. S.A. Arch. Bull., No. 52.
(1959). The Prehistory of Southern Africa.
Cole, S. M. (1954). The Prehistory of East Africa.
de Heinzelin, J. (1955). Le Fosse tectonique sous le Parallele d'Ishango. Inst. des
Parcs Nat. du Congo Beige. Brussels.
(1957). Les Fouilles d'Ishango. Inst. des Parcs Nat. du Congo
du Bois, C. G. B. (1959). Recent Volcanic Activity in the Kivu District, Belgian
Congo. Uganda J., 23, 118.
Emiliani, C., (1955). Pleistocene Temperatures. Jour. Geol., 63, 538-78.
Evernden, J. F. (1959). First results of research on dating of Tertiary and Pleistocene
rocks by the potassium/argon method. Proc. Geol. Soc. Lon., 1515, 22
Hirst, T. (1927). Summary of work carried out in the Kafu Valley in 1926. Ann.
Rept. Geol. Surv. Ug. for 1926, p. 15.
(1928). Summary of work carried out in 1927. Ann. Rept. Geol. Surv.
Ug. for 1927, p. 17.
Hopwood, A. T. (1926). Fossil Mammalia in Wayland (1926), pp. 13-36.
(1939). The Mammalian Fossils in O'Brien (1939), pp. 308-16.
Leakey, L. S. B. (1948). The Bolas in Africa. Man., April, no. 53, p. 48.
(1951). Olduvai Gorge. Cambridge.
Lepersonne, J. (1949). Le Fosse tectonique-Lac Albert-Semliki-Lac Edouard.
Ann. Soc. Geol. de Belge, 72 M, 1-92.
Lowe, C. Van Reit (1952). The Pleistocene Geology and Prehistory of Uganda.
Part II Prehistory. Mem. No. VI. Geol. Surv. Ug.
Oakley, K. P. (1957). Tools Makyth Man. Antiquity, 31, 191-210.
O'Brien, T. P. (1939). The Prehistory of Uganda Protectorate.
Posnansky, M. (1959). Progress and Prospects in Historical Archaeology in Uganda.
Discovering Africa's Past. Uganda Museum Occasional
Paper No. 4, pp. 83-9.
(1959). A Hope Fountain Site at Olorgesailie, Kenya Colony. South
Afr. Arch. Bull., 19, pp. 31-9.
Robinson, J. T. (1958). The Sterkfontein Tool Maker. The Leech, XXVIII,
Solomon, J. D. (1939). The Pleistocene succession in Uganda. O'Brien (1939),
PLEISTOCENE ENVIRONMENTS AND EARLY MAN 61
Wayland, E. J. (1925). Petroleum in Uganda. Mem. No. I. Geol. Surv. Ug.
(1927). A Possible age correlation of the Kafu gravels. Ann. Rept.
Geol. Surv. Uganda for 1926.
(1934). Rifts, Rivers ,Rains and Early Man in Uganda. Jour. Roy.
Anth. Inst., 64, p. 333.
et al. (1926). The Geology and Palaeontology of the Kaiso Bone Beds.
Occ. Paper No. 2. Geol. Surv. Ug.
EXCAVATION OF A ROCK SHELTER
AT HIPPO BAY, ENTEBBE
By R. M. BRACHI'
T HE object of this excavation was to establish a cultural sequence for
pottery from a site apparently of historic or protohistoric period. Hippo
Bay was chosen because it lies inside the present Livestock Experimental
Station, and there has therefore been no permanent African occupation of the
place for a number of years. Three sites were quickly found. Two are on steep,
open slopes and have suffered much from erosion. The third is a rock shelter,
and this site was chosen because of its protected nature and because of the
abundance of surface pottery found there.
The shelter is at the foot of a cliff forming the western headland of Hippo
Bay. (Fig. 2.) This bay, which faces south, is situated near the south-west end
of the Entebbe Peninsula (0 3' N. 32o 27' E.). The cliff is cut in typical nodular
lateritic ironstone, and the cave is near the end of the promontory at the rear
of a pronounced 'flat', which is a raised beach with a maximum height, at this
point, of nine feet above Lake Victoria. (Fig. 1.)
Pitting showed that the flat was underlain by two to three feet of variable
loamy soil, loamy sand, and a basal sandy gravel containing numerous laterite
nodules and occasional quartz pebbles. These deposits rest upon an uneven
surface of laterite, which is a former platform, cut by waves, similar to a
surface at present covered by shallow water off the end of the promontory.
As shown in Fig. 1, a lens of clean, medium to coarse, quartz sand, with
a maximum observed thickness of nine inches, and identical in lithology with
that forming the present beach of Hippo Bay, occurs in the cave mouth resting
upon laterite. This basal sand is undisturbed, slightly indurated by iron, and
sealed from the overlying occupation debris by a one to two-inch gravel layer
of rolled laterite and quartz pebbles. Despite this fact small 'sticks' and slightly
rolled nodules of charred wood occur distributed throughout the sand. The
source of the charcoal is not known but it seems probable that areas nearby
were attractive living sites in the period before the cave itself could be occu-
pied and that charred wood from cooking fires or accidental bush fires found
its way into the lake.
The evidence suggests that the sand was deposited and the charcoal frag-
ments incorporated at a time when Lake Victoria stood at least nine feet
higher than at present and was able to wash into the cave. The raised beach
is correlated with a former lake level found at other localities on the Entebbe
peninsular and along the western shore of Lake Victoria at an average height
of twelve feet above the lake.
This 'Hippo Bay level' is the lowest of a series of former 'strand lines' occur-
ring up to 200 feet above the western shore of the lake. Although the upper
levels show signs of tilting due to earth movement, the Hippo Bay level main-
tains a constant height above the lake. It is probably the result either of a
24 j Z
6 SANDY LOAM .- .
^_-AMY rL 8
2 LATERITE 2
LAKE LEVEL 0
25 20 40 60 80 100 120 140 160 L180o
FEET FROM LAKE MARGIN
Hippo Bay Cave Raised Beach. Cross Section. 0
64 UGANDA JOURNAL
change in precipitation within the drainage basin decreasing the rate of down-
cutting at the outflow, or of a particularly resistant band in the rock barrier
of the Ripon Falls outlet.
A sample of charcoal from the basal sand has been sent for radiocarbon
dating to the Yale University Geochronometric Laboratory. If a C14 age can
be obtained, it will establish a date for the deposition of the beach sand and
hence for the formation of the lowest former lake level. It will also provide
a maximum age for the archaeological material described below, for with the
fall in water level the cave became available as a possible lake-side dwelling
The filling of the cave was removed in six-inch spits from baulks thirty inches
square. Depths were measured from the land surface-a process justified in
this case since the major stratification lines run roughly parallel to the ground
surface. It was early seen that the quantity of pottery at the site was such
that collection of the pottery by the six-inch spits was essential. These collec-
tions were then washed, and all sherds which were not rims, bases, decorated,
or otherwise of interest were thrown away. Little restoration of the fragments
has therefore been attempted. Structural features were rare, and all were
destroyed after plotting.
The excavation was carried out over a period approaching two years, and
the effects of erosion were very considerable. Fallen pottery of interest was
collected at each visit, and marked accordingly.
The finds consisted of pottery, some polished stones, some vegetable matter,
either dried or charcoal, and a few iron objects.
Pottery was present in very large quantity and has been classified according
to the form of rim or decoration. The first two classes described are by far the
largest of those found.
ENTEBBE WARE. (Fig. 3, no. 1.) Rims chiefly of large crocks having a dis-
tinct bulge on the upper inner surface of the rim. This was formed, as can be
seen in a number of examples, by folding the upper surface of the clay inwards
before baking and smoothing it in an upward direction. Many of the examples,
which number more than 200, have on the inside continuous grooves of V
section, evidently applied with a comb. These grooves probably extended a
considerable distance down the inside of the vessel, but no base carrying
them was found. Their function is obscure. The outer surfaces have in the
majority of cases a herring-bone decoration, usually extending some two inches
below the rim. Below this is a number of horizontal grooves. Sometimes the
decoration is varied by waving the horizontal lines, and occasionally it is
altogether absent. This type of pottery was found in the vicinity of Entebbe
by Marshall (1954) at Waterworks Point and Bugonga Point. I have also found
Entebbe ware as far west as Ziba (surface finds at canoe landing) and as far
north as Lutembe It was not, however, found among many sherds at
Kazi. There is considerable variation in detail, both of section and of
WATER JARS (Fig. 4, no. 1.) This class is temporarily so named because of
a general resemblance to the familiar vessels used to bring water from the
waterhole to the house. They are roughly spherical, the rim being more or
-. ~1 .
General view of Hippo Bay Caves. The excavated shelter is on the right.
EXCAVATION OF A ROCK SHELTER AT HIPPO BAY 65
less flanged. The decoration usually consists of a twisted cord pattern extending
downward half an inch to two inches from the rim.
NICKED RIMS. The decoration consists of a series of notches placed around
the upper surface of the rim; the vessels are generally small. There is occasion-
ally also a twisted cord decoration.
FESTOON DECORATION. (Fig. 4, no. 2.) The decoration is probably applied
by a roller and consists of three or four lines of impressed holes close to the
rim; below this there is a similar number of rows of holes arranged in loops
or bights. The decoration is usually applied to fairly large vessels, the rims
resembling those of Entebbe ware.
2 3 4
Pottery from Hippo Bay Cave. 1, Entebbe ware; 2, Embossed pottery; 3 Lattice
decoration; 4, Black burnished ware.
EMBOSSED POTTERY. (Fig. 3. no. 2.) The decoration consists of rows of lumps,
usually small, applied close to the rim of the vessel, which again resembles
those of the Entebbe ware.
BOTTLE NECKS. A number of fragments of tall narrow necks were found.
LATTICE DECORATION. (Fig. 3, no. 3.) This decoration, applied by roller,
consists of impressed squares or rectangles. It is present on a considerable
number of sherds; a few are parts of small vessels, and their interiors have
definite red slips.
BLACK BURNISHED WARE. (Fig. 3, no. 4.) Only two fragments of this type
were found, and it is believed that they might be assignable to a period after
BASES. The number of definitely identifiable bases found is curiously small:
there can be little doubt that some rounded ones have been discarded unrecog-
nized. One specimen bears some resemblance to the 'dimple' bases found in the
66 UGANDA JOURNAL
Kavirondo area, but scarcely sufficient to justify the suggestion of a connec-
tion between the sites.
OTHER TYPES. (Fig. 4, nos. 3, 4, 5.) A number of other decorations were
found, though usually in very small numbers. A large collection of section
Pottery and small finds from Hippo Bay Cave. 1, Water jar type pottery; 2,
Festoon decoration; 3-5, Other pottery types; 6-7, Arrow heads; 8, Lip plug.
EXCAVATION OF A ROCK SHELTER AT HIPPO BAY 67
drawings and descriptions of all the pottery have been deposited in the Uganda
Table I gives the distribution of the various types of pottery at intervals of
six inches of depth. The only conclusions that can be drawn are that the period
of occupation of the site was short and that there was no variation of pottery
types during the occupation. The Entebbe ware, which shows considerable
variation in detail, might repay more detailed investigation.
Finds other than pottery included carbonized coffee beans, described by
R. E. T. Hobbs, then Scientific Adviser to Agricultural Enterprises, Ltd., as
a "typical sample of indigenous coffee", and mwafu nuts (Canarium schwein-
furthii, Engl.). These were found in the carbonized and dried states and are
believed to have been used as an incense. (Eggeling and Dale, 1940.) Frag-
ments of a number of unidentifiable dried fibres were also found.
Vertical Distribution of Pottery
0-6 84 25 7 4 3 1 4 1 1 1
6-12 70 26 4 5 4 3 2 1 1 1
12-18 45 24 3 1 6 3 1 1 1
18-24 71 29 2 1 1 1 1 1
24-30 42 24 4 4 3 1 1 3 1
30-36 18 13 3 1 1 1 1 -
36-42 5 10 1 1 2 1 -
42-55 1 -
Surface 2 1 -
Falls 71 26 13 3 1 4 3 2 1 1 1
Totals 409 177 33 21 16 12 12 8 8 9 6
Polished stones consisted mainly of pestles similar to those in use today to
grind foodstuffs. Two fragments bore grooves used for sharpening. The rock
types all occur locally. A small plug of a red stone, perhaps used as a lip
plug, was also found. (Fig. 4, no. 8.) Drawings and descriptions have been
deposited in the Uganda Museum. One mweso board was found on the laterite
surface at the top of the cliff above the cave.
Metallic objects. A portion of an iron knife and two arrow-heads were found
(Fig. 4, nos. 6, 7).
68 UGANDA JOURNAL
So far as the shallow deposits of this site are concerned, no development
of pottery types can be discerned. The general scarcity of pottery at the lower
levels may point to a short occupation period.
It is considered that after the fall in lake level, the cave, shallow though it
is, was occupied by bats, there being nine to twelve inches of guano lying on
the laterite floor inside the cave. Secondary changes have locally coloured the
guano brown with production of very insoluble ferric phosphate. Within the
cave, human occupation has resulted in the formation of ashy and peaty com-
pacted layers, some about six inches thick. The amount of pottery inside the
drip line is very much less than that outside.
The most recent feature of the cave must be the trenches. These cut through
all the strata inside the cave, and are about two feet wide. The filling is usually
dusty, with little pottery or other objects, and the walls seem to have been
little damaged by falls of soil, as were our own trenches. This fact implies that
they were refilled very soon after they were dug. It is difficult to assign any
purpose for them except perhaps removal of guano. There is no surface
indication of their presence.
The total absence of food residues such as bones and teeth is curious. It
has been suggested that the nature of the soil is such that all traces of bone
could be destroyed, but this seems unlikely in view of the recent date which
is probably to be assigned to the site. It is, however, possible that such residues
are grouped in another portion of the site, for only one-quarter to one-third of
the cave filling has been excavated.
The Hippo Bay rock shelter is still only the sixth site of its kind to have been
excavated and published in Uganda.2 Nevertheless from those few reports
previously published and from stray finds in various caves that have received
cursory examination it is possible to see some pattern emerging for the later
archaeology of Uganda.
The Entebbe type pottery previously found at Bugonga Point Cave and at
the Pumping Station Point (Marshall, 1954, pp. 52-3) and at various open sites
round the Lake would seem to be a type diagnostically associated with rock
shelter and lake-side sites. It will be noticed that this ware and the simple
'water jar' or cooking pot account for more than four-fifths of the total
assemblage. The assemblage lacks the sophistication of either the form or
variety of the pottery assemblages of the Bigo Culture. There are no huge
storage jars such as one might expect from a beer-drinking agricultural com-
munity, no complicated pottery forms like the pot stands of Masaka Hill or
Mubende, and of the few sherds bearing a slip, all are from small vessels,
probably drinking cups and bowls; pieces of the embossed and nicked-rim
pottery are also from small vessels, but basically the assemblage is of cooking
pots and as such is typical of a simple group of rock shelter folk using only
occasional finer ware obtained from their more settled neighbours.
The other finds afford no definite clue as to the age of the occupation of
the cave. As with those caves previously excavated, the lack of a decided
stratification would indicate a single phase occupation rather than occasional
occupation over a considerable period of time. Taking into consideration the
evidence from the Kenya portion of the Lake Victoria coast, which is unfortu-
EXCAVATION OF A ROCK SHELTER AT HIPPO BAY 69
nately still unpublished, it is possible to see a population, small in number,
living in the caves and rock shelters, and, as far as we can tell, subsisting
largely by hunting, fishing and strand looping. In the Kisumu area numerous
shell middens are known; at Nsongezi rock shelter, though hunting was pre-
ferred, two definite shell horizons were found. These hunter food-gatherers
had by the twentieth century been absorbed amongst the general population
and the caves were used only periodically by fishermen. Whether these hunter
food-gatherers could have been surviving Bushmen is impossible to say with-
out finding skeletal material. All that it is possible to suggest is that tales
of surviving Stone Age hunter food-gatherers occur and that the present and
yesterday hunters of East Africa like the Sandawe and Hadzapa of Tanganyika
and perhaps the Wanderobo of Kenya bear some affinities in physical type or
language to the Bushmen, who are thought of as having been the indigenous
inhabitants of much of southern and central Africa.3
It is possible to trace the assimilation of these hunter food-gatherer groups
into the expanding and more secure agricultural societies by their material
culture. Starting off as pure Stone Age societies with a Wilton material culture
of microliths and scrapers they acquire from their agricultural neighbours first
pottery, or less probably the practice of potting, and later iron tools. The
material from the rock shelter of Nsongezi and Aerodrome Cave comprises
Wilton microliths, pottery and iron tools. It is not known at what date iron
working or tools first came into Uganda, but from more detailed work in the
Rhodesias it must be assumed that iron tools were certainly in use by A.D.
1000-1200. Nevertheless Wilton tools are known to have been made from
gun flints in the Rhodesias and South Africa as late as perhaps the eighteenth
or early ninteenth centuries.4 The increasing use of iron meant that less reliance
was placed on microliths for arrow barbs and other hunting tools, and such
sites as Bugonga Point and Zirizi Cave5 include only crude microliths; at
Hippo Bay there were none at all. This fact suggests that the rock shelter at
Hippo Bay was occupied at a time when iron was in general usage.
Though none of the pottery can at present be dated with certainty, the red
slipped ware and polished black ware is of considerable interest. The latter
ware is nowhere found amongst assemblages of the Bigo Culture, and at Kibiro
is found only late in the sequence.6 It is not real Bunyoro graphite ware, though
the polished black wares of Buganda could well be imitations of this famous
and distinctive pottery. Tradition suggests that this pottery has no great
antiquity and the Kibiro and the general archaeological evidence again is sug-
gestive of a late date. The one dimple base bears no direct similarities in
fabric to the dimple based ware of Kavirondo which is itself probably of
relatively recent date. At Nsongezi some of the pottery has been compared by
Leakey (1948, p. 42) to the Kenya dimple-based ware and again this pottery
is a polished black ware. Of the decorative motifs employed, the string roller
patterns are still in use but were also used on the pots of the Bigo Culture.
The mammilated or embossed wares are of interest but are at present
The complete lack of microliths, the presence of iron tools, the presence of
black polished and red slipped wares and the use of motifs common also to the
70 UGANDA JOURNAL
cultural descendants of the Bigo Culture would all tend to suggest a late
date for occupation of the cave. How late it is hard to say, but probably some
time within the last two centuries. The total assemblage is of great value in
showing types of pottery in contemporaneous use at one spot for a short
period, and further work at other sites will almost certainly provide further
clues as to a more exact date.
1 With acknowledgments to Dr. W. W. Bishop, who contributed the paragraphs on
the geology of the site and drew Fig. 1; to Dr. M. Posnansky, who wrote the conclusions
and drew the small finds; to Mrs. Shinnie and Mr. French for the drawings of pottery;
to the many persons who helped with the excavations; and to the Officer in Charge of
the Livestock Experimental Station for permission to excavate.
2 The caves and rock shelters previously published comprise Nsongezi and Aerodrome
Cave, No. 1, Entebbe (Lowe, 1952, pp. 99-101): Kagadi (O'Brien, 1939, p. 263); Chui
Cave (O'Brien, 1939, p. 271); Bugonga Point Cave (Marshall, 1954. pp. 52-3).
3 de Heinzelin de Braucourt, J. (1957). Les Fouilles d'lshango. Brussels, p. 98.
4 Alimen, H. (1957). The Prehistory of Africa. London, p. 312.
5 Excavated in part by Dr. Bishop. 1956.
6 Personal communication from Professor J. Hiernaux who excavated at Kibiro in
Leakey, M. D., Owen, W. E., and Leakey, L. S. B. (1948). Dimple-Based Pottery
from Central Kavirondo, Kenya Colony. Nairobi.
Lowe, C. Van Riet (1952). The Pleistocene Geology and Prehistory of Uganda.
Part II. Prehistory. Geological Survey of Uganda. Memoir No. VI, Entebbe.
Marshall, K. (1954). The Prehistory of Entebbe Peninsula, Uganda J., 18, 44-57.
O'Brien, T. P. (1939). The Prehistory of Uganda Protectorate.
Posnansky, M. (1959). Progress and Prospects in Historical Archaeology in Uganda.
Proceedings of the Conference of Curators, Uganda Museum Publication.
EMIN PASHA AND BUNYORO-KITARA.
1877 TO 1889
By A. R. DUNBAR
THIS article' is concerned with the period of Egyptian influence on
Bunyoro-Kitara2 in which the central figure was Eduard Schnitzer, better
known to the world as Emin Pasha. The son of a Protestant merchant, he was
born at Oppeln in Prussia on 28 March 1840. After working as a doctor in
Turkey, Armenia, Syria and Albania he went in 1875 to Khartoum, entered the
Egyptian service under Gordon, the Governor of Equatoria, as a medical
officer, being known as Emin. His flair for foreign languages made
Muhammadans readily accept both him and his work.3 While still a medical
officer he was sent on three political missions, the first in 1876 to extricate
Nuehr Aga and his troops from Rubaga,4 the second in 1877 to Kabalega,
Mukama of Bunyoro-Kitara, to bring about a peaceful solution of existing
difficulties, and the third from November 1877 to April 1878 to assure Mutesa,
Kabaka of Buganda, that Egypt had abandoned her plans for a new route to
Equatoria by way of the Indian Ocean and Buganda.
Prout, an American, who succeeded Gordon as Governor of Equatoria in
November 1876, was followed in May 1877 by another American, Mason Bey.
Meanwhile Gordon had, in February 1877, become Governor-General of the
whole Sudan from Wadi Halfa to Lake Albert. Mason Bey having been with-
drawn in August 1877 was replaced by Egyptian Governors, whose incompe-
tence led Gordon to appoint Emin as Governor of Equatoria in July 1878.
Emin remained as Governor until May 1889 when he was reluctantly
'relieved' by Stanley, and the province thereupon relapsed into anarchy.
It was during Mason's brief regime that Emin received orders to undertake
his second political mission, and to come to terms with Kabalega whose
hostility was hindering the Egyptian administration in the south of Equatoria.
He travelled through northern Bunyoro by the paths linking the posts held by
small Sudanese garrisons under Egyptian officers. Early in August 1877 he
reached Mruli5 on the east bank of the Kafu where it joins the Nile.
Lengthy negotiations were necessary in order to obtain Kabalega's per-
mission to enter Bunyoro. Mindful of the description given by Baker of
Kabalega's begging propensities, Emin went in September to Kisuga on a hill
near Katagarukwa, where he left everything that could be dispensed with,
including even his rifle. He then marched to Londu situated on Kibiama Hill,
four miles south-west of Masindi and the furthest Egyptian outpost in
Kabalega's country. From there he marched to Mparo,6 near Hoima.7 On his
way he gave a description of what he saw: "the people are clothed in soft ox-
hides, from which the hair has been removed, except at the borders, where a
strip of hair of two fingers'-breadth has been left as an ornament; their costume
is completed by arm-rings and anklets made of brass and necklets composed
72 UGANDA JOURNAL
EMIN PASHA AND BUNYORO-KITARA 73
of roots. The head is not shaved-shaving is a sign of mourning . The
people impressed me favourably; they were modest and unpretentious, and
satisfied with anything that was given them . We then once more arrived
at clearings, where bananas, sweet potatoes, and cow-peas intermingled, and
here and there the green stalks of maize were seen, or the broad leaves of
Virginian tobacco . Probably in order to impress the stranger with the
immense size of the land, and therefore with the greatness of its ruler, he is
led about for days through the high grass, when the direct route would hardly
occupy three marching hours . ."
"In marked contrast to all the negro tribes inhabiting our territory . .
the Baganda and the Banyoro have brought commerce to a more advanced
state of development." The market at Mparo gave convincing proof-"the
collection . of the most diverse products, and a concourse of types of
nearly all the Eastern equatorial races".8 It is interesting to note that there was
already at this time a unit of currency in Buganda and Bunyoro created by the
Arabs, the cowrie shell, five hundred shells to a Maria Theresa thaler.9 Trade
in Bunyoro depended upon Buganda because Mutesa would not allow the
Arabs to go from Buganda to Bunyoro: it was much to his advantage to claim
possession of all arms and ammunition brought by the Arabs as trade goods.
As a result, neither Kamurasi nor Kabalega had been in a position to retaliate
for the raids made upon them by the Baganda. Mutesa was also jealous of
Kabalega whom he liked to represent to strangers and others unacquainted
with the real state of affairs, as his vassal. Despite the ban, two enterprising
Arabs had reached Bunyoro in 1872 and, in return for presenting half their
trade goods to the king, they obtained accommodation, food, and five times
the value of the goods in ivory. There was an export trade in coffee, bark-cloth,
hides, salt, pottery, iron, mats, fruits, resins, wood and female slaves.10
At Mparo Emin came face to face with Kabalega. "This, then, was Kabalega,
the cowardly, treacherous, beggarly drunkard described by Baker. The grace-
ful folds of a piece of fine salmon-coloured bark-cloth covered his body up to
the breast, above which it was perfectly bare, except the left shoulder, over
which was thrown like a plaid, a piece of darker-coloured bark-cloth. Two
burnt scars were visible on the temples of his well-formed, smoothly shorn
head . .his four lower incisor teeth were wanting . and the upper incisors
projected slightly, and were brilliantly white . A necklace of hairs from a
giraffe's tail, upon the middle of which was strung a single blue glass bead,
encircled his neck. A root amulet and an iron bracelet were the only ornaments
on his strong muscular arms; his hands were small and well kept. He is
strikingly fair, probably in consequence of his pure Kihuma blood. He made,
upon the whole a very favourable impression upon me, but there was a decided
voluptuous expression on his face . I next gave him the presents I had
brought with me and much enjoyed his pleasure in receiving them . My
soldier had a small revolver in his girdle; Kabalega requested permission to
view it, and comprehended at once its mechanism. He took it to pieces, put
it together again, and then gave it back to me. He then asked me to inform him
how I had enjoyed myself last year in Buganda, and what I had seen there, and
he was highly amused with my description of the court ceremonials which
74 UGANDA JOURNAL
obtain in that country. Threatening rain brought our conference to an end
before either of us wished its conclusion . .
"I have often visited Kabelega subsequently and cannot say that I ever heard
him speak an improper word or make an indecent gesture, or that he was ever
rude . Kabalega is cheerful, laughs readily and much, talks a great deal,
and does not appear to care to be bound by ceremony . I certainly cannot
charge Kabalega with begging; on the contrary he sent me daily, in the most
hospitable manner, stores of corn, meal, mwenge," etc., which although they
were intended to supply the wants of one day, could easily have been made to
last us a fortnight. During my repeated visits Kabalega gave me the impression
of being a thoroughly hospitable and intelligent man . he proved this in
a very noteworthy manner in connexion with an incident which might have
brought me into a very awkward position. Notwithstanding my strict orders
that no hostile action should be taken by the Egyptians during my visit to
Bunyoro, the soldiers in our nearest station, led by stupid, jealous officers,
made a raid upon the country, and killed several of Kabalega's people.
Katagrua12 was sent by the king to give me this information, and to assure me
at the same time, that, although this occurrence was highly displeasing to him,
it should in no way affect our personal relations! . I received a detailed
account of all the events that happened during Baker's visit, a curiously
different account from that given in Ismailia."13, 14
Emin wrote "I had to listen to a long recital of the doings of the Danaglas15
of Rionga, Anfina, and his deceased brother Tovoka,'6 the sum and substance
of all being that he had been continually provoked and attacked by them,
although he, as occupant of the throne, was entitled to rule over them. I there-
upon told him that we quite understood this, but that the others had made
friends with the Government and rendered many services, while he had always
remained hostile. He replied very plausibly that they had made friends with
the Government solely because they were compelled to do so for their own
safety. As for his own acts of hostility, it was true that he had killed some
Danagla and fought with Baker Pasha, but only in self-defence. He now begged
of me to tell him what were the Government's desires, as he would like to live
at peace with us. 'You do not know me; you say to each other far away from
here: "Kabalega is a robber, a murderer!" Has anyone of you ever come to
me? Has anyone ever satisfied himself as to the truth or falsehood of these
statements?' I then gave him to understand that the Government greatly
desired to see all the land which he formerly possessed settled and cultivated
by him."17 Emin carried out his mission successfully and made a settlement
under which Kabalega was guaranteed either an annual grant of money or
presents. Emin offered to escort Kabalega's ambassadors to Cairo, or if
Kabalega wished to go himself, to remain in Bunyoro as a hostage for him.
In addition, he agreed to withdraw Egyptian support from Ruyonga and
Mupena. "All this seemed to please him," wrote Emin, "and he thought at any
rate I was the most reasonable of all whom he had seen . and said 'We
Emin's success compared with the experience of many other Europeans in
dealing with Africans can be attributed to his gift of estimating them at their
EMIN PASHA AND BUNYORO-KITARA 75
right value. His knowledge of languages was a great help and he never lost
his personal ascendency.19 It is a tribute to his diplomacy that no other Euro-
pean ever achieved such cordial relations with Kabalega.20
At the end of his political mission Emin wrote "My official business was
brought to an end to our mutual satisfaction . I shall always remember
with pleasure the days I spent here."21 Kabalega appeared to accept that there
had been mutual provocation and was prepared to let bygones be bygones.
That Emin's hopes of a reasonable modus vivendi between Bunyoro and Egypt
failed to materialize was perhaps due to the Egyptian Administration's con-
tinued support of Ruyonga and Mupena contrary to the terms of agreement
and to the lack of effective supervision over the isolated garrisons in northern
Bunyoro. Emin left Kabalega's court on the 25 October and returned by way
of Londu and Kisuga to Mruli. Almost immediately afterwards he left for
Buganda to assure Mutesa that Egypt had abandoned her plans of a new route
to Equatoria by way of the Indian Ocean and Buganda. On his return journey
in April 1878, he travelled between Mruli and Magungo, taking canoe to
Foweira, and thereafter by land through northern Bunyoro. Mupena's zeriba
at Panyatoli was one of the halting places. "We there met Mupena himself,"
wrote Emin, "who again confirmed the good impression which he has always
made . He is the only negro 'gentleman' with whom I have
become acquainted in my wandering in these regions." Mupena was
chief of the semi-independent districts of Magungo, Chopi as well as of part
In July 1878 Emin was appointed Governor of Equatoria. The following
eighteen months saw the twilight of the Egyptian occupation of northern
Bunyoro which for a period facilitated access by European travellers to this
remote region. This occupation consisted of little more than the maintenance
of the Egyptian forts and-under strong escort-of routes converging on Mruli
from Magungo via Kikoroto, and from Foweira either by road or river. Move-
ment across country was possible through Mupena's territories between
Kikoroto and Foweira, whilst on Lake Albert, Emin's steamers, Khedive and
Nyanza, could visit Magungo and Kibero. In December the Rev. C. T. Wilson,
a member of the first Church Missionary Society party to reach Buganda from
Zanzibar in 1877, journeyed from Rubaga to Mruli and thence down the Nile
to Foweira to meet another party coming from the north by the Nile route to
Magungo. This party, which he met at Panyatoli, comprised Pearson, Litchfield
and Felkin, who recorded that Magungo had been removed from its old site,
since the land at the north-east corner of Lake Albert was gradually being
washed away making the station damp and unhealthy. The new Magungo was
a clean and well-built town, surrounded by a strong earthwork fortification
and a moat ten feet in depth. Kabalega forbade his people to act as porters
and, on their march towards Foweira, attacks and alarms were frequent. The
party were, therefore, glad to reach the protection of Kikoroto's strong stockade
which was situated in a large clearing in an immense forest.23 Led by Wilson,
they travelled to Buganda by way of Mruli in January 1879. In June of that
year, Wilson and Felkin escorted the Baganda ambassadors to Queen Victoria
by way of the Nile, Foweira, Fatiko and Dufile. During these journeys, Felkin
76 UGANDA JOURNAL
made the acquaintance of Emin and subsequently translated and edited his
letters and became the advocate of his cause in Great Britain.
The C.M.S missionaries were followed early in 1879 by Richard Buchta,
an Austrian photographer, who coming by way of Fatiko crossed the Nile at
Foweira into northern Bunyoro, visited the Murchison Falls and left Fort
Magungo by steamer northwards. His photographs must be the earliest
taken in Bunyoro.24
Towards the end of 1878 Gordon decided to evacuate the stations south
of Dufile, since the commanders persisted in raiding Kabalega and this
induced retaliation.25 Emin was opposed to this policy and delayed active
steps he had kept the Danagla in check, had established friendly relations
with tribal chiefs and had interested himself in the development of agricul-
ture, so that under his rule his province was attaining some measure of
But Emin found it prudent to comply with Gordon's directives. In November
1879, the Khedive took him from Dufile to Magungo, where he remained for
some weeks, superintending the withdrawals from the Bunyoro stations; and
it was doubtless at this time that the more distant outposts, Mruli, Koki
(Kodj),26 Kisuga and Kisindizi as well as Foweira were abandoned. The last
to be withdrawn was seemingly the garrison from Kikoroto which came into
Magungo early in December.
Emin spent a few days in the steamer visiting Mahagi on the west shore of
Lake Albert where a station was to be maintained among the Alur, and the
steamer took him back to Dufile by the middle of December 1879.
A responsible messenger from Pearson, the missionary in Buganda, with
mails for Europe reached Mruli towards the end of December but found the
place devastated and deserted, and had perforce to return. In March 1880
Litchfield and Pere Barbot of the White Fathers' Mission, both hoping to
obtain medical aid from Emin tried again, but having collided with Kabalega's
people near Mruli they also returned to Rubaga.
Meanwhile in Egypt, the Khedive Ismail, who had brought his regime to
bankruptcy, was in 1879 deposed and was succeeded by his son, Tewfik.
Shortly afterwards, there was a military revolt in Egypt and troops were with-
drawn from the Sudan, thus leaving the way clear for the Mahdi to seize power
and defeat the few troops that remained.
Gordon had resigned at the end of 1879 and so from 1880 onwards, Emin was
free to re-establish stations south of Dufile. In October he toured the Latuka
country and was present at the reoccupation of Foweira, whence in November
1880 he visited Mupena in north Bunyoro, and it was probably about this
time that the post of Foda on the Nile west of Foweira was built. Foweira and
Foda continued to be occupied until April 1884.
Emin's administration eventually covered fifty stations in Equatoria, three
having been established by Baker, and twelve by Gordon. The subordinate
officials at these stations were under strict orders to stop the slave trade and
to prevent inter-tribal fighting. They were, however, unreliable and morally
deficient for such work. The troops were little to be trusted, being harsh with
the people and insubordinate to their officers. The Governor could not see
EMIN PASHA AND BUNYORO-KITARA 77
everything, and despite the most stringent orders in his absence, slave dealing
and forced labour were connived at.2
In spite of this Emin's energy produced results in the form of a budget
surplus and the cultivation of wheat, coffee, rice, tobacco, nutmegs, indigo
and cotton. His attempts to achieve order out of chaos were not matched in
other provinces of the Sudan, as one by one the European administrators died
or retired. Their Egyptian successors failed to prevent a return to anarchy,
and discontent and unrest increased. A religious fanatic Muhammad Ahmed,
formerly a carpenter, proclaimed that he was the Mahdi, whose coming was
foretold, and established himself on an island in the Nile south of Khartoum.
Government troops were unable to dislodge him, and thousands, inspired by
religious zeal, rallied around him.
The last steamer communication with Khartoum left Lado in March 1883.
The British Government advised the Egyptian Government to abandon the
Sudan and Gordon arrived in Khartoum in February 1884 to evacuate the non-
Sudanese. He was not able to accomplish this and on 26 January 1885
Khartoum fell and Gordon was killed. The Sudan was abandoned and Emin
now completely isolated alone remained to uphold Egyptian authority in
Equatoria. The approach of the Mahdists compelled Dr. Junker, a Russo-
German explorer, who had been continuing Schweinfurth's work on the course
of the Welle28 and the Congo/Nile water-parting, to join Emin at Lado in
1884, and he was followed early in 1885 by Casati, an Italian, who had also
been exploring the Welle some four hundred miles west of Arua.
Casati alleged that Emin said "We, white men, shall escape-I answer for
it. We will give our black soldiers to my good friend Kabalega, king of
Bunyoro, and he will permit us to cross his boundaries."29 These words may
have been remembered by the troops and may account for the difficulties Emin
had later in persuading them to accept Stanley's proposals. Gradually the
Egyptian administration, threatened by the Mahdists from the Bahr el Ghazal
province, retreated to Wadelai. Emin arrived there on the 10 July 1885. Junker
stayed at Foda close to Mupena's headquarters from March to December of
that year, attempting to gain contact with the Europeans in Buganda and the
south. He was hindered by the hostility of Kabalega, who had entered into
trade with the Zanzibar merchants from whom he obtained arms and ammuni-
tion in exchange for ivory and slaves. Consequently he had no desire for "the
augmentation of his sovereign authority, or even the maintenance of it, from
the Egyptians. Kabalega . put pressure on Kamisoa30 . not to let them
re-enter his territory".31 In December, however, Kabalega asked that an
Egyptian Government representative should reside in Bunyoro and he let it
be known that he would permit the transit of correspondence.
As a result, in January 1886, having left Foda, Junker set out from Wadelai,
landed at Kibero and obtained an audience with Kabalega at Mparo after a
delay because of the new moon. He described the king as follows: "Kabalega
was seated on a raised bench, clothed in the national costume, a neatly worked
cowhide, with the ends gathered on his left shoulder. He was in the full vigour
of manhood, of stately presence, unadorned with any ornament, and with his
hair cropped short, as worn by all these people. His bright penetrating gaze be-
78 UGANDA JOURNAL
trayed nothing of the tyrant that he really was."32 At Mparo, Junker en-
countered a Tripoli-Turk merchant, Mohammed Biri, and from him received
first hand news of the course of events in Buganda. Biri returned almost at
once to Buganda. Due to the outbreak of hostilities between Bunyoro and
Buganda, which probably prompted Kabalega's request for a representative,
it was difficult to take advantage of the permission to pass through Buganda
given by Mwanga, Mutesa's successor as Kabaka of Buganda; but Kabalega
finally permitted Junker to travel on to Rubaga. On 26 February 1886 a
quantity of letters from Zanzibar and Buganda which Kabalega had allowed
to pass through his country was delivered to Emin at Wadelai. Among them
was an official notification of Egypt's abandonment of the Equatorial Province,
and suggesting a retreat to Egypt via Zanzibar. Later Junker arrived at
Zanzibar and announced Emin's survival to the world.
Emin was now seriously threatened by a Mahdist force from the Bahr el
Ghazal province. He began to think of a further retreat and he felt that his
most secure refuge would be with Kabalega.33 In May 1886, therefore, he sent
Casati as an ambassador to Bunyoro to come to terms with Kabalega. From
the 31 May to the 6 June, Emin toured Lake Albert arriving at Kibero on
2 June, where he spent some days. In his usual thorough manner he noted
the vegetation and the method of salt extraction.34 35
On the 2 June Casati was publicly received by Kabalega at Bujwahya near
Kasingo, and he gives the following description of the king: "he wore a dress
of elegant woollen cloth, finely worked and ornamented; his head was covered
with a red tarboosh . he was of colossal form, of gigantic stature, with a
smile more sarcastic than amiable".36 Kabalega coveted the ivory and ammuni-
tion of Equatoria and, now that Emin was in difficulties, was disposed to
feign friendship, especially as Bunyoro had been devastated by the Baganda
invasion, smallpox and famine. Kabalega agreed to the carrying of correspon-
dence through his territory and the transit of small parties of troops, but at
once the letters went astray and the troops were delayed. Then began a period
of protracted negotiations with the neighboring rulers to the south, whose
friendship and alliance Emin sought to gain, for he considered that should there
be a determined Mahdist advance from the north his only possible course
would be to retreat to the south via Lake Albert and Lake Tanganyika.37
On his arrival at Rubaga Junker, with the help of Mackay, had arranged
for Mohammed Biri to take goods to Emin; and Biri, conveyed by Emin's
steamer from Kibero, reached Wadelai in October 1886. With another caravan
he arrived at Wadelai again in July 1887, remaining for two months before
returning to Bunyoro.
Meanwhile Casati was ostracized and was virtually a prisoner at Kabalega's
court. Here he learnt of plans to attack Wadelai and to invite Emin to Bunyoro,
preparatory to murdering him. Casati was able to warn Emin, and the steamer
shelled the invading canoes. Casati commented: "Emin did not come, and I
did not move. Our dignity and interests required us to keep a footing in
Bunyoro."38 An uneasy peace resulted. War again broke out between Buganda
and Bunyoro. Kabalega eventually drove the Baganda out of his kingdom but
when harassed by them became increasingly hostile towards Casati.9
EMIN PASHA AND BUNYORO-KITARA 79
In March 1887 Kabalega moved to Buhimba, while Casati remained at
Bujwahya. In May he was accused of conspiring against Kabalega. Kabalega's
perfidy and persuasive charm exasperated him. Nevertheless, because it was
essential to maintain communications with Buganda, Casati remained and en-
tered into further negotiations, either for an alliance confirmed by exchange of
blood or for a concession to occupy Kibero and Kitana.40 This suited Kabalega
who desired an ally because he wished to occupy the lands of his enemies,
Mupena and Komwisa. As soon as he had defeated them, Casati was of no
further use to him, so there was an attempt to murder him. In November
Casati was joined by Biri who had returned from Wadelai and was hoping to
travel on to Buganda with a large quantity of his own and Emin's ivory. On
3 January 1888 rumours of the arrival of Europeans (whom they correctly
assumed to be Stanley's Relief Expedition) to the west of Lake Albert reached
them. This caused Kabalega to break off negotiations. Casati and Biri were
seized and their possessions rifled. Casati was permitted to escape to Kitana
and then Kibero. But Mohammed Biri was not immediately released, and
some months later was killed by the Banyoro.41
On the 16 January, Casati was rescued near Ndandamire by Emin in the
steamer. Casati was blamed for the failure of his mission, but Emin added:
"that the matter would have no serious consequences, that Kabalega had acted
only from personal aversion, that intercourse with him was about to be
resumed by the despatch of a more acceptable ambassador". This never hap-
pened and after Casati no European set eyes upon Kabalega until his capture
in south Lango in April 1899. Casati wished to retaliate but Emin told him
"he ought to be quite satisfied with having got out of the scrape alive, and the
most urgent matter was to resume friendly intercourse with Bunyoro" to ensure
"the transit of correspondence through Buganda".42
Emin had, in fact, been somewhat anxious about Casati. He wrote: "I have
seldom met so true and loyal a man. But, to judge from his last letters, he
seems to be rather at variance with Kabalega. The cause appears to me to be
that Casati is too candid with the king, and expects that Kabalega will re-
nounce his dodges and subterfuge, and treat him honestly and straightforwardly
S. .I apprehend that some day serious misunderstanding may arise."43
As Equatoria became self-supporting it became isolated. Junker, arriving
at Zanzibar in 1886, gave authentic news to Europe that Emin was still at his
post. He immediately became a hero who must be rescued whatever the cost.
Public opinion was to change later. When the rescue had been achieved and
his story told for him by others, his name became associated with ingratitude
The Emin Pasha Relief Expedition was under the command of H. M.
Stanley who decided to travel by way of the Congo; his ostensible reasons
were political, for his passage through East Africa was opposed by both the
French and Germans who were at that time suspicious of British intentions
in that part of Africa. It is also possible that he had personal or private reasons
for travelling by this route. Whatever the cause, he was criticized for it.44 Casati
wrote: "if the Expedition had followed the route (from the East Coast) pro-
posed by Felkin, and approved by Junker, as more feasible, shorter and less
80 UGANDA JOURNAL
perilous, it would have gained the lake in a shorter time, and with forces
nearly intact. The return journey to Bagamoyo of the column under Stanley
effected by this itinerary proves the truth of this assertion".45 On the 29 April
1888 at Nyamsassi on the west shore of Lake Albert there was an anti-climax;
the rescuer and his officers, in stained, shabby clothes and broken boots,
with virtually naked followers, were greeted by a slim neat man in a well
fitting white cotton suit, and accompanied by a bodyguard as well turned out
as himself. Stanley described the scene: "I expected to see a tall thin military-
looking figure, in faded Egyptian uniform, but instead of it I saw a small
spare figure in a well-kept fez and a clean suit of snowy cotton drilling, well-
ironed and of perfect fit. A dark grizzled beard bordered a face of a Magyar
cast, though a pair of spectacles lent it somewhat an Italian or Spanish
appearance."46 Casati wrote: "Emin did his best to supply shoes, linen, tobacco,
salt, honey, grain and simsim to the Expedition, equipped and sent out from
England to his aid."47
The Khedive's instructions, presented by Stanley, unmistakably renounced
the province, ordered the withdrawal of the Governor and his followers, but
did permit them to remain in isolated independence without help from Egypt.
But Stanley's Expedition had only brought thirty-four cases of ammunition,
two bales of half-spoilt clothes, and the Khedive's letter.48 The remainder of
the relief stores had either been lost or left with the rear party. To Stanley's
proposal that the province should be attached to the Congo Free State, Emin
replied that it would be impracticable to maintain communications through
an area that had nearly destroyed the Relief Expedition. Stanley then put
forward another proposition that a British Protectorate should be established
near Lake Victoria within reach of Zanzibar. Emin had himself suggested a
similar plan to Egypt on a previous occasion and the proposal seemed to afford
a possibility of preserving what he had accomplished for civilization, and cf
resuming and furthering that work in his province from a secure base well
connected with the outer world.49 He felt that it would be criminal to abandon
the tribes who had submitted to his government and who had helped it so
much. Before he met Stanley he had written "if a relief expedition comes to
us, I will on no account leave my people. We have passed through troublous
times together, and I consider it would be a shameful act on my part were I to
desert them".50 Stanley then went back to the Congo to fetch his rear column,
while Emin returned to his province accompanied by Mounteney-Jephson, one
of Stanley's officers, to sound his men about the withdrawal. Emin's soldiers
were already out of hand and in a rebellious mood. On the 30 May Sudanese
troops crossed Lake Albert to sack Kibero. This foray was practically the last
intercourse between Egypt and Bunyoro.51
Emin's soldiers had no wish to be rescued. They had households of women,
children and slaves, and were virtually immobile. They had heard of the losses
which the Relief Expedition had endured. There was little in Stanley's offer
to attract men born and bred in the southern Sudan. Mutiny broke out in
August, and Emin and Jephson, while trying to persuade them to retreat, were
made prisoners at Dufile and were only released and sent to Wadelai when the
Mahdists attacked the fort. Emin's influence was lost but fear of the Mahdists
EMIN PASHA AND BUNYORO-KITARA 81
made some of the rebellious troops think that it might be better to withdraw.
After weeks of delay, evacuation began, but since Emin had lost his army
the proposed Protectorate near Lake Victoria could not now be established.
Faced with Emin's inability to persuade his troops to evacuate, all Stanley
could do was to confess failure or to rescue Emin against his will; and rescue
him he did. Emin was too oriental and too polite to explain that he had never
asked to be rescued; that all he had wanted was to maintain his post; and that
now he had to relinquish it because Stanley, through choosing the wrong route,
had been forced to abandon the stores that would have permitted him to
remain in Equatoria. Stanley and Emin represented two opposite extremes.
Stanley was bluff, coercive, practical, self-reliant and tactless: Emin was con-
ciliatory, pliable, scientific and reticent. Before the stronger personality Emin
On the 10 April 1889 the retreat began and civilized rule in Equatoria
ceased. The journey was not a happy one and was only noteworthy for the
discovery of Ruwenzori, the Mountains of the Moon. The route was west of
Ruwenzori, across the Semliki to Katwe, thence to Lake Victoria and
Zanzibar from where Stanley returned to Europe alone, leaving Emin in the
German service in East Africa. Emin had received many offers of employment
but he accepted that of the Germans because it gave him a chance to remain
in the Africa that he loved.
During the ten years of his rule, Emin had governed a country 375 miles
from east to west, and 275 miles from north to south, with little help save that
of a band of Egyptian officers, half of whom were felons sent to Equatoria
instead of to prison. Emin's temperament and these circumstances made him
rule by conciliation rather than by coercion, and his subjects acknowledged
his patriarchal authority which was the one authority, apart from force of arms,
that they understood. "His kindly disposition and sense of justice, no less than
his ardour as an explorer and his diplomacy, impelled him to keep everything
in view, and above all, to make a close study of the natives . their feelings
and their thoughts, their habits and tribal laws."52 It was as the father of his
people that he cured the sick and instructed the sound in scientific, agricultural,
and practical matters. His spare time was spent in scientific research.53
Strangely enough, Emin wrote little about the health of the peoples over whom
he ruled. Perhaps his scientific studies were a means of relaxation and escape
from his daily practice of medicine.
Controversy continued in Europe and all concerned rushed into print with
the exception of Emin, who was already hastening back to the province from
which he had been so expensively detached. But Emin found service with
the German authorities uncongenial. In February 1891 he left his station at
Bukoba on Lake Victoria, planning to regain touch with his old province and
to cross Africa to the west coast as a freelance scientist. On nearing Equatoria,
however, he learnt that, though the bulk of the Sudanese remained loyal,
disorder and chaos prevailed, and his party was too small to attempt any-
thing. In any case the territory was now within the British sphere of influence.
Sickness and famine dogged the party and when smallpox struck, Emin ordered
Stuhlmann, his second in command, to retreat to safety with the stores and
82 UGANDA JOURNAL
the healthy people, while he remained with the sick. Continuing westwards, he
was murdered on the 23 October 1892 by some Arab slave traders. The motive
was, apparently, by the murder of a European to commit wavering Arabs to
resist approaching Belgian forces.
So died the one man who might have been able to persuade Kabalega,
Mukama of Bunyoro-Kitara, to come to terms with Western civilization and
thus preserve the dominion and the hegemony of Kitara in Uganda. Emin
was remarkable amongst Europeans for his compassion towards, and under-
standing of, Africans, and the circumstances outside his control which pre-
vented him from achieving his ideals have had far reaching results on the
development of Bunyoro-Kitara and the Uganda Protectorate.
The assistance given by R.A. Sir Tito Winyi, C.B.E., Omukama wa Bunyoro-
Kitara, Mr. J. W. Nyakatura, and Mr. L. M. Muganwa in reconciling certain
personal and place names, and the assistance given by H. B. Thomas, O.B.E.,
in suggesting certain references, is gratefully acknowledged.
1 The second of three articles on the early European travellers in Bunyoro-Kitara.
see Dunbar, A. R., European Travellers in Bunyoro-Kitara, 1862 to 1877, Uganda J.,
23 (1959), 101. From the Uganda Essay Competition 1958.
2 Bunyoro is the remnant of the ancient and extensive kingdom of Kitara, which
stretched southwards to the River Kagera, south-eastwards over Singo, Bulemezi, Buruli
and Bunyara (Bugerere), eastwards and northwards to the Victoria Nile, and westwards
beyond the River Semliki and Lake Albert. Thus Kitara refers to the full extent of the
kingdom and Bunyoro to the present district which has well defined geographical
boundaries, the Rivers Nile and Kafu/Nkusi and Lake Albert. R.A. Sir Tito Winyi.
C.B.E., Omukama wa Bunyoro-Kitara, personal communication.
3 Schweinfurth, G., et. al. (1888). Emin Pasha in Central Africa. Introduction, p. xii.
4 Dunbar, op. cit.
5 The spelling of names of people and places in Bunyoro-Kitara is taken from current
official maps, and Nyakatura J. (1947). Abakama ba Bunyoro-Kitara. St. Justin, Canada.
6 Emin refers to Mparo as Mpara-Nyamoga. There is only one Mparo in Bunyoro,
the burial place of Kabalega near Hoima. Nyamoga is part of the Mparo area. R.A. Sir
Tito Winyi, C.B.E., Omukama wa Bunyoro-Kitara, personal communication. Another
theory is that Mpara Nyamoga and Mparo are different places. See Thomas, H. B. (1932).
Chronological Notes on the Historical Geography of the present Bunyoro (Provisionally
Revised), District Commissioner, Bunyoro's File entitled: Bunyoro: Historical places.
7 Schweitzer, G. (1898). Emin Pasha: His Life and Work, vol. i, pp. 43-4.
8 Schweinfurth, et al., op. cit., pp. 52-9 and 111-12.
9 See U.J., 16 (1952), 96. The exchange value of the thaler in 1877 against the pound
sterling was 441 pence. The Manager, The British Linen Bank, London, personal
10 Schweinfurth, et al., op. cit., pp. 114-23.
12 Katama, R.A. Sir Tito Winyi, C.B.E., Omukama wa Bunyoro-Kitara, personal
13 Baker was unable to speak Runyoro and depended upon interpreters for informa-
tion. These interpreters may have misled him from personal, clan, or tribal reasons and
sometimes have unintentionally misinformed him through lack of proficiency. Emin
was able to speak Runyoro.
14 Schweinfurth, et al., op. cit., pp. 60-3.
15 Natives of Dongola, that is Sudanese irregular troops, or Nubians. Nubians became
a name for the Sudanese troops, even though they were composed of men recruited from
within Uganda, the Southern Sudan, or the Congo.
16 Should read Ruyonga, Mupena, and his deceased father Mpuhuka.
17 Schweitzer, op. cit., vol. i, pp. 48-9.